keentelengineering

Inside the PJM Rulebook: Markets, Governance & Interconnection

Wed, 15 Jul 2026 08:29:50 GMT

Jul 15, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Inside the PJM Rulebook

Challenge : Frequent false tripping using conventional electromechanical relays
Solution: SEL-487E integration with multi-terminal differential protection and dynamic inrush restraint
Result: 90% reduction in false trips, saving over $250,000 in downtime

1. Why the Rulebook Matters as Much as the Rules

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

How PJM Markets, Governance, and the Stakeholder Process Shape Interconnection Outcomes — and How Developers and Large Loads Can Engage

PJM Interconnection, L.L.C. is the regional transmission organization (RTO) and independent system operator (ISO) for all or part of thirteen states and the District of Columbia — spanning Pennsylvania, New Jersey, Maryland (the three states that give PJM its name), Delaware, Ohio, Michigan, Illinois, Indiana, Virginia, West Virginia, North Carolina, Kentucky, and a small portion of Tennessee. Headquartered in Valley Forge, Pennsylvania, PJM operated for decades as a utility power pool before becoming the nation's first fully functioning RTO in 2002. It is now the largest wholesale electricity market operator in North America by load served.

Three distinctions define what PJM actually does, and each one matters for how you engage with it:

Operator, not owner — PJM does not own the grid. Transmission-owning utilities that join PJM retain ownership of their lines, substations, and equipment, but transfer functional operational control to PJM. These participating transmission owners are compensated through PJM's Open Access Transmission Tariff when their facilities are used. This is why your interconnection project deals with both PJM (studies, agreements, queue administration) and a transmission owner (facility design standards, construction, cost estimates) — two parties, two sets of requirements, one project.
Wholesale, not retail — PJM operates the high-voltage bulk power system under FERC jurisdiction. Lower-voltage distribution facilities — generally 100 kV and below, though the precise boundary follows the facilities designated under the transmission owners' agreements — remain under state or local jurisdiction. This jurisdictional seam is exactly where many distributed generation, community solar, and behind-the-meter projects live, and it dictates whether your interconnection runs through PJM's process or a state-jurisdictional utility process.
Independent and non-discriminatory — PJM is a non-profit entity whose core mandate, under FERC oversight, is non-discriminatory grid access: any qualified buyer or seller of wholesale electricity connects and transacts under the same rules. That neutrality is also why PJM's rules change only through a structured, transparent stakeholder process — which is the subject of this article.

Keentel Perspective

The operator/owner split is the single most underestimated fact in PJM interconnection. Developers routinely budget for the PJM study process and are then surprised by transmission-owner facility requirements, design standards, and interconnection facilities costs. Treat the transmission owner as a full counterparty from day one — in your schedule, your single-line development, and your cost model.

Most developers, independent power producers, and large-load customers experience PJM Interconnection the same way: as a set of requirements that arrive fully formed. A model quality checklist. A study deposit schedule. A ride-through obligation. A capacity accreditation methodology. The rules land on your project, and your job is to comply.

But every one of those requirements was once a problem statement in a committee room. It was debated, redrafted, voted on by five stakeholder sectors, approved by an independent board, and filed with the Federal Energy Regulatory Commission (FERC) before it ever appeared in a tariff or a business practice manual. Understanding that machinery — how PJM is governed, how its markets are structured, and how its rules are actually made — is not academic. It determines when new requirements will hit your development pipeline, how much lead time you will have, and whether you have any voice in shaping them.

At Keentel Engineering, we have long argued that grid interconnection is a first-order engineering input, not a downstream administrative step. The same logic applies one level up: the regulatory and stakeholder processes that produce interconnection requirements are themselves a first-order planning input. A developer who understands the PJM stakeholder pipeline can see requirements coming twelve to twenty-four months before they bind.

One who does not will discover them in a study report, a deficiency letter, or a capacity auction result.

This article walks through PJM's role, jurisdiction, market architecture, governance structure, and rulemaking process — and then connects each of these to the practical decisions that generation developers, storage owners, and large electronic loads face today. A detailed FAQ follows in Part II.

2. What PJM Is and What It Is Not

3. The Market Architecture: Four Mechanisms, One Balance

5. How a PJM Rule Is Actually Made

About Keentel Engineering

Keentel Engineering is a power systems and grid interconnection consulting firm headquartered in Tampa, Florida, with offices in Austin, Sacramento, and Baltimore. Our service lines span grid interconnection engineering, substation and transmission design, power system studies NERC compliance, renewables and BESS engineering, EMT modeling, and owner's engineer services. We serve developers, independent power producers, utilities, and large-load customers across North American RTO/ISO footprints.

Part 2 — Frequently Asked Questions: Large Load Interconnection

The design intent is straightforward: no sector can dominate by turnout. If thirty end-use customers attend a vote against six members from each other sector, those thirty votes still control only the end-use sector's twenty percent. Coalition-building across sectors is therefore not optional — it is the arithmetic of how anything passes. A proposal needs broad, cross-sector support to reach the two-thirds threshold, which is why PJM rule changes tend to be negotiated packages rather than clean wins for any one interest.

Around the voting structure sit several bodies that shape outcomes without voting. The Organization of PJM States (OPSI) represents the utility regulatory commissions of the fourteen PJM jurisdictions, coordinating state positions and advising PJM on regulators' concerns. The Consumer Advocates of the PJM States (CAPS) represents the state consumer advocate offices — and is unique among RTO consumer organizations in being funded through a FERC-authorized tariff mechanism. User groups such as the Public Interest and Environmental Organizations User Group (PIEOUG) give non-member organizations structured access to PJM staff and the Board. FERC itself holds regulatory oversight over the entire structure and may place non-voting representatives on standing committees.

PJM is overseen by an independent Board of Managers consisting of ten members — nine elected members plus the PJM CEO serving ex officio. Members serve staggered three-year terms and must collectively bring expertise across utilities, finance, regulation, engineering, and markets. The Board's independence is structural: it is the body that decides whether stakeholder-endorsed rule changes are filed with FERC.

Beneath the Board sits the stakeholder apparatus where the substantive work happens. Two Senior Standing Committees anchor the structure: the Members Committee, composed of representatives of all PJM members and reporting directly to the Board, and the Markets and Reliability Committee, which consolidates input from the technical committees covering markets, operations, and planning. Both senior committees require a two-thirds supermajority to pass a proposal.

Voting in the senior committees uses sector-weighted voting across five member sectors, each carrying exactly twenty percent of the total vote regardless of how many members show up:

An electric grid must remain in continuous balance — generation onto the grid must equal consumption from it at every instant. PJM achieves this balance, and prices it, through a layered market architecture. Each layer operates on a different time horizon, and each one touches project economics differently.

Two features of this architecture deserve emphasis. First, security-constrained economic dispatch means the market model respects transmission limits: a higher-cost resource near a load center can clear ahead of a cheaper remote one when congestion binds. Location is priced, which is why interconnection point selection is an economic decision, not just a feasibility question. Second, the Reliability Pricing Model (RPM) capacity construct is a planning instrument, not just a revenue stream — its three-year-forward design (compressed in recent auction cycles) exists to signal where investment in new supply is needed. When capacity prices in a locational deliverability area spike, that is the market telling developers where the next tranche of projects will pencil.

Keentel Perspective

Market design and interconnection engineering are converging. Capacity accreditation now depends on demonstrated performance characteristics; ride-through obligations under IEEE 2800 and NERC PRC-029 determine whether an inverter-based resource stays online during the disturbances that capacity commitments are meant to cover; and EMT model quality determines whether PJM believes your plant does what your pro forma assumes. The study models you submit are, increasingly, market documents.

4. Governance: Who Actually Decides

The path from idea to enforceable tariff language follows a defined pipeline. Any stakeholder — including a non-member — can initiate it. The stages below are where a developer's engineering and commercial teams should be paying attention:

Stage 1: Problem statement and issue charge — A stakeholder brings a problem statement and issue charge to a technical committee. If approved, PJM and stakeholders develop a work plan: which committee owns the issue, and whether the end product is manual language or tariff revisions. This is the earliest signal — problem statements filed today describe the compliance obligations of two years from now.
Stage 2: Solution development — PJM typically develops a solutions package with options; other stakeholders may sponsor competing solutions. Technical debate happens here — modeling requirements, study assumptions, cost allocation formulas. This is where engineering input has the most leverage, because positions harden as packages advance.
Stage 3: Senior committee votes — A supported solution advances to the Markets and Reliability Committee and then the Members Committee, each requiring a two-thirds sector-weighted supermajority. Members vote for, against, or abstain.
Stage 4: Board action and FERC filing — The PJM Board decides whether to file the endorsed change with FERC under Section 205 of the Federal Power Act. Notably, the Members Committee can also vote (by two-thirds supermajority) to file a Section 205 change even without Board approval — and any individual or group can file a Section 206 complaint at FERC asserting that an existing PJM rule is unjust and unreasonable, with no committee approval required at all.

Why This Pipeline Matters to Developers

Nearly every major interconnection reform of recent years — cluster study transition, expedited interconnection pathways, EMT model requirements for inverter-based resources, large-load interconnection terms, capacity accreditation reform — moved through exactly this pipeline before it bound anyone. The stakeholder calendar is a forward schedule of your future compliance obligations. Monitoring it is cheap; discovering a new requirement mid-queue is not.

6. Participation Is Open and Underused

PJM stakeholder meetings are open to the public unless otherwise noted, with remote participation routinely available. Agendas, presentation materials, and issue-tracking histories are posted through PJM's Meeting Center, and non-members may attend and ask questions. Full membership — which carries voting rights — requires an application and an annual fee (currently $5,000 for most member classes), a modest sum against the value of a vote in processes that set study deposits, cost allocation, and performance obligations.

For most developers and large-load customers, the practical engagement model is layered: monitor the committees relevant to your asset class; coordinate with sector coalitions that share your position (experienced stakeholders rarely go it alone); engage your state commission through OPSI channels where state policy is implicated; and reserve direct membership and voting for organizations with sustained PJM exposure. The issues are technical, the history runs decades deep, and the language is tariff language — which is precisely why engineering-literate participation is disproportionately effective.

7. What This Means Right Now: The Issues Moving Through the Pipeline

The stakeholder process is currently the arena for the questions that will define PJM development economics for the rest of the decade:

Large loads and co-location — Data center growth has made large-load interconnection the defining reliability question in PJM. How co-located and grid-connected large loads are studied, charged, and obligated to ride through disturbances is being worked out in stakeholder forums now — in parallel with NERC's large-loads reliability work and ERCOT's NOGRR282 precedent on electronic load ride-through.
Inverter-based resource performance — PJM's EMT modeling requirements for inverter-based resources — PSCAD model quality, benchmarking against field behavior, and validation checklists — trace directly to NERC disturbance analyses and IEEE 2800 adoption. Model quality expectations continue to tighten, and resources with unbenchmarked or vendor-generic models face study delays and restudy risk.
Interconnection process reform — Queue reform continues to evolve, including expedited pathways for shovel-ready projects. Eligibility criteria and readiness deposits are stakeholder-negotiated parameters — meaning they can and will change as the backlog picture changes.
Capacity accreditation and RPM reform — Capacity market rules — accreditation methods, must-offer obligations, penalty structures — are under near-continuous revision following recent auction outcomes and FERC proceedings. For storage and renewables, accreditation methodology is often worth more to the pro forma than any energy-market design change.

8. How Keentel Engineering Fits In

PJM is not a black box. It is a transparent, rule-bound institution whose markets balance the largest grid footprint in North America and whose requirements are written in open committee rooms by the very sectors they govern. For developers, storage owners, and large loads, that transparency is an asset — but only for those who use it. Understand the market architecture, watch the stakeholder pipeline, engage where your interests are at stake, and build your engineering program to anticipate the rules rather than react to them.

Keentel Engineering supports clients across every stage of that program — from the first single-line diagram to the final compliance filing. Part II below answers the questions we hear most often.

Frequently Asked Questions

Keentel Engineering is a power systems and grid interconnection consulting firm headquartered in Tampa, Florida, with offices in Austin, Sacramento, and Baltimore. We work at exactly the intersection this article describes: where PJM's rules, studies, and stakeholder-driven requirements meet the engineering of real projects. Our service lines map to the PJM landscape as follows:

The through-line in all of it is the thesis we opened with: interconnection is a first-order engineering input. The developers who win in PJM treat the tariff, the stakeholder calendar, and the study models as design constraints from day one — not as paperwork at the end. Keentel exists to make that posture practical.

9. Conclusion

Disclaimer :

Keentel Engineering is an independent consulting firm and is not affiliated with, endorsed by, or sponsored by PJM Interconnection, L.L.C., the Federal Energy Regulatory Commission (FERC), the North American Electric Reliability Corporation (NERC), the Organization of PJM States (OPSI), Consumer Advocates of the PJM States (CAPS), or any other organization referenced in this document. All product names, standards, and trademarks are the property of their respective owners and are referenced for identification purposes only. This document summarizes publicly available regulatory and market-process information as of the date of preparation; market rules, tariff provisions, fees, and stakeholder procedures change frequently and readers should verify current requirements directly with PJM and FERC. Nothing herein constitutes engineering, legal, financial, or regulatory advice for any specific project or proceeding.

Gigawatt-Scale Loads & Grid Interconnection Engineering Guide

Tue, 14 Jul 2026 16:16:52 GMT

Jul 14, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Gigawatt-Scale Loads on the Bulk Power System

About This Publication

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

Behaviors, Capabilities, and Limitations of Data Centers, AI Training Facilities, Cryptocurrency Mining, and Hydrogen Electrolysis — An Interconnection Engineering Perspective

1. A New Class of Grid Citizen

For a century, power system planning treated load as the predictable side of the ledger. Generators were the complex machines that needed models, studies, and performance requirements; load was an aggregate — millions of motors, lights, and appliances whose statistical behavior smoothed itself out. That assumption has now been broken, decisively, by a new class of facility: single-site loads of 500 MW to 1.5 GW built around power electronics rather than iron and copper.

The July 2026 ESIG Large Loads Task Force report on large-load behaviors, capabilities, and limitations — one of eleven reports produced by that task force — confirms what interconnection engineers have been seeing in study queues for the past two years: the majority of forecast load growth over the next five years comes from hyperscale data centers, AI training campuses, cryptocurrency mining facilities, hydrogen electrolysis plants, and large manufacturing. These facilities do not behave like the loads embedded in legacy planning models. They ramp in seconds. They trip in milliseconds. They respond to token prices, electricity prices, and training-job schedulers rather than to weather and time of day. And because they interface with the grid through rectifiers, uninterruptible power supplies, and DC/DC converters, they bring converter dynamics — the same class of dynamics that reshaped generation interconnection over the past decade — to the demand side of the meter.

At Keentel Engineering, we have long argued that grid interconnection is a first-order engineering input, not a downstream administrative step. Nowhere is that more true than for gigawatt-class loads. A facility's internal power distribution architecture, UPS topology, cooling drive selection, and protection philosophy are made years before energization — and every one of those decisions determines how the facility will behave during the grid fault it will inevitably experience. This article walks through the behaviors, capabilities, and limitations of the three load classes reshaping interconnection queues — data centers (including AI campuses), cryptocurrency mining, and hydrogen electrolysis — and translates them into the study, modeling, and compliance work that developers and utilities must now perform.

2. Four Ways Large Electronic Loads Threaten Operational Security

Industry analysis converges on four distinct problem classes that large power-electronic loads introduce to bulk power system operation. Understanding which of the four applies to a given facility is the first step in scoping an interconnection study correctly.

The first is speed. Modern electronic loads can move hundreds of megawatts in seconds, and the power-electronics interface between the load and the grid introduces control-loop dynamics that legacy load models never contemplated. Highly variable AI training workloads, aggregated across a fleet of campuses, can consume balancing reserves that were procured against far slower load behavior.

The second is trip sensitivity. Electronic loads are protective by design — their protection settings exist to save servers, not to support the grid. When a transmission fault depresses voltage across a wide area, many large loads can reduce demand or disconnect simultaneously. Losing gigawatts of load in an instant produces the mirror image of a generation trip: over-frequency and over-voltage conditions with genuine cascading potential.

The third is forecastability. Cryptocurrency miners chase real-time energy prices. AI training facilities start and stop billion-parameter jobs at a scheduler's discretion. Neither behavior maps onto the weather-and-calendar regressions that underpin conventional short-term load forecasting, which complicates everything from reserve procurement to day-ahead unit commitment.

The fourth is power quality — harmonics and oscillations. Rectifier front ends, power factor correction stages, and variable frequency drives all emit harmonics and, under the wrong impedance conditions, can become the source of forced or converter-driven oscillations that interact with nearby synchronous generators. As we discuss below, these are no longer theoretical: they have been measured, replicated, and corrected at operating facilities.

Four electrolyzer technologies define the design space. Alkaline units are the mature workhorse, preferred at stable baseload, with liquid electrolytes that slow their response. PEM units use solid polymer membranes and respond fast enough to track renewables and provide frequency service. Solid oxide cells run hot on steam with strict thermal management, and anion exchange membrane technology aims to combine PEM-class response with alkaline-class capital cost. The operating envelope follows the chemistry: PEM ramps on the order of 1 to 20 percent of DC current per second, alkaline more like 0.5 to 2 percent per minute; warm starts recover full load in roughly 5 to 30 minutes while cold starts run from a few minutes for PEM to an hour or two for alkaline; and alkaline plants carry a hard minimum turndown — typically 5 to 25 percent — below which hydrogen-oxygen crossover creates an explosive safety limit, not merely an efficiency preference.

Demonstrations have established that electrolyzers can deliver real grid services: sub-second setpoint following, primary frequency response in both directions, synthetic inertial response, and reactive power support have all been shown in laboratory hardware-in-the-loop environments and at operating megawatt-scale plants. Whether those capabilities are compensated as ancillary services or mandated through interconnection requirements is a jurisdictional question — but the technical capability exists, and facility control hierarchies can be architected for it from day one.

Ride-through is the open frontier. Unlike computer loads with their de facto ITIC basis, electrolyzer facilities have no unified ride-through standard; capability is limited by the electrochemical process behind the stack rather than by the converters, jurisdictional requirements vary, and published example envelopes — riding through deep sags for tenths of a second and wide frequency bands for defined durations, with post-fault recovery ramps around 10 to 15 percent per second — are illustrative rather than prescriptive. Supplementary storage can extend capability at a project-specific cost. Interconnection agreements for hydrogen must therefore negotiate ride-through explicitly, informed by the specific stack technology, rectifier design, and safety chain — a genuinely bespoke engineering exercise on every project.

7. What This Means for Interconnection Engineering

Pull the threads together and a clear engineering agenda emerges for any party interconnecting — or hosting — a large load.

Modeling must reach inside the fence. Positive-sequence dynamic models remain necessary for planning studies, but they cannot capture rectifier PFC instability, harmonic interaction with nearby inverter-based resources, or sub-cycle trip behavior. For large facilities, especially in low short-circuit-ratio regions, validated EMT models representing the actual power conversion chain — UPS stages, sidecar converters, solid state transformers, VFD-driven cooling, rectifier plants — are becoming the standard of care, exactly as they did for inverter-based generation. Where the facility transfers to backup generation, minimum run times belong in restoration modeling.

Performance requirements must be written from the grid's perspective. The ITIC curve is a design convention, not a reliability standard. Utilities and ISOs are moving to dedicated large-load ride-through requirements, harmonic limits enforced at the POI under IEEE 519, oscillation screening, and telemetry obligations including PMU-class monitoring. Developers who engage these requirements at design — selecting transformer vector groups, specifying VFD ride-through parameters, sizing supplementary storage, architecting control hierarchies — interconnect faster and cheaper than those who retrofit.

Flexibility should be engineered, not assumed. AI training, crypto mining, and electrolysis each offer genuine flexibility, but bounded by contracts, chemistry, and economics respectively. The honest path is to characterize what the facility can actually deliver — ramp rates, notice requirements, minimum loads, restoration times — and embed it in the interconnection agreement and market registration.

Keentel Engineering works both sides of this problem: for developers, we prepare interconnection applications, facility one-lines, PSS/E and PSCAD/EMTDC models, ride-through assessments, harmonic and oscillation studies, and NERC compliance programs; for utilities and owners, we serve as owner's engineer reviewing large-load interconnection requests, drafting performance requirements, and validating the models submitted to us. Gigawatt-class loads are the defining interconnection challenge of this decade. The engineering to integrate them reliably exists — it simply has to be applied early, and applied well.

Work With Keentel Engineering

Planning a data center, AI campus, mining facility, or hydrogen plant interconnection — or receiving one into your system? Contact our power systems team at contact@keentelengineering.com or (813) 389-7871 to scope the studies, models, and compliance program your project needs.

Keentel Insight

These four problem classes map directly onto four study products: dynamic and EMT modeling for speed, ride-through and protection coordination review for trip sensitivity, operational load characterization for forecastability, and harmonic/oscillation screening for power quality. An interconnection scope that omits any of the four is incomplete for a facility above roughly 100 MW.

3. Inside the Modern Data Center Power Chain

A large data center connects to the transmission system and steps utility voltage down through a familiar chain: a main power transformer to a medium-voltage network around 11 kV to 15 kV, medium-voltage distribution throughout the building with backup generators paralleled to those buses, and a final transformation to roughly 400 V AC at the server rack level. What makes the facility electrically interesting is not the transformers — it is everything between the medium-voltage bus and the silicon.

Nearly all compute and power equipment inside the facility is designed to the ITIC (formerly CBEMA) voltage tolerance curve, which defines the envelope of voltage magnitude and duration that IT equipment must withstand without damage or shutdown. When grid voltage exits that envelope, the facility transfers to its UPS or battery backup, and if the disturbance persists, to on-site generation. A facility with no backup will simply disconnect along the ITIC boundary.

Here is the problem, and it is the single most important concept in large-load interconnection today: the ITIC curve was written to protect computers, not the grid. Its voltage tolerances are misaligned with the disturbance profiles the bulk power system actually produces. A normally cleared transmission fault — well within the planning criteria every utility studies — can push voltage outside the ITIC envelope long enough to shed an entire campus. The load "worked as designed," and the grid inherited a gigawatt-scale contingency it never studied. This is precisely why dedicated large-load voltage ride-through requirements, written from the grid's reliability perspective rather than the server's, are emerging across the industry — and why ERCOT's NOGRR282 and NERC's developing large-load work matter so much to project schedules.

Compounding the misalignment is opacity. Data center operators and their vendors treat IT equipment specifications as confidential, so the interconnecting utility frequently cannot determine the voltage level at which the facility will actually disconnect. In our study practice, the ITIC curve serves as a reasonable default proxy for fault response modeling — but the modeler should always request the facility's actual ride-through characterization, and interconnection agreements should compel its disclosure.

3.1 The UPS: A Small Converter Plant Hiding Inside the Load

The double-conversion UPS is the workhorse of enterprise, colocation, and hyperscale facilities. Its front-end rectifier regulates a DC bus and manages grid-side power factor; its back-end inverter feeds the data center bus, buffering the load from grid-side distortion; a battery hangs on the DC bus, always ready. Three operating modes — normal, eco/bypass, and battery — are transferred between seamlessly. UPS runtimes vary with the business served: on the order of one to two minutes at hyperscale facilities, around five minutes for cloud and colocation, and ten to fifteen minutes where financial customers demand it.

Under a grid voltage sag, the DC bus sags with it. Because the IT load is constant-power, current rises as voltage falls — and if the rectifier hits its current limit, it blocks and shuts down, transferring the facility to UPS power. Recovery is asymmetric: a facility riding on its UPS can return to the grid within seconds, but a facility that has transferred to backup generation cannot. Backup generators carry minimum run times, typically twenty to thirty minutes, which means the load does not come back when the fault clears. Operators and balancing authorities must plan restoration around that constraint.

AI training campuses increasingly skip the UPS altogether for IT load. Training jobs checkpoint their state to protected storage; if power fails, the run restores from the last snapshot, and the economics of a campus-scale UPS simply do not justify the training progress it would save. The grid-facing consequence is a facility with even less inherent ride-through than its enterprise cousins — a fact that must be represented honestly in study models.

3.2 Documented Oscillations: The Load as a Source, Not a Victim

The most consequential field evidence in the recent industry literature concerns oscillations that originate inside data centers. In one documented test program, an operator brought a full complement of servers to load and then idled them synchronously. At high load, a fleet of forty UPS units arranged in twenty pairs began to oscillate, with the onset point governed by feed inductance and unit loading. Removing one UPS pair from the utility feed — reducing the effective impedance seen by the remaining pairs — stopped the oscillation. The behavior was replicated in simulation and in factory testing with the manufacturer by representing the utility-side inductance, and the permanent fix was a reduction in the loop gain of the rectifier's power factor correction circuit to restore adequate phase margin.

The same operator later encountered a second, distinct instability in a power supply PFC stage — this one appearing at sites both with and without UPS units, at much higher frequency, in the ninth-to-eleventh harmonic range, and resolved through a vendor firmware change to the rectifier control. In a separate case, oscillations were triggered not by anything inside the facility but by the utility removing one of two parallel transformers, raising the source inductance feeding the UPS fleet.

Three lessons follow. First, converter-driven instability on the demand side is real, measured, and sensitive to grid impedance — meaning it is a function of the interconnection, not just the equipment. Second, models that omit the UPS PFC stage cannot screen for the phenomenon; positive-sequence phasor-domain models alone are structurally blind to it, which is why EMT representation of the power conversion chain is becoming a study requirement for large facilities in low short-circuit-ratio areas. Third, high-resolution monitoring — phasor measurement units and digital fault recorders at the point of interconnection and within the facility — is the only way operators gain situational awareness of these interactions before they propagate.

Keentel Insight

Every one of these documented events was diagnosed, replicated, and corrected through the combination of field measurement and validated EMT modeling — the same discipline the industry spent a decade building for inverter-based generation. Developers who invest in facility-level EMT models during design, rather than after an event, protect both their energization schedule and their neighbors on the grid.

3.3 Architectures in Motion: From 415 V AC to 800 V DC

Rack power density is forcing a redesign of the entire internal distribution chain, and each architecture generation changes the facility's dynamic signature at the point of interconnection. The conventional design delivers medium voltage to a step-down transformer, through a double-conversion UPS, to in-rack power supplies converting roughly 415/480 V AC to 54 V DC. The Open Compute Project's ORV3 generation moves the battery backup function into the rack alongside the power supply. As machine-learning racks head toward 500 kW and beyond, the industry is standardizing on 400 V DC distribution: a "sidecar" power rack takes in low-voltage AC, converts it to 400 V DC, and distributes it to adjacent compute racks, where DC/DC converters step down to 54 V. Full-bridge LLC resonant conversion inside 18 kW power supply modules, combined six to a shelf, enables roughly 400 kW per rack.

The next step is building-level DC. In the "Northstar"-class architecture, a solid state transformer — or a large rectifier — converts medium voltage AC (11 kV to 35 kV) directly to 400 V DC at the facility level, eliminating conversion stages. Nvidia has proposed an 800 V DC ecosystem for AI factories supporting megawatt-class racks: 13.8 kV AC is converted to 800 V DC at the data center perimeter, transmitted via busway through the halls, and stepped down to 54 V/12 V inside the rack, carrying substantially more power through the same conductor cross-section and shedding multiple AC/DC stages along the way.

For the power system engineer, each of these designs changes what must be modeled. The sidecar converter, the solid state transformer, the facility-level rectifier, and even the final DC/DC stage each contribute to the facility's fault response and its harmonic and oscillatory potential. Solid state transformers in particular are active power-electronic devices with their own control loops — coupled with storage, they can be engineered to deliver ride-through, which makes them both a modeling obligation and a compliance opportunity. Study models built for the conventional architecture cannot simply be reused for a 400 V DC or 800 V DC facility.

3.4 Cooling: The Quiet Ride-Through Liability

Cooling can consume 30 to 40 percent of a data center's electricity at the top of the range, and nearly every fan and pump in a modern facility is driven by a variable frequency drive or electronically commutated motor. VFDs carry their own ride-through settings, programmed independently of the IT equipment — and a cooling system that trips during a fault will force the compute floor down minutes later regardless of how well the servers rode through. Facility ride-through is therefore a chain with two links, and the weaker one governs. VFD parameters at large facilities should be reviewed and set so that normally cleared faults do not remove cooling, and facility-level models must represent the cooling drives, not just the IT power chain. Cooling design also drives PUE and therefore the total energy forecast the interconnecting utility must plan around.

4. AI Workloads: The Most Dynamic Load Ever Connected

AI campuses share the physical building blocks of conventional data centers, but their workloads produce electrical behavior with no precedent at this scale. The distinction begins with the silicon: GPU clusters optimized for matrix mathematics operate in large, synchronized groups, and that synchronization stamps itself onto the facility's power draw.

Training is the extreme case. The backpropagation cycle that dominates model training alternates between phases — a forward pass that spikes compute demand, a loss calculation that relaxes it, a backward pass that spikes it again, and a weight update that relaxes it — and the facility's aggregate load follows this rhythm in a repetitive sawtooth. Training campuses idle at roughly 60 to 70 percent of peak simply to maintain readiness, then swing between partial idle and full load as jobs progress. Because a synchronized training job requires every node, a single machine failure halts the entire run: 30 to 50 percent of active facility load can vanish within microseconds, then return as the job auto-reconfigures and restarts — a cycle that can repeat tens of times per day. At the individual GPU the power step is nearly instantaneous, though the supporting power supply's capacitance and control loops stretch the facility-level ramp to roughly fifty milliseconds.

Inference is gentler but far from benign. Its baseline is low, punctuated by request bursts that swing the load between idle and near-peak; aggregate inference fleets have exhibited power reductions of 80 to 90 percent in under a second. Fine-tuning sits between the two. And the trend line is adverse: rack-level measurements comparing prior-generation and current-generation GPUs show that power-swing potential is growing with each hardware generation, driven in part by liquid-cooled GPU architectures. Aggregation across racks and modules reshapes but does not eliminate the variability seen at the point of interconnection.

Distributed deep learning adds a systemic dimension. Training jobs now span multiple campuses — sometimes in different balancing areas. When one participating site experiences a grid fault and trips, it signals its partners to pause, and multiple multi-hundred-megawatt facilities may reduce demand together. A coordinated, effectively instantaneous loss of several gigawatts of load is an over-frequency event of the first order: it can force generators to run back or trip, and in the extreme, cascade. Planning coordinators need to treat correlated multi-site load loss as a studied contingency, not a hypothetical — and the size of that contingency is a moving target as training fleets grow.

This publication is part of the Keentel Engineering Technical Insight Series, developed to help utilities, developers, independent power producers, data center operators, and industrial energy users understand the engineering realities of interconnecting very large loads to the bulk power system. It synthesizes publicly available industry research — including the July 2026 report "Large Loads: Behaviors, Capabilities, and Limitations" published by the Energy Systems Integration Group (ESIG) Large Loads Task Force — together with Keentel Engineering's own interconnection, power system study, and compliance engineering experience.

All analysis, commentary, and recommendations in this document are Keentel Engineering's own. Where industry findings are discussed, they are summarized and interpreted in our own words; readers should consult the original ESIG publications at esig.energy for the source material.

Non-Affiliation Disclaimer

Keentel Engineering is an independent consulting engineering firm. Keentel Engineering is not affiliated with, endorsed by, or sponsored by the Energy Systems Integration Group (ESIG), NERC, ERCOT, PJM, CAISO, the Open Compute Project, ITIC, IEEE, or any equipment vendor, developer, or organization referenced in this document. All trademarks and report titles are the property of their respective owners and are referenced solely for identification and educational purposes.

Part 1 Technical Blog: When the Load Becomes the Largest Machine on the Grid

Part 2 — Frequently Asked Questions: Large Load Interconnection

The variability question ultimately becomes an oscillation question. Repetitive training cycles at the wrong frequency, at gigawatt scale, constitute a forced oscillation source that can excite inter-area modes or torsional modes at nearby synchronous machines. Frequency-domain screening of the facility's expected load spectrum against known system modes belongs in the interconnection study scope for any large AI campus.

4.1 The Flexibility Toolbox

The same characteristics that make AI training disruptive also make it the most promising flexibility resource among large loads. Training is an offline process without enterprise latency commitments: it can pause, throttle, and resume. A 2025 field demonstration in Phoenix by Emerald AI, Nvidia, the Salt River Project, and EPRI showed software-based load shaping using job pausing and dynamic frequency scaling of GPU clocks and voltage. Google has published compiler-level techniques that schedule and balance training jobs across time to smooth power flow, and non-critical jobs can be shifted to data centers on entirely different power systems as a form of demand response.

Hardware complements software. Supercapacitor systems — rack-mounted or facility-level — absorb and inject power over millions of cycles with response speeds that bridge the gap between server power supplies and the UPS, smoothing transients before they reach the grid; commercial offerings now span rack-mounted supercapacitor shelves to transmission-class E-STATCOM solutions. Battery systems, including containerized megapack-class units, add energy depth for longer smoothing horizons and steeper swings. In our assessment, code-level and scheduler-level mitigation is the least capital-intensive first line, with capacitive and battery hardware sized against whatever residual variability the software cannot remove — and the interconnection agreement should specify the performance the combination must achieve at the POI.

Flexibility has limits, and contracts define them. Colocation facilities carry customer uptime obligations, and in multi-tenant buildings the tenant, not the operator, controls the compute — and therefore the megawatts. Grid programs that assume the facility operator can curtail on request will fail at exactly the facilities where contracts say otherwise. Flexibility must be engineered and contracted at interconnection, not assumed afterward.

5. Cryptocurrency Mining: The Price-Following Gigawatt

Cryptocurrency mining is the purest economic load on the grid: an arbitrage that converts energy into currency. That single fact explains nearly all of its electrical behavior. Facilities are built lean — thousands of single-phase ASIC rigs fed through step-down transformers from a 34.5 kV-class distribution intertie, with no UPS and no backup generation, because every dollar of resilience equipment erodes a business that typically returns 20 to 40 percent on investment. Auxiliary load for safety, security, and environmental control runs only 1 to 5 percent of the facility.

Under normal conditions miners run flat out, targeting fleet PUEs as low as roughly 1.04, with load shaped mainly by ambient temperature's effect on cooling. But when energy prices rise above the value of the coin being mined, the load simply leaves. Most large miners in ERCOT participate in Load Acting as a Resource, others clear demand response day-ahead, and essentially all self-curtail on price. Facilities can typically shed to 20 percent of load on request — 5 percent if pressed — and can traverse from near-zero to full load in anywhere from 20 seconds to 20 minutes. Restoration after a trip or curtailment is similarly fast, usually within about 20 minutes as cooling systems re-establish. For the balancing authority, this is a load whose short-term forecast error is driven by commodity prices and whose ramps rival utility-scale batteries — which, not coincidentally, are the resource best suited to buffering them.

Ride-through behavior tracks the ITIC curve, with real hardware measured close to — and in places short of — its boundaries. Experimental testing of a widely deployed ASIC model demonstrated continuous operation from 66 to 130 percent of rated voltage, but a zero-voltage withstand of only 9.5 milliseconds against the 20 milliseconds contemplated by ITIC. Ride-through capability is set by on-board capacitor sizing, and retrofitting deeper capability is economically unattractive; ERCOT has already logged multiple events in which fleets of miners failed to ride through disturbances and dropped in unison, perturbing system frequency.

Simulation work has also shown that harmonic interaction with nearby inverter-based resources — a solar plant, in the published case — can trip mining power supplies in low short-circuit-ratio areas. VFD-driven cooling adds a second trip path: a documented facility tripped three times on single-phase voltage sags before its VFD ride-through parameters were corrected.

Power quality closes the picture. ASIC fleets present a leading power factor near 0.99 and emit characteristic triplen harmonics. With a delta-grounded-wye interconnection transformer, the delta winding traps triplen currents and the transmission system sees compliant distortion — measured facilities have recorded voltage THD near 4.3 percent and current THD near 9.2 percent, inside IEEE 519 limits. Without the delta winding, third and fifth harmonic injection can exceed IEEE 519 and demand filtering. Transformer vector group selection, harmonic screening against IEEE 519, and permanent power quality monitoring at the POI are therefore baseline interconnection requirements for mining facilities, not optional extras.

Keentel Insight

Mining interconnections succeed or fail on three studies: a harmonic assessment that reflects the actual substation transformer vector group, a ride-through review that treats the ITIC curve as a proxy pending vendor data, and an operational characterization that gives the balancing authority honest ramp and restoration parameters. All three are inexpensive relative to the cost of a post-energization mitigation retrofit.

6. Hydrogen Electrolysis: The Electrochemical Large Load

Electrolytic hydrogen remains a small fraction of global production today, but electrolyzer capacity is compounding — global installed capacity reached 1.4 GW in 2023, doubling in a single year, with several times that reaching final investment decision since. Individual projects now enter interconnection queues at 100 MW-plus, built up from modular stacks, and their electrical character differs fundamentally from the compute loads above: the constraint is electrochemistry, not silicon.

The facility's load anatomy is dominated by the stack itself, with the balance spread across conversion, compression, water treatment, and cooling. The rectifier plant is the grid interface and the design decision that matters most for power quality: large facilities have historically used thyristor-based silicon-controlled rectifiers in 6-, 12-, or 24-pulse configurations — robust and cost-effective at high power, but heavy consumers of reactive power, exposed to commutation failure, and reliant on pulse multiplication or DSTATCOM-class mitigation to control characteristic fifth and seventh harmonics. IGBT-based rectifiers deliver better harmonic performance, power factor, and weak-grid behavior, and are gaining share, though with current limits that constrain them at the largest stack ratings, higher switching losses, and higher cost. Paired rectifier transformers with staggered vector groups are the standard tool for canceling low-order harmonics and smoothing DC ripple to the stacks.

The questions below reflect what developers, utilities, and facility operators most often ask Keentel Engineering about interconnecting data centers AI campuses, cryptocurrency mining facilities, and hydrogen electrolysis plants to the bulk power system.

About Keentel Engineering

Keentel Engineering is a power systems and grid interconnection consulting firm headquartered in Tampa, Florida, with offices in Austin, Sacramento, and Baltimore. Our service lines span grid interconnection engineering, substation and transmission design, power system studies, NERC compliance, renewables and battery energy storage engineering, EMT modeling, and owner's engineer services. We help developers, utilities, and large energy users treat interconnection as a first-order engineering input — because on today's grid, that is exactly what it is.

Keentel Engineering | 400 N Ashley Dr STE 2600, Tampa, FL | (813) 389-7871 | contact@keentelengineering.com | FL Firm Registration No. 36853

Disclaimer

This document is provided for general informational and educational purposes only and does not constitute engineering advice for any specific project. Interconnection requirements, study scopes, and performance standards vary by jurisdiction and change over time; consult a licensed professional engineer regarding your specific facility. Keentel Engineering is not affiliated with ESIG, NERC, ERCOT, or any organization, standard body, or vendor referenced herein.

ERCOT BESS Interconnection Guide | TNMP & AEP Texas

Mon, 13 Jul 2026 14:45:41 GMT

Jul 13, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Interconnecting a Utility-Scale BESS in ERCOT: The Complete Guide to the TNMP and AEP Texas Process

Why the 10 MW Threshold Changes Everything

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

Battery Energy Storage Systems (BESS) are being deployed across Texas at a record pace, and for good reason. ERCOT's volatile energy prices, growing ancillary service markets, and rapid load growth make the grid one of the most attractive storage markets in the world. But between a signed land lease and a battery earning revenue sits one of the most consequential — and most frequently underestimated — phases of any project: interconnection.

This guide walks through the full interconnection process for a utility-scale BESS in the ERCOT region, with a specific focus on projects in Texas-New Mexico Power (TNMP) and AEP Texas service territories. It covers the classification thresholds that determine your pathway, the ERCOT study and registration process, the utility-side coordination, fees, timelines, and the commissioning gauntlet at the end. We close with answers to the questions developers ask us most, and four anonymized case studies from real projects in TNMP and AEP territory.

ERCOT is unusual among North American grid operators. There is no traditional cluster queue; each project moves on its own schedule through a defined sequence of studies, though projects can still be affected by others studying the same part of the grid. The process is administered through ERCOT's online Resource Integration and Ongoing Operations – Interconnection Services (RIOO-IS) portal, and the governing requirements live in Planning Guide Section 5.

Phase 0: Preparation and Application

The entity submitting the request — the Interconnecting Entity (IE) — begins by creating a RIOO-IS user account, which requires registration and multi-factor authentication setup. The GINR application itself demands a well-developed project definition: the point of interconnection (POI), the proposed in-service date, equipment selections, and preliminary electrical design. Application quality matters. Deficient or incomplete submissions get kicked back, and in a process where a missed window can cost a full quarter, rework is expensive.

Applicable fees are set by the ERCOT Fee Schedule. For a project at or above 10 MW, the administrative fees paid directly to ERCOT stack up as follows:

For a project in TNMP territory, the TDSP coordination track runs alongside the ERCOT process. TNMP administers interconnection applications through its PowerClerk-based portal, where the applicant registers an account and works through a structured application covering customer and installer information, the service point and its ESI ID (TNMP ESI IDs are 17-digit identifiers), and the generation configuration — with Energy Storage as its own generator type, including configurations for integrated storage systems and separate inverter-plus-battery arrangements, and DC-coupled versus AC-coupled designs for hybrid projects.

The application requires the technical package that defines the project electrically: a one-line diagram, a layout sketch showing a lockable, visible disconnect device, equipment certifications (IEEE 1547 / UL 1741), voltage and phase configuration, kVA rating and power factor, export intentions, and confirmation that the inverter manufacturer has supplied dynamic modeling values to the utility. The application concludes with selection of the interconnection agreement party structure — TNMP's agreement contemplates four options depending on whether the end-use customer, a separate generator owner, the premises owner, or a contracted energy-rights holder acts as the counterparty.

From there, TNMP's workflow proceeds through defined stages: application review (TNMP targets returning an executable interconnection agreement within about ten business days for standard applications), electronic signatures, construction, field inspection, service order generation, meter installation, and finally Permission to Operate. For utility-scale projects, expect substantially deeper engagement than the standard portal flow implies — direct coordination with TNMP engineering on substation or feeder connection design, easement and right-of-way review where TNMP facilities cross the project property or third-party land, and encroachment review for any construction near existing TNMP facilities. And as noted above, for a ≥10 MW project TNMP is also the entity conducting the ERCOT Full Interconnection Study, which makes a constructive working relationship with TNMP's planning engineers doubly valuable.

Interconnection is necessary but not sufficient. Before energization, a BESS project must also clear local and state permitting, and battery storage presents a wrinkle: most local zoning ordinances were written before BESS existed, so the use is rarely listed. Projects frequently need a Conditional Use Permit (CUP) or Special Use Permit (SUP) from the local authority having jurisdiction, which introduces a public-facing, discretionary approval into the schedule.

On the code side, the installation must comply with the National Electrical Code and, centrally for BESS, NFPA 855, the Standard for the Installation of Stationary Energy Storage Systems. NFPA 855 drives siting and separation distances, fire suppression and detection design, deflagration protection, emergency response planning, and the hazard mitigation analysis that fire marshals increasingly demand. Fire protection engineering is not a checkbox at the end of design — separation requirements and suppression infrastructure affect site layout, and site layout affects the one-line, the collector system design, and even the interconnection application. It belongs in the design process from day one.

Case Studies

The following case studies are drawn from utility-scale BESS interconnection projects in ERCOT. Client names, locations, and identifying details have been anonymized.

Case Study 1 — TNMP Territory: The Threshold Decision That Saved a Year

A developer approached us with a planned 10 MW / 20 MWh standalone BESS on a distribution-adjacent site in west Texas TNMP territory, assuming the project would follow the utility's standard DG process. Our first-week screening flagged the problem: at exactly 10 MW, the project sat on the wrong side of ERCOT's DG threshold and was headed into the full GINR pathway — roughly $17,000 in ERCOT fees, a half-million-dollar security posting, and an 18-to-30-month clock, against a revenue model built on a 12-month energization target.

We ran a two-track analysis: the market revenue impact of derating to a 9.9 MW self-limiting configuration versus the carrying cost and schedule risk of the full GINR route. For this project's merchant strategy, the lost margin on 0.1 MW was immaterial next to a year of earlier revenue. We redesigned the power conversion system controls and protection settings to enforce the self-limiting configuration at the POI, documented it to ERCOT's self-limiting facility requirements, and filed through TNMP's process with a complete one-line, layout with lockable visible disconnect, and manufacturer dynamic modeling data on day one. The project cleared studies without a restudy cycle and energized inside its original financing window — approximately 14 months ahead of the GINR counterfactual.

Case Study 2 — TNMP Territory: Recovering a Stalled Application

A storage owner-operator engaged us to rescue a BESS interconnection in TNMP territory that had been in process for months without reaching an executable interconnection agreement. Our review found the application had been submitted with an incomplete technical package: the one-line diagram did not match the as-designed collector system, the layout sketch omitted the required lockable visible disconnect detail, and the inverter manufacturer's dynamic modeling values had never been transmitted to the utility — each deficiency individually minor, collectively fatal to review progress.

We rebuilt the submission as a coherent package: corrected one-line and site layout, IEEE 1547/UL 1741 certification documentation, complete generation detail with verified kVA and power factor data, and direct coordination with the inverter OEM to deliver validated dynamic models. We also restructured the interconnection agreement party designation, which had been filed under the wrong option for the project's ownership structure. The resubmitted application moved through TNMP review to an executable agreement in under three weeks, and the project proceeded through inspection and meter installation to Permission to Operate without further deficiency notices. The lesson we emphasize from this engagement: in interconnection, application quality is schedule.

Case Study 3 — AEP Texas Territory: Parallel-Tracking Utility and ERCOT Workstreams

A developer of a 9.95 MW / 20 MWh BESS in AEP Texas territory engaged us as owner's engineer with a hard commercial operation deadline driven by an offtake commitment. The project qualified for AEP's streamlined DER pathway, but the developer's baseline schedule treated the process as sequential: AEP application, then studies, then agreement, then ERCOT registration, then metering, then testing. Run that way, the timeline missed the deadline by roughly a quarter.

We restructured the program to run the tracks in parallel. While AEP's engineers conducted their impact and safety studies, we simultaneously prepared the ERCOT registration package — RARF data, collector system information, and dynamic models — and pre-coordinated the EPS settlement metering design so meter procurement started before the Interconnection Service Agreement was signed. Ride-through capability documentation was assembled from manufacturer test data during construction rather than after it. When AEP issued Permission to Operate, the ERCOT side was already staged: commissioning tests were scheduled, telemetry was validated within days, and the project reached commercial operation about ten weeks sooner than the sequential baseline — inside the offtake deadline.

Case Study 4 — AEP Texas Territory: Fire Protection as a Design Driver, Not an Afterthought

A 10 MW-class BESS project in a growing AEP Texas municipality came to us after receiving pushback from the local fire marshal on its preliminary site plan. The original layout, produced before any fire-protection engineering, placed battery enclosures at spacings that did not satisfy NFPA 855 separation requirements once the enclosure listing and fire-testing documentation were examined — and the municipality, with limited prior BESS experience, was requiring a full hazard mitigation analysis and emergency response plan before it would advance the project's Special Use Permit.

We took over the fire protection engineering scope and integrated it with the electrical design rather than bolting it on. The site was re-planned around compliant separation distances and fire-apparatus access; enclosure-level detection, suppression, and deflagration protection were specified against the manufacturer's large-scale fire test data; and we produced the hazard mitigation analysis and emergency response plan, then supported the developer in working sessions with the fire marshal and at the public SUP hearing. The revised layout required rerouting the DC collection and shifting the interconnection switchgear — changes that would have been catastrophic if discovered after the utility design package was submitted, but manageable because the site plan, one-line, and AEP application were finalized together. The project secured its Special Use Permit, cleared AEP's review on the corrected package, and avoided what we conservatively estimate would have been a six-month redesign loop had the original plan reached the utility first.

These figures reflect the fee schedule at the time of writing and should always be verified against the current ERCOT Protocols before budgeting, as they are periodically revised. They also exclude TDSP study costs, third-party engineering, and — critically — financial security.

Phase 1: Screening Study (approximately 90 days)

Once the application is validated, ERCOT performs a screening evaluation of the proposed POI, looking for immediate fatal flaws and system constraints. Results are typically returned within 90 calendar days of a validated application.

After screening, the project must post interconnection financial security of $50,000 per MW to hold its position — $500,000 for a 10 MW project. A portion (roughly 20%) is refundable if the project withdraws early in the process, but the security converts to fully non-refundable once construction-stage agreements are finalized. This deposit, far more than the administrative fees, is the number that disciplines speculative projects out of the ERCOT development pipeline.

Phase 2: Full Interconnection Study (roughly 120 to 180 days)

The FIS is the technical heart of the process, and here is a detail many developers miss: the FIS is conducted by the TDSP — TNMP or AEP Texas for the projects this guide covers — under ERCOT's framework. The study package includes steady-state power flow analysis, dynamic stability simulation, short-circuit analysis, and facility design evaluation, with individual studies typically running 45 to 60 days each in sequence across a four-to-six-month block.

The FIS phase is also where the project's modeling obligations come due. Within 90 days of application, the IE must deliver comprehensive technical models: steady-state, dynamic, and transient inverter models suitable for ERCOT's simulation platforms. ERCOT's Model Quality Guide — which now includes formal Model Quality Testing requirements along with UDM and PSCAD model guidelines — governs what is acceptable. Inverter-based resources like BESS receive particular scrutiny here, because their behavior during grid disturbances is defined by control software rather than physics, and poorly validated models are one of the most common causes of study delays and restudy cycles.

Depending on the location and technology, ERCOT may also require a subsynchronous resonance (SSR) vulnerability study and will require a Reactive Power study demonstrating that the resource meets the reactive capability requirements of Nodal Protocol Section 3.15.

Phase 3: Agreements, Registration, and the QSA Cycle

With studies complete and any required network upgrades defined, the project executes a Standard Generation Interconnection Agreement (SGIA) with the TDSP. Note for current projects: all SGIAs executed on or after January 1, 2026 must use the updated SGIA form dated November 6, 2025, which incorporates changes approved by the Public Utility Commission of Texas in Project 58211. The older 2019 form survives only for amendments to pre-2026 agreements.

After the interconnection agreement is signed, BESS projects submit the Battery Request for Information Template through RIOO-IS, and the project enters ERCOT's Quarterly Stability Assessment (QSA) cycle, with submission requirements defined in Planning Guide 5.3.5. The QSA runs on fixed quarterly windows, and this is one of the least forgiving schedule mechanics in the entire process: miss a data-submission deadline and the project does not slip by a week — it slips into the next 90-day quarterly window. Disciplined data management during this phase is worth real money.

In parallel, once the planned resource has met the requirements of Planning Guide Section 6.9 for addition to the planning models, it can be registered with ERCOT. The entity that registers and takes responsibility for the resource — the Resource Entity (RE) — may be the same as the IE or a different entity. Registration involves the Resource Asset Registration Forms (RARF), which feed ERCOT's network operations model, along with the Collector System Template covering the site's internal cable segments and the Dynamic Model Templates with the Model Quality Test report. Projects planning to co-locate load under 75 MW with the storage resource also complete a Load Information Form to coordinate that arrangement during interconnection.

Phase 4: Commissioning and Energization (roughly 90 to 120 days)

The final phase transforms a construction project into a market-registered, dispatchable ERCOT resource. Key elements include the Commissioning Plan, submitted for ERCOT approval to energize, synchronize, and complete required testing; AVR (automatic voltage regulator) testing tailored to BESS resources, with data submitted on ERCOT's standard templates; PMU (phasor measurement unit) sample data submission for transmission-connected resources; frequency ride-through and voltage ride-through capability reporting; the New Generator Commissioning Checklist, executed with the project's Qualified Scheduling Entity (QSE) across ERCOT's market and control systems; and the attestation required under the Lone Star Infrastructure Protection Act.

Only after this sequence — plus telemetry validation and settlement metering verification — does ERCOT grant approval to connect and, ultimately, commercial operation status that allows the battery to transact in the wholesale energy market and provide ancillary services.

AEP Texas administers its interconnection intake through the AEP Texas interconnection portal, with an application fee and a documentation package that will look familiar: one-line wiring diagram, inverter specification sheets, site plans, and a visible AC disconnect diagram. AEP engineering performs impact and safety studies evaluating whether the project will overload local facilities or cause voltage disruptions, and approved projects execute an Interconnection Service Agreement (ISA) with AEP, including payment for any physical grid upgrades the studies identify.

For projects below the DG threshold, AEP's process is the streamlined path — the utility-level review substitutes for the full ERCOT study sequence, and the ERCOT side reduces to registration, EPS (ERCOT-Polled Settlement) metering, and model verification with ride-through testing. For projects at 10 MW and above, the AEP track functions like the TNMP track described above: the utility coordination and ISA run alongside the full GINR process, with AEP as the studying TDSP.

In both territories, the closing sequence is the same shape: construction, utility field verification, bidirectional/settlement meter installation, the utility's Permission to Operate letter, and then ERCOT commercial operation status.

Beyond the Utilities: Permitting, Codes, and Fire Protection

The Bottom Line

The AEP Texas Track: What to Expect in AEP Territory

Before anything else, a developer needs to know which of two very different pathways their project falls into, because the answer drives the timeline, the fees, the study burden, and the engineering scope.

In ERCOT, generation and storage resources below 10 MW are generally treated as Distributed Generation (DG). These projects avoid the full transmission-level review and are handled primarily through the local utility's (the Transmission/Distribution Service Provider, or TDSP's) streamlined interconnection process. The evaluation focuses on the local distribution system — line loading, voltage impacts, protection coordination — rather than the statewide transmission grid. DG projects commonly clear the process in 8 to 12 months.

Projects at 10 MW and above are a different animal. They enter ERCOT's Generation Interconnection or Change Request (GINR) process — the full large-resource pathway — with a rigorous sequence of interconnection studies, comprehensive modeling requirements, a standardized interconnection agreement, quarterly stability assessments, and a formal commissioning and registration program. These projects typically take 18 to 30 months from application to commercial operation.

An important nuance that trips up many developers: the threshold is measured in megawatts of power, not megawatt-hours of energy. A 10 MW / 20 MWh system is a 10 MW resource for classification purposes — the 20 MWh simply means it can sustain full output for two hours. And because the DG pathway applies to resources under 10 MW, a project at exactly 10 MW does not qualify for the streamlined treatment. It goes through the full GINR process.

This is also why nameplate strategy is a genuine engineering and commercial decision. ERCOT's framework recognizes Self-Limiting Facilities — projects that are physically and contractually limited to inject below a threshold at the point of interconnection. For some developers, configuring a project at 9.9 MW instead of 10 MW is the difference between a one-year process and a three-year process. For others, offtake commitments, market participation plans, or project economics make the full 10 MW (or more) worth the longer road. An experienced owner's engineer should pressure-test this decision before the first application is filed, because changing course mid-queue is expensive.

The remainder of this guide assumes a project of 10 MW or greater following the full GINR pathway, with notes on where the TNMP and AEP tracks differ.

The ERCOT GINR Process, Phase by Phase

Frequently Asked Questions

Interconnecting a utility-scale BESS in ERCOT is a multi-front campaign: ERCOT's study, modeling, registration, and commissioning machinery on one front; the TDSP's application, design review, agreement, and inspection process on another; and local permitting, codes, and fire protection on a third. The projects that reach commercial operation on schedule are the ones that treat these fronts as one integrated program — where the nameplate strategy is settled before the application is filed, the models are validated before ERCOT asks, the QSA deadlines are managed like the schedule cliffs they are, and NFPA 855 shapes the site plan instead of tearing it up.

That integration is the core of what an owner's engineer does. If you are developing battery storage in TNMP, AEP Texas, or anywhere in the ERCOT region — whether you are sizing your first project against the 10 MW threshold or untangling a stalled application — our team has walked this road and can help you plan it, de-risk it, and deliver it.

The TNMP Track: What to Expect in Texas-New Mexico Power Territory

SCADA in Field Services & Digital Substation Architecture

Mon, 13 Jul 2026 13:17:37 GMT

Jul 13, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

SCADA in Field Services and Substation SCADA Architecture

1. Why SCADA Is Now a First-Order Design Input

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

From RTUs and Data Concentrators to IEC 61850 Digital Substations — Designing, Commissioning, and Securing the Systems That Run the Modern Grid

At its core, every SCADA system performs the same loop: sense, digitize, transmit, interpret, decide, and act. Field sensors and instrument transformers measure analog quantities — voltages, currents, temperatures, pressures, tap positions. Signal conditioning and analog-to-digital conversion turn those measurements into digital values. Communication links move the data to a master station, where operators and applications interpret it, and control commands travel the same path in reverse to breakers, switches, tap changers, and inverter controllers.

SCADA has evolved through four recognizable generations. First-generation monolithic systems ran on standalone minicomputers with proprietary protocols and no external connectivity.

Second-generation distributed systems spread processing across networked stations on a LAN. Third-generation networked SCADA embraced open standards — DNP3, IEC 60870-5, and eventually IEC 61850 — allowing multi-vendor interoperability across wide-area networks. The fourth generation folds in IP-native architectures, cloud-adjacent analytics, and IIoT-style edge devices. Most utilities today operate somewhere between the third and fourth generation, with legacy islands from the first two still in service — which is exactly why field-services engineering and protocol integration remain so consequential.

Keentel Perspective

Generational mixing is the norm, not the exception. A typical brownfield substation we encounter has a 20-year-old RTU polling serial IEDs, a partially deployed station bus, and a corporate mandate to reach IEC 61850. The engineering challenge is not choosing the destination — it is sequencing the migration so protection, control, and NERC compliance never lapse along the way.

3.1 Remote Terminal Units and Gateways

The RTU remains the workhorse of substation data acquisition. A modern RTU is a modular platform — CPU, analog input/output cards, digital status inputs, control outputs, redundant power supplies, and multiple communication interfaces — hardened for the substation environment and increasingly expected to carry NERC CIP -relevant security functions. Critically, the modern RTU is no longer just an I/O box: it acts as the substation data concentrator and secure gateway, polling dozens of IEDs over the station LAN, performing protocol conversion for legacy devices, hosting a web-based local HMI, and presenting a single hardened connection point to the wide-area network.

3.2 Intelligent Electronic Devices

Microprocessor-based protective relays, meters, bay controllers, and power-quality monitors — collectively IEDs — have absorbed functions that once required racks of discrete hardware. A single modern relay provides protection elements, metering, event and disturbance recording, breaker monitoring, and native Ethernet communications. From a SCADA architecture standpoint, IEDs are both the richest data sources in the substation and the devices whose configuration discipline most directly determines data quality at the master station.

3.3 Merging Units and the Process Bus

Merging units extend the digital architecture from the control house into the switchyard. By digitizing current and voltage signals at the primary equipment and publishing time-synchronized sampled values over fiber, the merging unit replaces bundles of copper CT and VT wiring with a process bus. The payoff is significant: reduced copper, reduced CT saturation and open-circuit hazards, simplified panel wiring, and a foundation for fully digital substations

The cost is a new engineering discipline — process bus network design, time synchronization (PTP/IEEE 1588), and sampled-value performance validation all become field-services responsibilities that did not exist in the copper world.

3.4 Master Stations and the HMI

At the top of the hierarchy sits the master control center: SCADA servers, application servers, historians, communication front ends, ICCP servers for inter-utility data exchange, network management, and dispatcher training simulators. The human-machine interface is where architecture meets operations — mimic diagrams, alarm handling, and situational-awareness tools determine whether an operator can distinguish a genuine cascading event from an alarm flood. Intelligent alarm filtering and rationalized alarm philosophies are design deliverables, not afterthoughts.

8. Securing the Substation SCADA System

Because SCADA directly supervises energized equipment, its reliability philosophy must be explicit. Three related but distinct concepts govern the design:

Fail-safe — A design that, upon failure, defaults to a state that cannot cause harm — normally-closed alarm circuits that annunciate on wire break, complementary signal pairs that expose sensor faults, and control outputs that revert to safe states on loss of communication.
Fault-tolerant — A design that continues full operation through component failure — duplexed CPUs, mirrored disks and historians, redundant power supplies, and hot-standby servers with automatic changeover.
High availability — A design philosophy for the overall system — dual independent LANs, redundant communication channels and routes to the control center, primary/secondary master configurations, and geographically separated backup control centers.

In power system SCADA, redundancy must be traced end to end: a redundant server farm buys nothing if a single serial link, single antenna, or single station gateway sits in the path of critical data. Channel redundancy from the RTU outward, CPU and port redundancy in the RTU itself, dual station LANs, and diverse WAN routing are the recurring requirements for mission-critical points — and the items most frequently value-engineered out under budget pressure. Our position is straightforward: define the critical-function list first, apply full redundancy to that list without compromise, and economize elsewhere.

SCADA security differs fundamentally from enterprise IT security: availability and integrity outrank confidentiality, devices live for decades, and patching windows are constrained by outage schedules. The threat catalog is well established — false data injection, replay and covert attacks against control loops, protocol exploits against unauthenticated DNP3/Modbus, malicious firmware, and removable-media vectors into engineering workstations. History's marquee ICS malware families all exploited the same underlying gaps: flat networks, unauthenticated protocols, and unmanaged remote access.

The countermeasure framework is defense-in-depth, applied specifically to the substation context:

Zone-based architecture — Segment the architecture into security zones with defined conduits: process bus, station bus, station-level DMZ, WAN, and control center — each boundary enforced by firewalls of differing makes with deny-by-default rulesets, and no direct sessions permitted from less-trusted zones into the SCADA core.
ICS-aware monitoring — Deploy SCADA-aware intrusion detection that understands DNP3, 61850, and Modbus semantics — flagging not just malformed packets but valid-but-anomalous control sequences.
Endpoint hardening — Harden the endpoints: RTU and IED account management, disabled unused ports and services, signed firmware, controlled removable media, and secured local HMI access.
Protocol-layer security — Wrap the protocols: DNP3 Secure Authentication and IEC 62351 profiles where the installed base supports them; encrypted VPN transport for WAN links regardless.
Compliance as architecture — Anchor the program in NERC CIP: asset categorization, electronic security perimeters, interactive remote access management, configuration change monitoring, and incid ent response — engineered into the architecture rather than bolted on for audit season.

6. Protocols: Choosing and Integrating the Right Ones

Protocol selection shapes everything downstream — point mapping effort, event time-tagging fidelity, bandwidth requirements, and the cybersecurity surface. The table below summarizes the protocols a substation SCADA engineer in North America actually encounters:

3. The Building Blocks: RTUs, IEDs, Data Concentrators, and Merging Units

Above the station level, the utility control center connects via DNP3 (dominant in North America) or IEC 60870-5-104 (dominant internationally), while control centers exchange data with neighbors and reliability coordinators over ICCP (IEC 60870-6 / TASE.2). The architecture within the station can be arranged in several topologies — cascaded (daisy-chained) switches for small stations, ring architectures with rapid spanning tree for medium stations, and fully redundant designs using PRP (Parallel Redundancy Protocol) or HSR (High-availability Seamless Redundancy) where zero-recovery-time failover is required for protection-class traffic.

Design Rule of Thumb

Match the network redundancy class to the traffic class. Supervisory polling can tolerate seconds of network reconvergence; GOOSE-based breaker-failure tripping and sampled values cannot. If protection functions ride the network, the network is protection equipment — design, test, and maintain it to that standard.

5. Reliability Engineering: Fail-Safe, Fault-Tolerant, and High-Availability Design

7. SCADA in Field Services: Where Architecture Meets Reality

9. The Migration Question: Legacy, Hybrid, or Fully Digital?

legacy Modbus and serial DNP3 devices concentrated behind a gateway that normalizes them into the station data model; IEC 61850 deployed for the station bus where it earns its complexity; and a clean, well-documented DNP3 or 104 point map presented to the control center. Complementary standards round out the picture — IEEE C37.118 for synchrophasor streams, IEC 62351 for protocol-layer security, and IEC 61968/61970 (CIM) for enterprise data integration.

4. Substation SCADA Architecture: Levels, Buses, and Topologies

Keentel Engineering delivers substation SCADA and automation as an integrated engineering scope — not a bolt-on to someone else's design. Our services span:

Architecture and design — SCADA architecture development and 30/60/90/IFC design integration for transmission and distribution substations, renewable plants, BESS facilities, and large-load interconnections.
Integration engineering — RTAC/gateway configuration, IEC 61850 system engineering (SCL, GOOSE, sampled values), DNP3 and Modbus point mapping, and legacy protocol integration.
Field services and commissioning — FAT/SAT planning and execution, point-to-point testing, redundancy and failover validation, and commissioning support through energization.
Compliance and cybersecurity — NERC CIP-aligned network zoning, secure remote access design, and disturbance-monitoring compliance engineering for conventional and inverter-based resources.
Owner's engineer services — Owner's engineer oversight of vendor SCADA scopes — reviewing architectures, point databases, and test records so the owner's interests are protected from specification through final acceptance.

Whether you are planning a greenfield digital substation, migrating a legacy RTU fleet, or untangling an integration problem discovered at commissioning, our team brings protection, automation, and interconnection engineering under one roof.

Let's Talk

Contact Keentel Engineering at 813-389-7871 or contact@keentelengineering.com to discuss your substation SCADA architecture, integration, or field-services scope. Offices in Tampa, Austin, Sacramento, and Baltimore. www.keentelengineering.com

Disclaimer

This publication is provided by Keentel Engineering for general informational purposes only and does not constitute engineering advice for any specific project. Keentel Engineering is not affiliated with, endorsed by, or sponsored by any standards body, equipment manufacturer, publisher, or author referenced herein. All trademarks, standards designations, and product names are the property of their respective owners. Readers should consult a licensed professional engineer regarding their specific applications.

10. How Keentel Engineering Supports SCADA and Substation Automation

Very few utilities can justify rip-and-replace. The economically rational path is almost always hybrid: new IEDs, data concentrators, and fiber deployed on an expansion or protection-refresh project, coexisting with the legacy RTU-based system while protocol conversion bridges the two. The engineering discipline lies in defining the target architecture first — station bus topology, data model, security zones, naming conventions — so each incremental project is a step toward the target rather than another one-off island. For greenfield stations, the calculus increasingly favors going digital from the start: IEC 61850 station and process bus, merging units, PRP/HSR networking, and an HMI with modern situational awareness, specified as an integrated system rather than assembled by change order.

Supervisory Control and Data Acquisition has quietly become the nervous system of the North American grid. Every unmanned substation, every utility-scale solar plant, every battery energy storage facility, and every large-load interconnection depends on a SCADA architecture that acquires field data reliably, transports it securely, and presents it to operators in a form that supports fast, correct decisions. Yet on far too many projects, SCADA is still treated the way interconnection studies once were: as a late-stage administrative task, scoped after the one-line is frozen and the protection panels are ordered.

At Keentel Engineering, we take the opposite view. SCADA architecture — point counts, protocol selection, network topology, redundancy philosophy, and cybersecurity zoning — belongs at the 30% design milestone, alongside the protection and control philosophy. Retrofitting a communications architecture into a substation that was not designed for it is expensive; retrofitting cybersecurity into a SCADA system that was not zoned for it is worse. This blog walks through what SCADA actually does in the field, how modern substation SCADA architectures are structured, and where field-services engineering makes or breaks the outcome.

2. SCADA Fundamentals: Data Acquisition to Supervisory Control

Frequently Asked Questions

A SCADA architecture is only as good as its commissioning. Field services is where the design assumptions — point lists, protocol mappings, network settings, failover logic — are proven against energized reality. In our experience, the field scope breaks into six recurring disciplines:

Point-to-point and end-to-end testing — Verifying every analog scale factor, every status point, and every control output from the field device through the RTU/gateway to the control center display. Point-to-point testing is tedious and non-negotiable; a mis-scaled MW point or reversed breaker status discovered during a system event is an operational hazard.
Integration and protocol conversion — Configuring and validating protocol conversions at the gateway — Modbus registers to DNP3 points, serial IED data into the 61850 station model — including exception behaviors, quality flags, and time-tag propagation that only surface under fault and restart conditions.
FAT and SAT — Structured factory acceptance testing against the approved point database and HMI screens, followed by site acceptance testing with the real field wiring, real communication channels, and real master station — with documented punch lists closing the loop between the two.
Redundancy validation — Failover and redundancy proving: pulling primary power supplies, failing network links, forcing server changeovers, and confirming the system degrades and recovers exactly as the reliability philosophy promises — with no loss of event data and no spurious controls.
Timing and SOE integrity — Time synchronization verification across GPS clocks, IRIG-B distribution, PTP grandmasters, and NTP fallbacks, so sequence-of-events records and synchrophasor data are trustworthy for post-event analysis and NERC disturbance-monitoring obligations.
Migration and cutover — Cutover planning for brownfield migrations — sequencing legacy RTU retirement, parallel operation windows, rollback criteria, and outage coordination so the control c enter never loses visibility of an energized station.

Field Reality Check

The most common field finding is not a broken device — it is an undocumented assumption. A point list that drifted from the relay settings, a gateway map edited on-site without a database update, a spare input landed on the wrong terminal. Rigorous configuration management, disciplined as-built documentation, and independent point-to-point verification are the cheapest insurance a utility can buy.

IEEE Standards in Power Engineering | Keentel Engineering

Mon, 13 Jul 2026 11:11:06 GMT

Jul 13, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

IEEE Standards and How Keentel Engineering Puts Them to Work

Introduction: Standards Are Design Inputs, Not Checkboxes

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

Disclaimer

Keentel Perspective

A standards-aligned design is cheaper than a standards-remediated one. Nearly every costly field change order we have reviewed as owner's engineer traces back to a standard that was applied late — a grounding grid sized before soil data existed, a battery sized without IEEE 485 duty cycles, or an arc-flash label printed from an outdated IEEE 1584 model.

Strip away the vendor language and every data center resolves into the same functional blocks. A secure envelope encloses one or more data halls of IT cabinets. Utility power enters through an intake substation, flows through medium- and low-voltage switchgear, is conditioned and bridged by uninterruptible power supply (UPS) systems, backed by standby generation, and delivered to racks through power distribution units (PDUs), busway, or rack power panels. Heat leaves through a cooling chain — room or row-level air handlers, chilled water or refrigerant loops, and external heat rejection — sized to remove essentially every watt the electrical system delivers. Around this core sit the support spaces: plant rooms, battery rooms, network intake rooms at diverse building entries, loading and build/test areas, a network operations center, and the security layers that control movement among them all.

Rack power density is the design variable that drives everything else. Legacy enterprise cabinets at 2–5 kW, virtualized and blade environments at 8–15 kW, dense compute at 20–40 kW, and current-generation AI training racks at 80–150 kW and climbing represent fundamentally different electrical distribution, cooling, and structural problems — not scaled versions of one problem. A hall designed for uniform 8 kW air-cooled racks cannot absorb a 120 kW liquid-cooled AI pod without touching the power chain, the cooling topology, the floor loading, and often the utility supply itself. Because IT refreshes every two to five years while the building and plant last decades, the highest-value design decision is the one that keeps density growth from becoming a rebuild: oversized risers and containment routes, plant rooms with expansion positions, electrical architectures that scale in modules, and a site with secured headroom at the point of interconnection.

What the standard covers

IEEE 80 is the governing guide for the design of substation grounding systems. It defines the safety criteria — tolerable touch voltage, step voltage, and ground potential rise (GPR) — based on human body current limits, fault duration, and surface material, and it provides the methodology for sizing grounding conductors, arranging grid meshes and ground rods, and verifying that a fault on the system does not create lethal potentials for personnel or the public.

How Keentel Engineering aligns

Every Keentel substation and collector-station grounding design begins with an IEEE 80 safety analysis performed in dedicated grounding software (WinIGS/CDEGS-class tools), modeling the actual grid geometry, soil model, and split-factor fault current rather than hand approximations.
We size grid conductors and connections for the X/R-adjusted fault current and clearing time from the protection study — coordinating IEEE 80 with the C37-based relay settings so the assumed fault duration is the real one.
Touch and step voltages are checked at fences, gates, equipment operating handles, and adjacent metallic structures, with crushed-rock surfacing derating applied per IEEE 80.
Deliverables include a stamped grounding study report with GPR, tolerable-limit calculations, mesh/touch/step contour plots, and a construction-ready grounding plan.

What the standard covers

IEEE 81 defines how to measure the things IEEE 80 needs as inputs: soil resistivity (Wenner four-point and Schlumberger arrays), grounding grid impedance (fall-of-potential method), and continuity of grounding connections. A grounding design is only as good as the soil model beneath it, and IEEE 81 is the standard that keeps that model honest.

How Keentel Engineering aligns

Keentel specifies IEEE 81 Wenner-array soil resistivity testing at multiple traverses and probe spacings during the site-investigation phase — before 30% design — so the two-layer (or multilayer) soil model is a measured quantity, not an assumption.
We develop test plans and review contractor field data for validity: probe spacing versus depth of interest, seasonal moisture effects, and interference from buried metallic objects.
For energized or existing stations, we specify fall-of-potential grid impedance testing per IEEE 81 to validate as-built performance against the IEEE 80 design model.
Measured data is reduced into layered soil models directly inside our grounding software, closing the loop between field measurement and design.

IEEE 1547 Interconnection of Distributed Energy Resources

What the standard covers

IEEE 1584 provides the empirically derived model for calculating arc-flash incident energy and arc-flash boundary distances in low- and medium-voltage equipment. The 2018 edition substantially revised the model — introducing electrode configuration factors and enclosure-size corrections — and remains the basis for the labels, PPE selection, and safe work boundaries required by workplace electrical safety programs.

How Keentel Engineering aligns

Keentel performs arc-flash studies using the current IEEE 1584 model in ETAP/SKM-class software, built on a verified short-circuit model and the actual protective device settings from our coordination study — never on assumed clearing times.
We evaluate incident-energy reduction options as part of the study: maintenance-mode switches, zone-selective interlocking, differential zones, and faster settings validated against selectivity requirements.
Deliverables include equipment labels, an incident-energy table, and an engineering report identifying every bus above target energy thresholds with concrete mitigation recommendations.
Legacy studies performed under the 2002 model are flagged for re-study — the 2018 equations can move results materially in either direction.

What the standard covers

IEEE 1547 (with its 2018 revision and the 1547.1 test standard) is the national framework for interconnecting distributed energy resources — solar PV, BESS, and other inverter-based resources — with distribution and area electric power systems. It defines voltage and frequency ride-through, voltage regulation functions (volt-var, volt-watt), power quality requirements, and interoperability.

How Keentel Engineering aligns

Keentel prepares IEEE 1547-conformant interconnection packages for distribution-connected solar and storage, including our community-solar and sub-5 MW project work, mapping each utility's technical requirements to the standard's performance categories (Category A/B and I/II/III).
We specify and verify smart-inverter function settings — ride-through curves, volt-var and frequency-watt parameters — against both IEEE 1547-2018 and the interconnecting utility's adopted profile.
Our power system studies (load flow, short circuit, protection impact) quantify the effects the standard is designed to manage: voltage regulation interaction, unintentional islanding risk, and protection desensitization.
For transmission-connected IBRs we carry the same discipline into the NERC world — ride-through and disturbance-monitoring requirements such as PRC-029 and PRC-028 — so distribution and bulk-system compliance are handled as one continuum.

IEEE 485 Battery Sizing for Stationary Applications

What the standard covers

IEEE 485 defines the method for sizing vented and valve-regulated lead-acid batteries for stationary duty — the DC systems that trip breakers, run protection relays, and keep SCADA alive when the station itself is dead. It formalizes the duty cycle, the section-by-section sizing calculation, and the aging, temperature, and design margins that must be applied.

How Keentel Engineering aligns

Keentel builds a complete IEEE 485 duty cycle for every substation DC system we design: continuous relay and RTU load, momentary breaker-trip inrush, and the emergency-lighting or communications tail load over the specified backup period.
We apply the standard's aging (typically 125%), temperature, and design-margin factors and document them transparently so the owner can see exactly why the battery is the size it is.
Battery sizing is coordinated with charger sizing, DC voltage-window checks at the relay terminals (including end-of-discharge voltage drop), and the C37-driven trip duty of the actual breaker population.
For non-lead-acid chemistries we apply the companion IEEE practices and manufacturer data while preserving the IEEE 485 duty-cycle discipline.

IEEE 81 Earth Resistivity and Grounding Impedance Measurement

What the standard covers

The IEEE Green Book addresses grounding philosophy inside the fence and inside the plant: whether a system should be solidly grounded, low- or high-resistance grounded, or ungrounded; how equipment grounding and bonding should be executed; and how static, lightning, and electronic-equipment grounding interact with the power system.

How Keentel Engineering aligns

Keentel applies IEEE 142 when selecting the grounding method for collector systems, auxiliary power systems, and industrial facilities — for example, low-resistance grounding of a 34.5 kV collector neutral to limit ground-fault damage while keeping faults detectable.
Our medium-voltage design packages document the grounding decision explicitly, with fault-current, protection-sensitivity, and transient-overvoltage justifications drawn from the Green Book.
We coordinate IEEE 142 equipment grounding and bonding details with the IEEE 80 grid design so the below-grade and above-grade systems behave as one network.

IEEE 1584 Arc-Flash Hazard Calculations

What the standard covers

IEEE 519 establishes the recommended limits for harmonic voltage distortion supplied by the utility and harmonic current distortion injected by users, measured at the point of common coupling (PCC). With inverter-based resources, VFD-heavy industrial loads, and data centers proliferating, IEEE 519 compliance has become a routine interconnection requirement.

How Keentel Engineering aligns

Keentel performs harmonic studies for solar, BESS, and large-load interconnections, modeling inverter harmonic spectra against the frequency-dependent network impedance — including resonance introduced by collector cable capacitance and harmonic filters.
We evaluate compliance at the correct PCC with the correct short-circuit-ratio-based current limits from the standard's tables, not generic thresholds.
Where limits are exceeded, we design mitigation — tuned filters, C-type filters, or control-side solutions — and verify performance across contingency network conditions.
Harmonic study results feed directly into our power quality monitoring recommendations under IEEE 1159, so compliance can be demonstrated after energization, not just predicted before it.

IEEE 1159 Monitoring Electric Power Quality

IEEE 519 Harmonic Control in Electric Power Systems

This document is published by Keentel Engineering for educational and informational purposes. Keentel Engineering is an independent consulting engineering firm and is not affiliated with, endorsed by, or sponsored by IEEE, the Bureau of Indian Standards (BIS), the International Electrotechnical Commission (IEC), NERC, FERC, or any equipment manufacturer or standards body referenced herein. All trademarks and standard designations are the property of their respective owners. Standards are revised periodically; readers should always verify the current edition and any amendments before applying requirements to a specific project. The case studies in this document are fully anonymized composites drawn from representative project experience; names, locations, ratings, and identifying details have been altered.

IEEE 142 (Green Book) Grounding of Industrial and Commercial Power Systems

What the standard covers

IEEE 693 provides the recommended practice for the seismic qualification of substation equipment — bushings, transformers, disconnect switches, instrument transformers, and their support structures — defining qualification levels (low, moderate, high) and the analysis or shake-table testing required to demonstrate performance.

How Keentel Engineering aligns

For projects in seismically active regions — including our West Coast work — Keentel specifies the appropriate IEEE 693 qualification level in equipment procurement specifications and reviews vendor qualification reports for conformance.
We coordinate equipment anchorage, foundation design inputs, and flexible-bus/conductor slack requirements with the structural engineer so seismic qualification is preserved by the installation, not defeated by it.
As owner's engineer we verify that seismic documentation is complete at factory-acceptance and delivery — a detail that is inexpensive to check early and extremely expensive to discover missing later.

IEEE 693 Seismic Design of Substations

What the standard covers

IEEE 1159 is the recommended practice for monitoring and classifying power quality phenomena — sags, swells, interruptions, transients, harmonics, flicker, and unbalance — with consistent definitions, magnitude/duration categories, and monitoring methodology.

How Keentel Engineering aligns

Keentel uses IEEE 1159 categories as the common language in every power quality investigation, so an event described in our reports means the same thing to the utility, the owner, and the equipment vendor.
We develop monitoring plans — instrument placement, trigger thresholds, and recording durations — for commissioning of IBR plants and for diagnosing malfunction complaints at industrial facilities and data centers.
Disturbance records from relays and dedicated PQ meters are analyzed against IEEE 1159 classifications and cross-referenced with our EMT and dynamic models to identify root cause rather than just symptoms.

IEEE 1815 DNP3 Communication Protocol

Every substation, solar plant, battery energy storage system (BESS), and industrial facility that Keentel Engineering touches is shaped by a common body of engineering law: the IEEE standards. These documents are not paperwork to be cited at the end of a project — they are first-order design inputs that determine conductor sizes, grounding grid geometry, relay settings, battery capacity, harmonic filters, and even the anchor bolts under a transformer. At Keentel Engineering, we treat grid interconnection and power system studies the same way: as engineering that must be done early, correctly, and to the letter of the governing standard.

This blog walks through eleven of the most consequential IEEE standards in power engineering — the same standards that appear on virtually every utility technical-requirements document and interconnection agreement — and explains, standard by standard, how Keentel Engineering's workflows, software toolchain, and deliverables align with each one.

IEEE C37 Series Switchgear, Circuit Breakers, and Protection

What the standard covers

IEEE 1815 standardizes DNP3, the dominant SCADA communications protocol in North American utility systems. It defines the data models, event reporting, time synchronization, and — critically — Secure Authentication for communications between master stations, RTUs, and IEDs.

How Keentel Engineering aligns

Keentel designs substation automation architectures — including SEL RTAC- and Axion-based systems — with IEEE 1815/DNP3 point maps engineered deliberately: class assignments, deadbands, event buffering, and time-sync strategy documented for the operating utility.
We build integration test plans that exercise the DNP3 profile point-by-point during factory and site acceptance testing, so SCADA cutover is a verification exercise rather than a discovery exercise.
Where projects mix IEC 61850 GOOSE/MMS inside the fence with DNP3 to the control center, we engineer the gateway mapping so no protection-critical or compliance-critical point is lost in translation.
Secure Authentication and network segmentation recommendations are aligned with the owner's NERC CIP posture.

What the standard covers

The C37 family is the backbone of protection and switchgear engineering: C37.04/.06/.09 for circuit breaker ratings and testing, C37.010 for application, C37.2 for device function numbers, C37.90 series for relay standards, C37.91–.119 protection guides for transformers, lines, buses, and generators, C37.20 series for switchgear assemblies, and C37.111 (COMTRADE) for disturbance records, among many others.

How Keentel Engineering aligns

Keentel's short-circuit studies apply C37.010/C37.5 methodology so breaker interrupting duties are compared against nameplate ratings on the correct symmetrical/asymmetrical basis, including X/R correction.
Our protection and coordination designs follow the applicable C37 protection guides — transformer differential and overcurrent per C37.91, line protection per C37.113, bus protection per C37.234 — and every one-line and schematic uses C37.2 device numbering.
Relay settings files, coordination curves, and setting-basis documents are delivered as an auditable package that supports NERC PRC compliance obligations.
Disturbance monitoring deliverables (DFR/relay event records) are specified in COMTRADE format per C37.111, which also underpins PRC-028 disturbance-monitoring compliance for inverter-based resources.

Pulling It Together: One Model, Many Standards

The most important thing about these eleven standards is that they are coupled . The IEEE 80 grounding design depends on the C37-based clearing time. The IEEE 1584 arc-flash result depends on the same relay settings. The IEEE 485 battery must trip the C37-rated breakers. The IEEE 519 harmonic study and the IEEE 1547 interconnection functions interact through the same inverter controls. Keentel Engineering's practice is built around this coupling: one verified system model, one protection philosophy, and one set of studies that feed each other — delivered by licensed engineers who work with these standards every day across grid interconnection substation design, power system studies NERC compliance, and owner's engineer engagements.

About Keentel Engineering

Key takeaway

Most 'full' legacy facilities are full on paper and stranded in practice. Measured capacity accounting, disciplined air management, and a block-architected power and liquid design turned a decade-old 6 kW hall into a home for 90 kW AI racks — live, without a new building, and with the whole facility running more efficiently than before the project started.

Whether you need an IEEE 80/81 grounding study, an IEEE 1584 arc-flash refresh, IEEE 519/1547 interconnection studies for a solar or BESS project, or a full standards-aligned substation design Keentel Engineering delivers stamped, utility-ready engineering. Call 813-389-7871 or email contact@keentelengineering.com.

Anonymized Case Studies

IEEE 80 Safety in AC Substation Grounding

The following case studies are fully anonymized composites representative of Keentel Engineering's project experience. Client names, locations, voltages, and ratings have been altered to protect confidentiality; the engineering issues and outcomes are authentic.

Case Study 1: IEEE 80/81 Grounding Redesign for a Utility-Scale Solar Collector Substation

The Challenge

The EPC's preliminary grounding design had been sized using an assumed uniform 100 Ω·m soil model carried over from a different site. Late-stage IEEE 81 Wenner testing revealed a high-resistivity upper layer over a conductive lower layer — a profile that dramatically changes current distribution in the grid and pushes touch voltages up at the grid perimeter. The utility's interconnection reviewer rejected the submitted study, putting the energization date at risk.

The Approach

Keentel re-reduced the raw IEEE 81 field data into a validated two-layer soil model, rebuilt the grid in full-geometry grounding software, and imported the actual single-line-to-ground fault current and split factor from the short-circuit study — with clearing time taken from the real C37-coordinated relay settings rather than a conservative placeholder. The analysis showed perimeter mesh voltages exceeding IEEE 80 tolerable touch limits at the fence and at two equipment operating positions.

The Solution

Rather than a costly wholesale re-trench, Keentel designed targeted reinforcement: perimeter conductor densification, deep-driven rods reaching the conductive lower layer, gradient-control loops at the fence and gates, and a verified crushed-rock surfacing specification with documented derating per IEEE 80. The revised design was validated in the model and packaged with GPR, tolerable-limit, and contour-plot documentation.

Results

Touch and step voltages brought within IEEE 80 tolerable limits at all locations, with documented margin.
Utility review passed on first resubmittal; energization date preserved.
Reinforcement cost was a small fraction of the EPC's initially feared full-grid redesign.
Post-construction IEEE 81 fall-of-potential testing confirmed grid impedance within 10% of the model prediction.

Lesson Learned

Soil resistivity data is the foundation of every grounding design. Measured IEEE 81 data must precede — not follow — IEEE 80 design. Copying a soil model between sites is one of the most expensive shortcuts in substation engineering.

Case Study 2: IEEE 1584 Arc-Flash and C37 Protection Modernization for an Industrial Facility

The Challenge

The plant's arc-flash labels dated to a 2012 study performed under the IEEE 1584-2002 model, and the utility had since increased available fault current at the service entrance. Several main-tie-main lineups carried dangerously optimistic labels, and maintenance work practices were built around them. The owner also suspected miscoordination after a feeder fault had tripped a main breaker and dropped half the plant.

The Approach

Keentel rebuilt the system model from verified field data — nameplates, cable schedules, current relay settings files — and re-ran short-circuit duties per C37.010 methodology against breaker ratings. The coordination study was redone across all voltage levels, and arc flash was recalculated under IEEE 1584-2018 with correct electrode configurations and enclosure dimensions. The station battery supporting the 69 kV breakers was checked against an IEEE 485 duty cycle as part of the same engagement.

The Solution

The study identified two breakers with interrupting duties above nameplate at the new utility fault level, the miscoordinated feeder/main pair responsible for the plant-wide trip, and eleven buses where 2018-model incident energies exceeded the old labels — three of them severely. Keentel delivered revised C37-coordinated settings, maintenance-mode (arc-energy-reduction) switching on the worst lineups, replacement recommendations for the over-dutied breakers, updated labels for every bus, and a resized station battery calculation.

Results

Full selectivity restored — subsequent feeder faults cleared at the feeder level with no plant-wide interruptions reported.
Worst-case incident energy on the main lineups reduced by more than 60% in maintenance mode.
Over-dutied breakers identified and replaced before failure rather than after.
A current, defensible IEEE 1584-2018 label set supporting the owner's electrical safety program.

Lesson Learned

Arc-flash, short-circuit, coordination, and DC-system studies are one coupled problem. Updating labels without re-verifying breaker duties and coordination — or vice versa — leaves the most dangerous gaps untouched.

Case Study 3: IEEE 1547/519 Interconnection and Power Quality Engineering for a Distribution-Connected Solar-Plus-Storage Portfolio

The Challenge

A distributed-generation developer faced three different utilities with three different IEEE 1547-2018 adoption profiles — different ride-through categories, different volt-var default curves, and different study requirements. One flagship site also sat at the end of a long rural feeder with a weak short-circuit ratio, raising utility concerns about voltage regulation interaction and harmonic distortion at the PCC.

The Approach

Keentel built a standard interconnection engineering package that mapped each utility's technical requirements to the IEEE 1547-2018 performance categories, then tailored inverter certification documentation, settings tables, and study scopes per jurisdiction. For the weak-feeder site, a harmonic scan of the feeder impedance versus the certified inverter spectra was run across grid contingency conditions, and IEEE 519 current-distortion limits were evaluated at the utility-confirmed PCC using the actual short-circuit ratio.

The Solution

The studies demonstrated IEEE 519 compliance without external filters once inverter quantity-versus-loading behavior at low output was accounted for, and identified a volt-var curve adjustment that resolved the utility's regulation-interaction concern in simulation. Keentel documented IEEE 1547 Category settings site-by-site, specified DNP3 (IEEE 1815) point maps for the utilities requiring direct transfer-trip and curtailment interfaces, and defined an IEEE 1159-based post-energization monitoring plan to verify predicted performance.

Results

All portfolio sites received interconnection approval without redesign; the weak-feeder flagship site avoided a six-figure harmonic filter through study-demonstrated compliance.
A reusable, jurisdiction-mapped IEEE 1547 settings and documentation framework cut engineering cycle time on subsequent sites materially.
Post-energization IEEE 1159 monitoring confirmed voltage regulation and distortion performance within study predictions, closing out utility conditions.

Lesson Learned

IEEE 1547-2018 is a framework, not a single rulebook — every utility adopts it differently. Portfolio developers save the most money when interconnection engineering is standardized around the framework and localized per utility, instead of reinvented site by site.

Frequently Asked Questions

Work With Keentel Engineering

Keentel Engineering is a power systems and grid interconnection consulting firm headquartered in Tampa, Florida, with offices in Austin, Sacramento, and Baltimore (FL Engineering Firm Registry No. 36853). Our services span grid interconnection and POI engineering, power system studies, substation and transmission design, EMT modeling, renewables and BESS engineering, NERC compliance, and owner's engineer services — all delivered by licensed engineers who treat interconnection and standards compliance as first-order design inputs, not late-stage administrative steps.

PRC-028 Compliance Guide | Disturbance Monitoring for IBRs

Sun, 12 Jul 2026 14:43:04 GMT

Jul 12, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

PRC-028 Substation Compliance

Chapter 2 — Anatomy of PRC-028-1: Applicability and the Eight Requirements

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

2.2 The Requirement Structure

PRC-028-1 applies to Generator Owners and, for certain interconnection-point obligations, Transmission Owners of inverter-based resources. The facility threshold captures both Bulk Electric System IBRs and a second category of non-BES IBRs: facilities with an aggregate nameplate rating of 20 MVA or greater interconnected at 60 kV or above. In practice this sweeps in essentially every utility-scale solar PV plant, Type 3 and Type 4 wind facility, battery energy storage system, HVDC-connected resource, and hybrid facility of commercial significance. Facilities near the BES boundary should confirm applicability with their Regional Entity, and owners should note that resources below the bright-line thresholds can still be designated for compliance based on reliability need.

The word "aggregate" matters. A 12 MVA solar block co-located with a 10 MVA battery behind one interconnection is a 22 MVA facility for applicability purposes. Hybrid facilities should evaluate the combined nameplate at the point of interconnection, not the individual technology blocks.

The standard is organized as eight requirements that follow a deliberate logic: three data types, each with a "where and what" requirement and a "how well" requirement, followed by data management and equipment maintenance obligations.

Each requirement carries a corresponding Measure defining acceptable evidence — actual data recordings or derivations, or documents describing device specifications, configurations, and settings. That evidentiary dual-path is significant for program design: a well-documented design standard, applied consistently and kept current, is itself compliance evidence. Chapter 11 returns to what an audit-ready evidence package looks like.

PRC-028-1 is built on three distinct recording disciplines. They answer different questions, run at different speeds, and impose different equipment demands. Conflating them is the most common early design error.

3.1 Sequence of Events Recording (SER)

7.1 Performance Envelope Versus the Standard

PRC-028-1 became effective April 1, 2025, with implementation phased by facility category. The structure rewards early movers and punishes owners who treat the far dates as the start date.

7.2 Platform Selection

DFR functionality requires the modern RTAC compute tier running R153-V0 or later firmware: the SEL-3555, SEL-3560, SEL-3350, or the SEL-2241-2 Axion CPU module. Legacy controllers — the SEL-3505 class and the original SEL-3530 generation — do not support the fault-recording and continuous-recording engines and must be replaced (with project conversion) as part of a compliance retrofit. On the SEL-3555 and SEL-3560E, an additional PCIe card — the SEL-3390E4 network adapter or the SEL-3390T time-and-Ethernet adapter — supplies the dedicated EtherCAT interface to the Axion I/O; the 3390T additionally brings a high-accuracy PTP time input suitable for synchrophasor-class synchronization. Firmware licensing must include the FileIO, Dynamic Disturbance Recording, and Continuous Recorder options — a procurement line item that is easy to miss and painful to discover at commissioning.

Three planning realities deserve emphasis. First, the new-resource schedule is effectively "operational at COD": a facility reaching commercial operation after mid-2026 must energize with its disturbance monitoring working, which puts the DFR design on the critical path of the substation package rather than in a post-COD punch list. Second, the 50% milestone for existing fleets is a portfolio decision — owners with dozens of sites need a prioritization logic (interconnection voltage, plant size, regional scrutiny, outage windows) and a repeatable site design, because bespoke engineering at every plant will not scale to the 2030 date. Third, R8 is already live: any facility with recording capability today carries the 90-day restore-or-CAP obligation now, which means recorder health monitoring and failure-discovery documentation cannot wait for the R1–R7 dates.

Owners should also account for the market realities around these dates: qualified engineering resources, recorder hardware, and instrument-transformer lead times are all being consumed by an entire industry converging on the same deadlines. Extension mechanisms exist for circumstances genuinely beyond an owner's control, but an extension request supported by a documented, in-flight program is a very different conversation with the Regional Entity than one that begins after the milestone has passed.

Chapter 6 — PRC-028 in Context: Neighboring Standards and Regional Overlays

Chapter 3 — The Data Triad: SER, FR, and DDR

4.1 The DDR Performance Floor

Requirement R5 sets the two numbers that dominate recorder selection: input sampling of at least 960 samples per second and an output recording rate for electrical quantities of at least 60 times per second. The 960 sps floor — 16 samples per cycle at 60 Hz — exists because IBR control dynamics live at speeds conventional disturbance recorders were never specified to resolve. Any equipment evaluation should treat these as minimums to be exceeded with margin, not targets to be met exactly: recording headroom is cheap at design time and unobtainable during an event investigation.

4.2 Fault-Recording Triggers

Requirement R3 enumerates the trigger classes fault recording must implement — neutral (residual) overcurrent, AC phase overvoltage and undervoltage, and over/underfrequency. A compliant design implements these as configured trigger elements with documented thresholds and pickup times, and the configuration record itself becomes Measure evidence. Good practice layers additional triggers the standard does not require — sequence-component and rate-of-change-of-frequency elements — because the marginal cost is a settings entry and the marginal value during an oscillation or ride-through investigation is substantial.

4.3 Time Synchronization

Requirement R6 exists because untimed data is nearly worthless for event reconstruction: records from the plant, the transmission owner's recorders, and neighboring facilities must align on a common time base traceable to UTC. The engineering implication is a facility time architecture — a GNSS-disciplined clock, distribution via IRIG-B and/or Precision Time Protocol, and verification that every recording element is actually locked. Millisecond-class alignment satisfies event sequencing; the reference architecture in this book delivers microsecond-class synchronization, which additionally enables synchrophasor-grade measurement from the same hardware.

4.4 Data Formats and Retention

Requirement R7 standardizes the deliverable: records in COMTRADE per IEEE C37.111 (1999 revision or later) with file names conforming to IEEE C37.232 COMNAME (2011 revision or later), retained so they remain retrievable for the standard's retention window and produced upon request from the Reliability Coordinator, Regional Entity, or NERC within the required response period. Two operational consequences follow. First, retention is a rolling engineering property of the recorder — the continuous archive must be sized so the retention window survives worst-case recording load. Second, retrieval must be a documented, rehearsed procedure, not an improvisation: an owner who cannot produce a requested record inside the response window has a violation regardless of whether the data technically exists.

Verify the Enforceable Text

Retention windows, response periods, and phased dates in this book reflect the adopted PRC-028-1 and its Implementation Plan as published. NERC standards are living documents subject to revision and regional variance; every compliance program should tie its internal procedures to the currently enforceable version retrieved from the NERC Reliability Standards library, and re-verify at each program milestone.

Chapter 5 — The Compliance Clock: Phased Implementation

Chapter 7 — Reference Architecture: SEL RTAC + Axion Digital Fault Recorder

6.1 PRC-002-5: The Synchronous Sibling

PRC-002 remains the disturbance-monitoring standard for the conventional system, and its current revision, PRC-002-5, clarifies the boundary by excluding IBRs — which now live under PRC-028-1. The two standards are parallel in architecture (SER, FR, DDR; formats; retention) but differ where IBR physics demands it, most notably in unit-level SER reach into inverter controls and in the recording performance floor. Transmission Owners with recorders at IBR interconnection points, and owners of mixed portfolios, must track obligations under both standards; the reference architecture in this book is documented by its manufacturer as designed to exceed both PRC-002 and PRC-028 recording requirements, which is precisely why a single platform standard across a mixed fleet is attractive.

6.2 PRC-029-1 and PRC-030-1: The Consumers of the Data

PRC-029-1 establishes voltage and frequency ride-through performance requirements for IBRs — remaining connected and continuing to deliver current through defined disturbance zones, with momentary cessation prohibited in the no-trip region — with design obligations effective October 1, 2026, and performance demonstration obligations that depend on disturbance monitoring data. PRC-030-1 obligates analysis and correction of unexpected IBR performance following disturbances. Both standards presume the existence of exactly the records PRC-028 mandates: an owner cannot prove ride-through, and cannot perform credible post-event analysis, without SER, FR, and DDR. Compliance programs should therefore treat the three standards as one data-driven lifecycle — record (PRC-028), perform and demonstrate (PRC-029), analyze and correct (PRC-030) — and design the monitoring architecture for all three consumers at once.

6.3 Regional Overlays

Regional frameworks can layer additional or accelerated obligations on the same equipment. In ERCOT, for example, dynamic disturbance and event-recording obligations under the region's own rule changes apply alongside PRC-028-1, with their own retention and provision duties — some already in force. A single, well-designed disturbance monitoring installation can satisfy multiple rulebooks simultaneously, but the compliance evidence must trace to each framework's citations separately. Owners operating across regions should maintain a requirements matrix per site, mapping every recording capability to every obligation it discharges.

Chapter 4 — Technical Specifications That Drive Equipment Selection

8.1 Channel Math Before Hardware

The design sequence that controls cost runs: single-line diagram → monitored areas → measurement points → module count → chassis and platform. Working from the facility SLD, the engineer marks every location where R1, R2, and R4 quantities live — the point of interconnection, main power transformers (both sides where required), collector feeders, reactive devices, and the breakers and disconnects whose status feeds SER. Each three-phase V+I measurement point consumes an AC protection module input group; each block of status points consumes digital-input channels. Only after that map exists should anyone pick a platform, because the channel count — not habit — determines whether the site is a single 10-slot node or a multi-node network, and whether a 3350 suffices or a 3555's capacity is warranted.

8.2 Storage Engineering for Continuous Recording

Continuous recording is where storage design becomes compliance design. Each voltage or current channel configured for continuous recording at 3 kHz consumes roughly 600–700 MB per day. A modest plant recording, say, 24 analog channels continuously is producing on the order of 15–17 GB per day — about half a terabyte per month — before fault records, PMU streams, and SOE archives. Drives up to 8 TB are available on the SEL-3555, SEL-3560E, and SEL-3350, and the design deliverable is a simple table: channels × rate × retention target, with margin for event bursts and with the R7 retrievability window as the floor, not the goal. On the 3555, RAID-1 mirroring of the data drives adds loss protection for the archive that is the entire point of the installation.

8.3 Instrument Transformers: The Long-Lead Reality

The recorder is rarely the schedule risk; the copper is. Brownfield sites frequently lack spare CT cores or accessible PT secondaries at the required locations, and adding them means outages, and sometimes procurement with the longest lead time in the project. Two design rules follow. On existing facilities, exhaust the reuse options first — existing protection-class CTs and PTs and the event-capable relays already on collector feeders can discharge a substantial share of the FR obligation with reconfiguration rather than construction. On greenfield facilities, specify dedicated measurement cores before steel is ordered: a spare CT core costs almost nothing at procurement and a fortune after energization. Ratios are entered in the DFR configuration as X:1 (a 1200:5 CT is entered as 240), and the CT/PT data sheet for every source belongs in the compliance evidence file.

Chapter 8 — Sizing the System: Channels, Storage, and Instrument Transformers

7.3 The Measurement Front End

The Axion side of the architecture is a modular chassis system — SEL-2242 backplanes in 10-slot, 4-slot, and dual 4-slot formats — populated per the site's channel map: SEL-2243 power couplers (which also carry the node-to-node EtherCAT links), SEL-2245-42 AC protection modules terminating CT and PT secondaries for the FR/DDR quantities, SEL-2245-43 analog input modules for auxiliary quantities, and SEL-2244 digital input/output modules for breaker and disconnect status, alarms, and DFR status outputs. A digital input variant with battery-monitoring capability (0.5% DC accuracy) lets the same installation watch the station battery — one of the auxiliary quantities every disturbance investigation eventually asks about. The platform scales to 60 modules and beyond a thousand digital inputs, so channel-count growth is a configuration exercise rather than a replacement project.

7.4 EtherCAT Network Rules

The RTAC-to-Axion link is EtherCAT, and its rules are strict enough to deserve their own paragraph in every design package: exactly one Ethernet interface on the controller is dedicated to EtherCAT; the connection from that port to EtherCAT Port 1 on the first node's power coupler (Slot A) is direct — the protocol is non-routable and no switches are permitted; and subsequent Axion nodes daisy-chain coupler to coupler. Modern platform documentation additionally describes star, sequential, and combination topologies for multi-node layouts; the governing document for any given firmware is the SEL-5033 manual, and the topology should be fixed during design, not discovered during commissioning.

Chapter 9 — Configuration: From Single-Line Diagram to Running Recorder

The RTAC DFR extension reduces what was historically days of recorder engineering to a menu-driven configuration measured in hours, with no code written. The workflow below mirrors the sequence Keentel executes on a typical IBR collector substation; every artifact it produces is simultaneously compliance evidence.

9.1 The Asset Model

The extension organizes the plant as assets of three types. Bus assets carry voltage-only measurement. Generic assets carry voltage and current — the right model for transformers, inverters, reactors, and capacitor banks. Transmission line assets carry voltage and current plus impedance-based fault location, with line impedance parameters and restraint factors. Feeders are modeled as line assets; where a layout permits current measurement on a bus, modeling it as a generic asset captures both quantities. Line assets accept current summation from two CT sources — the mechanism that handles ring-bus and breaker-and-a-half positions where a line's current is the sum of two breaker CTs. The asset names chosen here propagate into channel names, PMU names, and event records, so the naming convention is an audit artifact: name assets after the plant's real equipment designations.

9.2 The Configuration Sequence

The end-to-end sequence: insert the Digital Fault Recorder extension into an R153+ project; set the global parameters (station and company identity — which flow into COMTRADE headers — nominal frequency, phase rotation, PMU data rate, continuous-recording retention days, fault-recording rate and record lengths with pre-trigger time); define the Axion nodes with chassis size and module layout, entering module variants and CT/PT ratios; define the substation assets and bind each to its measurement modules; configure trigger conditions per asset; map digital inputs (breakers on falling-edge triggers so a 52A dropout produces a fault record, disconnects monitored without triggering, unused points disabled); assign local monitoring outputs; and finally toggle Build DFR to true. The build step auto-generates every device object, all required settings, the IEC 61131-3 logic, a continuous recording group referencing the EtherCAT modules, and a PMU per asset — the objects that would once have been weeks of manual engineering, created consistently in minutes.

9.3 Triggers Mapped to R3 — and Beyond

9.4 Verification and In-Service

Commissioning follows the same three-layer philosophy Keentel applies to all substation communications: transport (the EtherCAT network and every module nominal), data (live-data inspection of every channel against primary injection or known system values, forced values where appropriate), and application (staged trigger tests producing real COMTRADE records that are pulled through the actual retrieval workflow and opened in the event-analysis tool). The project compiles to zero errors and zero warnings before download; after Go Online and download, the configuration runs in service independent of the engineering connection. The commissioning records — injection sheets, trigger test records, retrieved sample COMTRADE files, time-sync verification — are retained as the foundation of the R-series evidence package.

Appendix A — Requirement-to-Capability Mapping

Chapter 10 — Time Architecture: Satisfying R6 With Margin

Chapter 11 — Operating the Program: Retention, Retrieval, R8, and the Audit File

11.1 Retrieval as a Rehearsed Procedure

On the reference platform, every retrieval path is browser-based. Triggered fault records accumulate as COMTRADE files under the controller's event collection; continuous-recording extracts are generated on demand — operator selects start time, duration, and channels (3 kHz oscillography, PMU, digital) and the system produces a COMTRADE record for download; SOE logs and impedance fault-location results export as CSV from the file manager. The compliance procedure wraps these mechanics: who retrieves, within what internal deadline (set comfortably inside the standard's response window), in what naming discipline (COMNAME conformance is a requirement, not a style choice), delivered to whom, with what transmittal record. Run the procedure end-to-end at commissioning and annually thereafter; the rehearsal records are themselves evidence.

11.2 R8: Failure Management Is a Live Obligation

R8 starts a 90-day clock at the discovery of a recording-capability failure: restore, or initiate a Corrective Action Plan. Two program elements make R8 survivable. First, automated failure discovery: the DFR's alarm logic asserts on EtherCAT network abnormality, controller CPU burden above 75% for a sustained minute, storage below thresholds (10% or 4 GB), or memory exhaustion — and its Enabled/Ready/Recording/Synchronized/Alarm indications should be mapped to SCADA so a recording failure is a same-day alarm, not an audit-day surprise. Second, a documented discovery-to-disposition workflow: date-stamped discovery record, restoration work order or CAP initiation, and closure evidence. The watchdog behavior of the platform deserves a line in the operating procedure as well: a runtime-exception watchdog that reverts the controller to a default project is exactly the class of event that constitutes a recording-capability failure, and operations staff should know the recovery-and-report path in advance.

11.3 Security and Change Control Around the Recorder

The DFR is a cyber asset in a compliance-critical role, and the platform's native controls should be engaged rather than admired: allowlist-based execution control so only authorized applications run, role-based accounts with the factory-default credentials retired before energization, X.509 certificate management for encrypted access, syslog and network-audit reporting into the owner's monitoring, and the controller's ability to serve as an encrypted engineering-access gateway to downstream devices. Configuration changes to the recorder follow the same change-control rigor as protection settings — versioned projects, comparison between revisions, and an approval record — because an undocumented settings change that silently disabled a trigger is both a reliability failure and an evidentiary one.

11.4 The Audit-Ready Evidence Package

Assembled once and maintained, the package that makes a PRC-028 audit uneventful contains: the applicability determination; the site single-line with monitored areas and measurement points; the CT/PT schedule with ratios and data sheets; the design standard mapping each requirement to the implementing equipment and settings; the recorder configuration exports (asset definitions, trigger settings, recording rates, retention settings); time-architecture documentation and sync-verification records; commissioning and trigger-test records with sample COMTRADE/COMNAME-conformant files; the retention and retrieval procedure with rehearsal records; the R8 discovery-and-disposition log; and dated transmittals for any data actually provided on request. Each item traces to a specific Measure. The discipline is not glamorous; it is the difference between an audit finding and an audit compliment.

Chapter 12 — Program Roadmap: Brownfield, Greenfield, and How Keentel Helps

12.1 Brownfield Strategy

For an existing fleet, the economical program runs: portfolio applicability screen → per-site gap assessment against R1–R8 (existing relays, RTUs, CT/PT inventory, time sources, communications) → interim-compliance configuration of existing capability (event-capable relays and existing controllers configured and retained now, protecting the owner under already-live retention and R8 obligations during the construction window) → standardized retrofit design (a repeatable RTAC + Axion pattern with a site-specific channel map) → prioritized rollout sequenced to the 50% and 100% milestones and to outage availability. Legacy controllers below the modern compute tier are replaced with project conversion; existing feeder relays keep their protection role and contribute FR coverage where their recording meets the specifications.

12.2 Greenfield Strategy

On a new facility, the cheapest disturbance monitoring program is the one specified before procurement closes: dedicated measurement CT cores in the substation package; unit-level SER capability written into the inverter OEM scope during factory coordination — new units get no capability relief; the DFR channel math sizing controller and storage; monitoring operational and commissioned at COD because the new-resource schedule effectively requires energizing compliant; and the evidence package assembled as construction documentation rather than reconstructed afterward. A facility designed this way satisfies PRC-028 as a byproduct of good substation engineering — and arrives pre-instrumented for the PRC-029 ride-through demonstrations and PRC-030 analyses that follow.

12.3 The One-Platform Dividend

The architecture in this book is deliberately not a single-purpose recorder. The same controller that discharges PRC-028 carries the plant RTU and SCADA concentration, protocol conversion across DNP3, IEC 61850, Modbus, and IEC 60870; the HTML5 HMI; secure engineering access to every IED; PMUs for wide-area visibility; and the IEC 61131 logic engine for plant automation. Owners who treat the compliance project as the modernization project — one platform, one software environment across every RTAC model, one evidence discipline — recover much of the compliance cost in operating-and-maintenance simplification.

How Keentel Helps .

About Keentel Engineering

Keentel Engineering is a power systems and grid interconnection consulting firm headquartered in Tampa, Florida, with offices in Austin, Sacramento, and Baltimore. Service lines include point-of-interconnection and grid interconnection engineering, power system studies, substation and transmission design, EMT modeling and power quality, renewables and battery energy storage engineering, NERC compliance, and owner's engineer services. Florida Engineering Firm Registry No. 36853.

2.1 Who and What Is Covered

Engineering Note — Read the Final Standard, Not the Drafts

The unit-level recording obligations changed materially between the circulated drafts and the adopted PRC-028-1, particularly regarding inverter-level data and "if capable" relief for existing equipment. Programs scoped against draft language have, in Keentel's experience, carried unnecessary hardware cost. Re-baseline every requirement interpretation to the final enforceable text and its Implementation Plan before committing procurement.

SER is the time-stamped log of discrete state changes: a breaker opened, a disconnect closed, an inverter posted a fault code, a ride-through mode asserted. Its value is causal ordering — establishing, to millisecond resolution, what happened first. For IBR facilities the standard's reach into inverter-level information is the novel element: fault codes, alarms, and ride-through status originate inside the OEM's control system, which means SER design is as much a vendor-coordination exercise as a wiring exercise. Existing units may have qualified relief where the equipment is not capable; new units should have unit-level SER capability specified in the procurement documents, because no such relief attaches to equipment an owner chooses to buy.

3.2 Fault Recording (FR)

FR is triggered, high-speed oscillography: point-on-wave voltage and current captured around a disturbance, the raw material of protection analysis. The standard defines both where FR must exist — the facility's main power transformers, collector feeders, and reactive devices — and what must wake it up: residual overcurrent, phase over- and undervoltage, and frequency excursions. FR answers what the electrical system actually did during the fault window: current magnitudes and asymmetry, voltage depression and recovery, breaker clearing.

3.3 Dynamic Disturbance Recording (DDR )

DDR is continuous, phasor-domain recording of the quantities that describe system dynamics — voltage, current, real and reactive power, frequency — captured without interruption so that slow phenomena invisible to a triggered recorder are preserved: power oscillations, frequency events, voltage excursions, and the IBR control-system behavior between and after faults. DDR is what makes ride-through performance demonstrable under PRC-029 and what feeds model validation. Because it never stops, DDR is fundamentally a storage-engineering problem, which Chapter 8 quantifies.

The architecture Keentel deploys most often for PRC-028 compliance pairs an SEL Real-Time Automation Controller with SEL-2240 Axion I/O over a dedicated EtherCAT network. One platform then discharges all three data obligations: Axion protection-class modules provide the high-fidelity measurement front end, the RTAC's Digital Fault Recorder extension and continuous recording groups provide FR and DDR, the tag and SOE infrastructure provides SER, and the same controller carries the plant's SCADA, protocol conversion, and HMI duties it would host anyway. The DFR capability is documented by the manufacturer as designed to exceed the recording requirements of both PRC-002 and PRC-028.

A defensible R6 design has three layers. The reference is a GNSS-disciplined substation clock. Distribution reaches the recorder as demodulated IRIG-B and/or Precision Time Protocol — on the modern RTAC platforms, PTP is supported on standard Ethernet ports, and the SEL-3390T card accepts a high-accuracy PTP input with sub-microsecond performance, then redistributes time downstream over IRIG-B and time protocols so every connected IED shares the reference. Verification is continuous: the DFR's local monitoring includes a Synchronized status derived from the presence of high-quality IRIG-B or PTP sync on the controller and on every attached AC protection module, and that status belongs on SCADA and in the alarm philosophy — because unsynchronized records discovered after an event are a loss that cannot be repaired retroactively.

Designing to microsecond-class accuracy rather than the millisecond class that event sequencing strictly requires buys two things: the plant's DDR channels are simultaneously synchrophasor-grade (each DFR asset already instantiates a PMU), and the facility is future-proofed against oscillation-analysis and wide-area monitoring use cases that arrive with the PRC-029/PRC-030 lifecycle. The time architecture drawing, clock data sheet, and sync-verification records are all Measure evidence.

Keentel Engineering delivers PRC-028 compliance as turnkey engineering: portfolio applicability screening and per-site gap assessments against R1–R8; disturbance monitoring equipment specification and procurement support; detailed design of the RTAC + Axion DFR architecture including channel maps, storage sizing, time synchronization, and network design; interim-compliance configuration of existing relays and controllers; DFR extension configuration, commissioning, and three-layer verification; and audit-ready documentation packages mapping every requirement to its evidence.

Because the same standards lifecycle continues into PRC-029 ride-through demonstration and PRC-030 disturbance analysis — and because Keentel's practice spans grid interconnection, EMT modeling, protection, and NERC compliance — the disturbance monitoring program we build is designed for everything the data will be asked to prove. Contact us at 813-389-7871 or contact@keentelengineering.com.

CRG — Continuous Recording Group — the RTAC's continuous-recording construct enabling on-demand COMTRADE extraction from a rolling high-rate archive.

DDR — Dynamic Disturbance Recording — continuous recording of phasor-domain quantities (V, I, P, Q, f) describing system dynamic behavior.

DFR — Digital Fault Recorder — the integrated recording function; on this platform, an ACSELERATOR RTAC extension that builds the complete recorder configuration.

EtherCAT — The deterministic industrial Ethernet fieldbus connecting the RTAC to Axion I/O modules; dedicated port, non-routable, direct connections only.

FR — Fault Recording — triggered, high-speed point-on-wave capture of voltages and currents around a disturbance.

IBR — Inverter-Based Resource — solar PV, Type 3/4 wind, battery storage, HVDC-connected and hybrid facilities interfacing through power electronics.

PMU — Phasor Measurement Unit — synchrophasor source; the DFR extension auto-generates one per configured asset.

POI — Point of Interconnection — the boundary at which the facility connects to the transmission system; primary DDR location.

PRCTPT — The Axion AC protection module class (SEL-2245-42) terminating CT/PT secondaries for recording and measurement.

SER / SOE — Sequence of Events Recording — time-stamped discrete event logging; SER is the standard's term, SOE the platform's.

Appendix B — Glossary

Get the complete guide to PRC-028 monitoring, recording, and compliance requirements.

For a century, the machines that generated North America's electricity also explained themselves. A synchronous generator responds to a grid fault with physics — inertia, field dynamics, damper windings — and the protective relays around it capture oscillography as a matter of course. When something went wrong, investigators had data.

Inverter-based resources broke that assumption. A solar plant, wind facility, or battery energy storage system responds to a disturbance not with rotating mass but with control code, changing active and reactive power output in milliseconds according to firmware logic that varies by manufacturer, model, and settings file. When fleets of IBRs began tripping or reducing output during transmission faults — the Blue Cut Fire event in California, the Canyon 2 Fire event, and the Odessa disturbances in Texas — investigators repeatedly hit the same wall: the plants that misbehaved had little or no high-resolution recording. Facilities had been built with minimal oscillography, no unit-level visibility into inverter fault codes, and inconsistent time synchronization. Event investigations stalled, dynamic models could not be validated against measured behavior, and the same failure modes repeated across the fleet because nobody could prove exactly what had happened.

The regulatory response arrived through FERC Order No. 901, which directed NERC to close the IBR reliability gaps, and through a family of new Protection and Control standards developed under NERC's IBR work: PRC-028-1 for disturbance monitoring, PRC-029-1 for ride-through performance, and PRC-030-1 for post-disturbance analysis and mitigation. Of the three, PRC-028-1 is the foundation — the other two standards depend on the measured data it mandates. Ride-through performance cannot be demonstrated, and unexpected behavior cannot be analyzed, without records.

PRC-028-1, Disturbance Monitoring and Reporting Requirements for Inverter-Based Resources, became effective on April 1, 2025. For owners of utility-scale solar, wind, storage, and hybrid facilities, the practical question is no longer whether to install disturbance monitoring equipment, but how to do it economically — at existing sites that were never designed for it, and at new sites where the monitoring must be operational at commercial operation. This book lays out the standard's anatomy, the engineering specifications behind each requirement, and a reference architecture built on the SEL Real-Time Automation Controller (RTAC) and SEL-2240 Axion platform that Keentel Engineering deploys to achieve — and document — compliance.

Chapter 1 — Why PRC-028 Exists: The Inverter-Based Resource Visibility Gap

Data Center Design Guide | Grid, Power & Infrastructure

Sat, 11 Jul 2026 20:14:18 GMT

Jul 11, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Data Center Concepts and Design From Grid Interconnection to the Rack

A data center is a building wrapped around an electrical system. Every other discipline — cooling, fire protection, security, networking, controls — exists to keep conditioned, uninterrupted, high-quality power flowing into IT equipment whose failure the business cannot tolerate. That framing matters because it determines where design risk actually concentrates: not in the architecture, not even primarily in the mechanical plant, but in the power chain that begins at the utility point of interconnection and ends at the server power supply. In the current market — where individual AI training halls request more power than small cities and utility interconnection queues stretch for years — the grid end of that chain has become the single largest determinant of whether a data center project succeeds, and on what schedule.

Keentel Engineering works this problem from both ends: as power systems and interconnection engineers who take large loads through utility and regional reliability processes, and as electrical design engineers who carry the campus from the intake substation through the UPS to the rack. This article lays out the complete design chain for a modern data center — siting, resilience philosophy, electrical and mechanical systems, air and thermal management, fire safety, commissioning, and operations — with particular attention to the decisions that are hardest to reverse later.

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

Executive Overview

2. Site Selection: Power Availability Is the New Location

Traditional site-selection criteria fiber diversity, latency to markets, climate for free cooling, geotechnical and flood risk, planning environment, aviation and security exclusions, workforce access all still apply. But the ordering has changed. In today's market, the availability, schedule, and firmness of utility power is the first filter, and it eliminates more candidate sites than every other criterion combined. Interconnection of a large load is no longer a service application; it is a multi-year engineering and regulatory process involving utility system impact studies, potential network upgrades, contribution-to-cost allocations, and increasingly regional reliability requirements aimed specifically at large loads, including ride-through expectations, telemetry obligations, and curtailment provisions.

Power due diligence Interrogate the utility early and quantitatively: available capacity at nearby transmission and distribution nodes, study queue position and timelines, upgrade scope and cost exposure, and the utility's posture on phased energization.
Climate and water Climate drives operating cost through free-cooling hours and water availability; both are now board-level issues as water-use effectiveness joins PUE in sustainability reporting.
Connectivity Dual diverse fiber entries, carrier presence, and latency or synchronous-replication distance to paired facilities constrain the shortlist further.
Planning risk Noise (nighttime plant and generator testing), fuel storage volumes, visual massing, and traffic are the recurring planning friction points; early pre-application engagement measurably shortens determination.
Risk analysis A formal risk schedule external threats, environmental exposures, utility futures, EMI sources, expansion capability compiled with all stakeholders and carried through design is the discipline that keeps siting decisions defensible.

The Keentel view

We treat the interconnection study and the site decision as one workstream. A site with land, fiber, and water but a five-year power timeline is not a site; it is an option. Quantifying that timeline with utility engagement, load-flow screening, and a realistic upgrade scope before land commitment is the cheapest risk mitigation available in data center development.

Resilience language is standardized around a simple vocabulary. N is the capacity required to serve the load. N+1 adds redundant components so any single unit can fail or be maintained. A concurrently maintainable design allows any component or distribution path to be taken out of service, planned, without interrupting the critical load which forces dual paths at least in alternate form. A fault-tolerant design withstands any single unplanned failure, anywhere, without load interruption which forces two simultaneously active paths (2N or 2(N+1)) and compartmentalization so one event cannot take both. The industry tier framework (Tier I through IV) maps onto exactly these distinctions: single path; single path with redundant components; concurrently maintainable; fault tolerant.

The essential discipline is that a tier level is a business decision expressed in engineering, not an engineering aspiration. Fault tolerance roughly doubles the electrical plant, expands space and cost accordingly, and imposes operating complexity that must be staffed and maintained for decades. A facility serving genuinely continuous, intolerant workloads earns that cost; a facility whose workloads can migrate or tolerate structured maintenance windows does not. Distributed-redundant and block-redundant architectures occupy the middle ground, sharing reserve capacity across systems to approach dual-path availability at lower plant count. What is not acceptable is ambiguity: the resilience level must be declared in the brief, because it drives planning consent, space program, utility capacity, cost, and commissioning scope from day one and mixed halls (a 2N hall and an N+1 hall in one building) are legitimate and increasingly common, provided the boundaries are explicit.

8. Fire Safety: Strategy First, Systems Second

Cooling design begins with the environmental envelope. Modern IT hardware tolerates supply air across a wide band — the industry-consensus recommended range extends to 27 °C at the rack inlet and allowable ranges reach considerably higher — and every degree of supply-air setpoint increase expands free-cooling hours and refrigeration efficiency. The constraints are legacy hardware that cannot tolerate the modern envelope (and must be zoned for, not designed around), humidity limits that modern equipment has relaxed substantially, rate-of-change limits, and — critically — the thermal ride-through question: at higher setpoints, the time between a cooling failure and IT inlet over-temperature shrinks, so the failure scenario, continuous-cooling provisions (UPS-backed fans and pumps, chilled-water buffer vessels), and rate-of-rise analysis must be engineered as deliberately as the steady state.

System selection then follows density and scale. Perimeter CRAC/CRAH units with underfloor supply serve conventional air-cooled halls; in-row and rear-door heat exchangers extend air cooling into the 20–40 kW band; and above that, liquid arrives — direct-to-chip cold plates handling the majority of rack heat with an air-cooled residual, or immersion for the densest and most specialized deployments. The chilled-water plant behind these choices carries its own resilience design: N+1 or better chillers and economizers, buffer vessels sized for compressor-restart ride-through, dual risers and valved sectionalizing so any unit or pipe segment can be maintained concurrently, and free-cooling economizers — air-side, water-side, or adiabatic-assisted — that in temperate climates now carry the majority of annual cooling hours.

Electronically commutated fans, floor-void pressure control, and variable-speed pumping close the loop on part-load efficiency, which matters because facilities spend years below design load, and plant that only performs at nameplate wastes energy through the entire fill-up curve.

Data center fire engineering starts with a documented strategy — business criticality, compliance basis, evacuation, smoke control, firefighting access, and protection philosophy — agreed among the operator, insurer, and authorities before systems are selected, because the systems are only the strategy's instruments. Detection in critical spaces is built on aspirating smoke detection (ASD/VESDA-class): networks of sampling pipes drawing air continuously to laser detectors sensitive enough to alarm at the incipient, pre-visible stage, positioned with the airflow (including in the return path to cooling units) rather than against a static ceiling assumption. Analogue addressable panels provide the per-device status and location precision the risk demands, and the entire response logic — first-knock investigation alarms, second-knock actions, air-handling shutdowns, damper closures, door releases, suppression release — is captured in a cause-and-effect matrix that is tested line-by-line at commissioning and after every subsequent change.

Suppression is layered. Pre-action sprinklers (dry pipes, water admitted only on independent detection) protect the building risk while making accidental discharge into live halls a two-failure event. Clean-agent gaseous systems (inert blends or fluoroketone agents) protect hall volumes without conductive residue — with design attention to room integrity testing, pressure relief venting, post-discharge purge, and the acoustic effect of discharge nozzles on hard drives, which is a real and documented failure mode addressed by low-noise nozzle design. Water mist suits generator and plant rooms; oxygen-reduction (hypoxic) systems, which hold the protected volume below combustion-supporting oxygen levels continuously, have matured into a credible prevention option for lights-out spaces. Containment changes all of it: a contained aisle is a separate compartment from the detection and suppression system's point of view, and retrofitting containment without re-engineering fire protection is a recurring and serious audit finding.

6. Air Management: The Cheapest Megawatt in the Building

In an air-cooled hall, the gap between installed cooling capacity and delivered cooling capacity is almost entirely an air management problem. Two loss mechanisms dominate. Bypass is conditioned air that returns to the cooling units without passing through IT equipment — leaking through cable cutouts, unsealed floor penetrations, and oversupplied grilles. Recirculation is IT exhaust that re-enters IT inlets — over the tops of rows, around row ends, through empty rack positions missing blanking panels. Bypass wastes fan energy and starves the hall; recirculation creates the hot spots that pressure operators into lowering setpoints, which destroys efficiency across the whole facility. It is entirely typical for a legacy hall to circulate twice the airflow its IT load requires and still run hot in the wrong places.

The remedies are physical segregation and measurement. Hot-aisle or cold-aisle containment, rack chimneys, blanking panels, brush-sealed penetrations, and balanced floor grilles convert the hall from a mixing volume into a ducted system; computational fluid dynamics (CFD) modeling validates layouts before racks land and diagnoses problems after; and inlet-temperature instrumentation at the rack face — not just at the cooling unit return — provides the control signal that lets setpoints rise safely. Containment interacts with fire protection (detection and suppression coverage inside contained aisles must be engineered, not assumed) and with occupational limits on hot-aisle working temperatures. Air management does not save energy by itself; it is the enabler that unlocks higher setpoints, lower fan volumes, and free cooling — which is why it is the highest-return retrofit in almost every legacy facility we assess.

3. Resilience Philosophy: Choosing a Topology Before Choosing Equipment

4.1 From the intake down

Design proceeds top-down from the utility supply: voltage level, capacity and growth terms, metering point, and on-site substation scope are negotiated with the utility and frequently pace the entire program, since transmission-voltage interconnections for large campuses involve utility-grade substation engineering — bus configurations, protection and relaying, revenue metering, SCADA/RTU telemetry to the utility — that is precisely conventional power engineering, not building services. From there, medium-voltage switchgear architecture (main-tie-main, ring, or breaker-and-a-half at the largest campuses), transformation to utilization voltage using cast-resin or ester-filled units, and low-voltage switchgear rated for the fault duty and arc-flash requirements of very large busses complete the distribution frame. Busbar ratings, breaker interrupting duties, and selective coordination across the entire chain must be proven by short-circuit, coordination, and arc-flash studies — the same study discipline Keentel applies on utility systems, applied inside the fence.

4.2 UPS, energy storage, and standby generation

The UPS bridges the gap between a utility disturbance and generator assumption of load, while continuously conditioning power quality against sags, swells, and harmonics. Static (double-conversion) UPS with lithium-ion or advanced lead batteries dominates new construction, with modern units offering high-efficiency eco or dynamic modes that materially move facility PUE. Rotary and diesel-rotary (DRUPS) machines remain compelling at large unit sizes for their ride-through physics and fault-current contribution, at the cost of mechanical maintenance regimes. Configuration — N+1 within a system, 2N across systems, or distributed redundant — must mirror the declared resilience topology, and neutral-earthing arrangements across UPS, bypass, and generator operating modes require deliberate design: three- versus four-pole switching decisions made casually at design time surface later as nuisance trips, circulating neutral currents, or unsafe isolation conditions.

Standby generation is sized not just for the IT load but for the mechanical plant that must restart with it, with starting-sequence studies proving the block-load and motor-starting behavior. Fuel storage sizing (commonly 24–72 hours at full load), fuel polishing and refill logistics, emissions permitting, and — increasingly — the question of whether the generation asset can provide grid services or operate on renewable diesel are all design-stage decisions. The final meters of the chain — PDUs with static transfer switches, busway with tap-off units, rack power panels, dual-corded IT with A/B feeds — determine whether the facility's theoretical redundancy actually survives at the rack, and single-corded legacy equipment quietly defeats dual-path architectures unless caught by point-of-use transfer switches.

4.3 Monitoring and controls

A modern facility carries an electrical power monitoring system (EPMS/SCADA) over the switchgear, UPS, and generation; a building management system (BMS) over the mechanical plant; and increasingly a data center infrastructure management (DCIM) layer correlating both against IT load at the rack. Designing these as one instrumented architecture — consistent naming, synchronized time, a single alarm philosophy, sub-metering granularity that supports both PUE reporting and capacity management — is dramatically cheaper than reconciling three vendor silos after energization. The same discipline Keentel applies to substation SCADA points lists applies here verbatim.

5. Mechanical Systems: Cooling as a Chain, Not a Component

Power usage effectiveness — total facility energy divided by IT energy, as an annualized value — remains the industry's shorthand metric. Legacy facilities commonly ran at 2.0 and above, dominated by compressor energy; current best-practice designs deliver 1.2–1.4, and free-cooling-led architectures in favorable climates press toward 1.1. The strategy stack is hierarchical and interdependent: rationalize and utilize the IT itself (the only watts that do revenue-bearing work); manage air so setpoints can rise; raise setpoints so refrigeration works less and economizers work more; strip electrical losses (high-efficiency UPS modes, right-loaded transformers — including amorphous-core units that cut no-load losses dramatically in the lightly-loaded redundant configurations data centers force — and LED lighting on controls); and only then consider supply-side measures such as on-site generation, heat recovery to district or campus users, and renewable procurement. PUE is also widely misused: it fluctuates with IT load and climate, reaches design value only at design load, and says nothing about water or carbon — so mature operators now report PUE alongside water usage effectiveness and carbon intensity, and increasingly against embodied-carbon assessments of the M&E plant itself, which typically exceeds the building structure's embodied footprint.

9. Network and Security: The Other Two Utilities

7. Energy, PUE, and the Sustainability Ledger

This content package presents Keentel Engineering's integrated perspective on data center concepts and design, organized in three parts. A long-form technical blog walks the complete design chain: what a data center functionally is; site selection in a power-constrained market; resilience philosophy and topology selection; electrical systems from the utility intake through UPS, generation, and rack distribution; mechanical and cooling systems; air and thermal management; energy, PUE, and the widening sustainability ledger; fire safety; network and physical security; commissioning through integrated systems testing; operational governance; and the industry's trajectory under AI-scale demand. A practitioner FAQ addresses the fifteen questions developers, enterprises, and operators raise most often. Three anonymized case studies then demonstrate the approach on representative project archetypes: a hyperscale AI campus interconnection, an enterprise commissioning rescue, and a high-density liquid-cooling retrofit in a live legacy facility.

The unifying thesis: the data center industry's binding constraint has moved outside the building. Power availability, interconnection timelines, large-load reliability obligations, and grid-interactive load behavior now pace projects more than any construction trade — which makes utility-grade power engineering, applied on both sides of the meter as one system, the discipline that determines whether a data center project succeeds and on what schedule.

4. Electrical Systems: The Spine of the Facility

A data center's reliability is not designed in; it is proven in. The completion phase is structured as a five-level progression: factory acceptance tests on major plant (witnessed, at full and part load); installation verification by contractors; witnessed component and system demonstrations to the owner's team; inter-system interface testing — UPS to generator to BMS to fire cause-and-effect — proving the control logic among systems; and finally integrated systems testing (IST), the choreographed demonstration that the whole facility rides through every credible failure scenario at load, typically against heat-bank load simulating the IT. The black-building test — deliberately opening the utility supply and watching the facility carry itself — is the graduation exercise. IST is scheduled first and defended absolutely, because it is the only phase that ever gets compressed when construction slips, and every uncommissioned failure mode is simply an outage on deferral.

Completion management wraps the testing in evidence: a commissioning plan with defined key deliverables, a roles-and-responsibilities matrix across owner, contractors, and commissioning agent, test documentation packs, O&M manuals with plant replacement strategies, as-built records, clinical-clean certification of halls before IT wiring, operator training with proof of competency, and soft-landings support through the first operating seasons. The deliverable is not a certificate; it is an operations team that has already seen the facility fail safely, on purpose, before the business moved in.

10. Commissioning and Completion: Where Paper Resilience Becomes Real

Network provision follows the same resilience logic as power: diverse building entries into physically separated intake rooms, carrier and route diversity verified to the street (not just contractually asserted), structured cabling engineered to current fiber and copper standards with containment, bend-radius, and capacity headroom for growth, and a core/distribution/access architecture that can absorb speed generations without recabling. Physical security is engineered in concentric layers — perimeter fence and PIDS, vehicle interlocks and crash-rated gates, gatehouse and lobby procedures, mantrap or interlocked corridor entries, and data hall access on multi-factor credentials with anti-tailgating measures — each layer instrumented with surveillance positioned to identify rather than merely observe. The design principle is that security, like resilience, is a documented threat-and-risk posture expressed in construction, and the loading bay — the one place where the perimeter deliberately opens — deserves the most careful interlock design in the building.

11. Operations: Engineering the Decades After Handover

1. What a Data Center Actually Is: The Functional Anatomy

12. Where the Industry Is Going — and Why It Leads Back to the Grid

The disciplines that keep a commissioned facility at its designed availability are procedural, and they decay without governance. Change management treats every add, move, and maintenance activity in the critical envelope as a controlled event with method statements and approvals. Capacity management gates every new rack against measured — not nameplate — power and cooling headroom, preventing the slow accretion of hot spots and stranded capacity.

Environmental monitoring holds the hall to its envelope continuously; emergency operating procedures are drilled as walk-throughs until the night shift can execute them from memory; escalation paths are documented from first detection through business-continuity notification; and cable hygiene under floors and in containment is enforced so that the airflow paths and spare pathways the design assumed still exist in year twelve. Facilities of identical design diverge in availability over a decade almost entirely on the strength of these procedures — which is why operational readiness belongs in the design scope, not after it.

Every current trend line points the same direction. AI-class racks have pushed hall densities past what air can carry, making direct-to-chip liquid cooling standard in new high-performance builds and hybrid air/liquid halls the transitional norm. Higher IT thermal tolerance and better air management have pushed PUE toward its practical floor, moving the sustainability frontier to water, embodied carbon, and heat reuse. Modular and phased construction lets infrastructure follow load instead of preceding it by years. And facility scale has grown to the point where individual campuses are transmission-system events: hundreds of megawatts to gigawatts, drawing regional reliability attention, ride-through and telemetry requirements, curtailment frameworks, and interconnection study queues that now pace the industry more than any construction trade. Emerging supply-side options — on-site fuel cells, gas turbines with heat recovery, storage participating in grid markets, and eventually small modular reactors — are all, at bottom, grid-interconnection engineering problems wearing a real-estate costume.

The Keentel thesis

The data center industry's binding constraint has moved outside the building. The engineering disciplines that now determine project success — interconnection studies, transmission and substation design, protection and SCADA, large-load reliability compliance, power quality — are utility power engineering, applied to a private campus. Keentel brings exactly that practice to data center developers and operators: we take the project through the utility process and design the power infrastructure on both sides of the meter as one system, from the transmission tap to the busway tap-off.

Case Studies

Key takeaway

At campus scale, the interconnection is the project. Treating the grid side as a managed engineering program — phased studies, owner's-engineer substation design, load-behavior mitigation, and compliance mapping — converted a pro-forma-threatening timeline into a stageable one, and the building program inherited a schedule it could actually build to.

The following case studies are drawn from representative Keentel Engineering project experience. All client names, project names, locations, and identifying details have been anonymized or generalized; quantitative parameters have been rounded or adjusted to preserve confidentiality while retaining engineering fidelity.

Case Study 1: Grid Interconnection and Power Infrastructure for a Hyperscale AI Campus

Background

A data center developer secured land for a multi-phase campus targeting AI training workloads — roughly 300 MW at ultimate build-out, phased from an initial 60 MW hall. The site's economics depended on a transmission-voltage interconnection to a 230 kV corridor, a dedicated customer substation, and an energization schedule that delivered first power inside the anchor tenant's contractual window. The developer's building and mechanical design was well advanced; the grid side of the project — utility studies, substation engineering reliability obligations — had been treated as a utility deliverable and was unmanaged.

The Challenge

The utility's initial screening indicated the full 300 MW would trigger network upgrades with a timeline extending years past the anchor tenant's requirement, threatening the entire pro forma.
The regional reliability framework imposed emerging large-load requirements — telemetry to the transmission operator, voltage and frequency ride-through expectations, and participation in a curtailment framework — that no one on the project had scoped, priced, or designed for.
The AI load profile itself raised utility concern: multi-megawatt step changes as training jobs start, stop, and checkpoint, with power-quality and oscillatory behavior the utility required to be studied before it would finalize the interconnection agreement.
Long-lead equipment — the 230 kV power transformers above all — had procurement lead times that would consume the entire schedule if released against the building program rather than the interconnection program.

Keentel's Approach

Restructured the interconnection request into a phased capacity profile matched to hall build-out, allowing the utility to study and approve an initial tranche served by existing system capability while upgrades for later phases proceeded in parallel — converting a single cliff-edge study into a stageable program.
Served as owner's engineer for the customer 230 kV substation: bus configuration selection, protection and relaying design coordinated with the utility's remote terminals, revenue metering, station SCADA/RTU points engineered to the transmission operator's telemetry specification, and grounding design from field-measured soil resistivity.
Performed the load-behavior studies the utility required — modeling aggregate rectifier front-end harmonics against the applicable power-quality standard and characterizing training-load step and swing behavior — and engineered the mitigation: on-site BESS buffering specified to smooth the fastest load transients, with ride-through settings coordinated against the reliability framework's envelopes.
Built the compliance map for the campus's large-load obligations — telemetry, ride-through, curtailment participation, and data submissions — into a single owner-facing document with design responsibilities assigned across the developer, utility, and tenant.
Issued the long-lead procurement package (transformers, 230 kV breakers, MV switchgear) against interconnection milestones with factory-test witnessing scoped into each order.

Results

Key takeaway

Component tests prove equipment; only integrated failure testing proves a facility. An independent power engineer's review — topology against the resilience claim, studies against the installed equipment, and an IST program defended at full scope — is the difference between a certificate at handover and a facility that has already failed safely, on purpose, before the business moved in.

Case Study 2: Electrical Resilience Verification and Commissioning Rescue for an Enterprise Facility

Background

A financial-services enterprise was weeks from occupying a new 8 MW facility marketed to its board as concurrently maintainable with fault-tolerant power. Construction was substantially complete and individual systems had passed their component tests, but the integrated systems testing window had been compressed by construction slippage to a fraction of its planned duration, and the owner's operations team — inheriting a facility they had not operated — asked for an independent review before accepting handover. Keentel was engaged as the owner's independent power engineer.

The Challenge

Design review against the declared resilience claims found three topology defects: a shared mechanical switchboard feeding both 'independent' chilled-water pump groups, a single non-redundant control power source supplying the paralleling switchgear controls, and UPS neutral-earthing arrangements that produced an undefined earth reference in one bypass operating mode.
The short-circuit and coordination study of record predated a mid-construction switchgear substitution; breaker settings in the installed equipment did not match the study, and selective coordination between UPS output boards and downstream PDU breakers was unproven.
The compressed IST plan covered normal transfers only — no black-building test, no failure-during-failure scenarios, no fire cause-and-effect integration under load.
The generator plant had never been proven against the true block-loading sequence including mechanical restart, and heat-bank capacity on site was insufficient to load the facility realistically.

Keentel's Approach

Issued a topology remediation package for the three defects — an independent second source and transfer arrangement for the shared mechanical board, redundant DC control power for the paralleling controls, and a corrected earthing scheme with four-pole switching at the defined transition points — sequenced so remediation ran parallel to remaining construction rather than behind it.
Rebuilt the short-circuit, coordination, and arc-flash studies around the as-installed equipment, issued corrected settings across the chain, and verified selective coordination from the utility intake to the PDU branch level, including the UPS current-limit behavior that the original study had ignored.
Re-scoped IST around a scenario matrix graded by risk: full black-building test, single-failure scenarios in every distribution block, failure-during-maintenance cases matching the concurrently-maintainable claim, and fire cause-and-effect actions executed with the plant at load — with rented heat banks bringing test load to a representative level.
Ran the owner's operations team through every scenario as operators, not observers, converting IST into the training program the compressed schedule had eliminated; emergency operating procedures were corrected in real time as tests exposed gaps.

Results

Key takeaway

Case Study 3: High-Density AI Retrofit — Power, Cooling, and Capacity Recovery in a Legacy Hall

Background

A colocation operator needed to land an AI inference customer requiring a pod of liquid-cooled racks at roughly 90 kW each inside an operating facility designed a decade earlier for 6 kW air-cooled cabinets. The facility was contractually full on paper — its power and cooling capacity fully allocated — yet measured utilization told a different story: hall-level metering showed substantial headroom stranded by conservative nameplate allocations, poor air management, and density drift. The operator's question was whether the pod could be accommodated without a new building phase, and without disturbing existing customers.

The Challenge

Nameplate-based capacity accounting had allocated the electrical plant to roughly its rating while measured demand ran far lower — but recovering that headroom contractually and technically required defensible measurement, not assertion.
The hall's air management was degraded in the classic pattern: missing blanking panels, unsealed floor penetrations, uncontained aisles, and cooling units in fan-speed competition — producing hot spots that had already forced setpoints down and would make any density addition appear impossible.
The 90 kW pod needed a liquid cooling loop — coolant distribution units, secondary piping, water-quality management — introduced into a live hall with no shutdown tolerance, plus busway-fed A/B power at currents the original PDU architecture could not serve.
Fire detection and suppression had been engineered for the original open-hall airflow; the retrofit's containment and the pod's enclosed architecture invalidated those assumptions.

Keentel's Approach

Established the measured baseline first: three months of coincident-demand data at hall, UPS, and PDU level, reconciled against contractual allocations, quantifying recoverable electrical headroom with an engineering margin the operator's counsel could defend in customer discussions.
Executed an air-management recovery program across the existing hall — containment on the highest-density existing rows, blanking and sealing throughout, floor-grille rebalancing, and unit fan coordination — validated by CFD before and instrumented rack-inlet measurement after, which eliminated the hot spots and allowed a staged setpoint increase.
Designed the pod's power path as a dedicated block: new busway from a recovered UPS output position, A/B feeds to rack level, updated short-circuit, coordination, and arc-flash studies for the modified distribution, and EPMS integration so the pod's fast load swings were visible and alarmed distinctly from the legacy hall.
Engineered the liquid loop as a closed secondary system with N+1 CDUs, leak detection zoned to isolate automatically, and commissioning procedures written for live-hall execution under the operator's change-management framework.
Re-engineered fire protection for the new geometry: aspirating detection extended into the contained aisles and pod enclosure, cause-and-effect matrix revised and retested line-by-line, and suppression coverage analysis updated for the changed airflow.

Results

Frequently Asked Questions: Data Center Design, Power, and Operations

Disclaimer

This document is published by Keentel Engineering for general informational and educational purposes. It does not constitute engineering services, and no engineering, procurement, or investment decisions should be made on the basis of this document without project-specific analysis performed by a licensed professional engineer. Standards, utility processes, reliability requirements, and industry practices referenced herein evolve; readers must consult the current requirements applicable to their jurisdiction, utility, and facility.

All product names, standards designations, organization names, rating frameworks, and trademarks referenced in this document are the property of their respective owners. Keentel Engineering is not affiliated with, endorsed by, or sponsored by any standards body, certification organization, institution, utility, equipment manufacturer, or other organization referenced herein. References are made solely for identification and educational purposes.

Case studies in this document are anonymized composites drawn from representative project experience; identifying details have been removed or altered, and quantitative results are illustrative of typical outcomes rather than guarantees of future performance.

IEC 61850 SCADA Engineering Guide | ACSELERATOR Architect

Sat, 11 Jul 2026 12:45:25 GMT

Jul 11, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

IEC 61850 SCADA Engineering With ACSELERATOR Architect

1. From Point Maps to Semantic Models

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

From Server Models to GOOSE That Behaves in Production

2. The Protocol Triad: GOOSE, MMS, and Sampled Values

For decades, SCADA integration meant point mapping. DNP3 asks the integrator to assign index numbers; Modbus asks for register addresses. Every device, every gateway, and every master station carries its own spreadsheet of numbers, and every firmware change or panel addition reopens that spreadsheet. The mapping itself carries no meaning index 47 is a breaker status only because a human wrote it down somewhere.

IEC 61850 inverts this model. Devices are self-describing: the standard defines a hierarchical, vendor-agnostic data model Logical Devices, Logical Nodes, Data Objects, and Data Attributes so that a breaker position, a frequency measurement, or a distance element operation carries the same semantic name regardless of which manufacturer built the relay. A latch bit in a feeder relay is not "index 47"; it is a status value inside an annunciation logical node, discoverable by any compliant client. The engineering consequence is profound: instead of maintaining point maps, the integrator engineers a model, and the model travels with the project in standardized files.

IEC 61850 is not one protocol but a suite, and each member has a distinct role in the substation communications architecture. Selecting the right protocol for each data flow is the first architectural decision in any 61850 design.

Substation Configuration Language (SCL) files are the interchange currency of IEC 61850 engineering. Four file types matter, and understanding which file owns which stage of the workflow prevents the most common integration failures.

Architect imports and exports all four types across Edition 1, Edition 2, and Edition 2.1 of the standard a practical necessity, because brownfield substations routinely mix Edition 1 legacy devices with Edition 2.1 replacements, and the system tool must speak to all of them. The workflow is symmetric across vendors: SEL ICD files are imported, configured for GOOSE, SV, or MMS, and pushed to devices as CID files; third-party SCL files are imported so SEL devices can subscribe to other manufacturers' GOOSE, SV, or MMS publications. That bidirectional interoperability is the entire value proposition of the standard, and the system configuration tool is where it is either realized or lost.

8. Production Hardening: The Checklist the Screenshots Skip

The most consequential capability in recent versions of Architect is the flexible server model: the ability not merely to view but to edit the IEC 61850 server model of supported SEL devices. Model editability is gated by the device's Class File Version (CFV). CFV 006 files corresponding to SEL's initial Edition 2 implementation can be viewed; CFV 010 and later files corresponding to the initial Edition 2.1 implementation can be edited, with real-time automation controllers editable from CFV 006 onward. Edition 2.1 is more than a version bump: it extends the model so an IED can represent the primary power system equipment it is connected to, along with enhanced system controls and feedback monitoring.

5.1 Logical Devices and Logical Nodes

Custom logical devices let the engineer organize logical nodes around the physical functions of the installation a line, a transformer bay, a reclosing scheme rather than accepting only the manufacturer's factory grouping. The recommended discipline is to add new logical devices and logical nodes for customization rather than editing manufacturer defaults. Default-model editing is deliberately gated behind a project setting with an explicit warning, for good reason: the edits are stored in the project file, an upgrade tool may silently revert them, and the device's model will no longer match its instruction manual. Treat the default model like vendor firmware extend it, don't rewrite it.

5.2 Logical Node Variants and NSD Namespaces

When adding a logical node, the engineer chooses among variants with explicit provenance: manufacturer-defined variants tuned to specific product families, pure IEC 61850-7-4 nodes imported from standardized namespace definition (NSD) files that include every optional data object the leaner variants omit, and user-defined variants maintained in a project data-type library. NSD import is the mechanism by which an organization's own standardized data model the logical node conventions a utility has refined across years of projects becomes a reusable engineering asset inside the tool.

5.3 Common Data Class Variants and the Verification Burden

Each data object carries a Common Data Class (CDC), and the tool allows CDC variant selection per object. Here a caution is warranted: the dropdown may list CDCs the target device cannot actually support, particularly when user-supplied type libraries introduce attributes or enumerations outside the device's implementation. The instruction manual not the configuration tool remains the authority on what the device supports. Good practice is to stay with default-populated variants unless there is a specific, verified reason to deviate.

5.4 Binding the Model to the Device: sAddr Mapping

The final act of server-model engineering is binding IEC 61850 data attributes to the device's internal reality relay word bits and analog quantities via the short address (sAddr) column. This mapping is freeform text and the tool performs no validation of what is entered: a mistyped bit name produces a model that validates against the schema and silently reports nothing. This is precisely the kind of gap that disciplined design review exists to close every sAddr binding should be verified against the device documentation and then proven end-to-end during commissioning, never assumed.

For large point counts, the spreadsheet layout with full CSV export and re-import turns model engineering into a bulk operation: export the model, engineer the bindings in a controlled spreadsheet with review sign-off, and import the result back an auditable workflow instead of hundreds of manual clicks.

Several disciplines separate a configured system from a hardened one. Factory-default device credentials must be retired before energization configuration transfer workflows will happily use them, and so will an adversary. Configuration revision (confRev) values should auto-increment with notification enabled, so every subscriber can detect that a publisher's dataset changed shape and configuration drift is caught at the moment it is created rather than during a misoperation investigation. SCL files should pass schema validation before site work, and semantic comparison between revisions should be part of change control, so that every modification to the communications system is as reviewable as a relay settings change. Finally, the printed subscription and report documentation belongs in the project turnover package the communications model is a deliverable, not a byproduct.

How Keentel Helps

Keentel Engineering delivers IEC 61850 substation communications design as an integrated part of protection, SCADA, and grid interconnection engineering not as an afterthought. Our engineers develop the full SCL project architecture, engineer server models and GOOSE schemes that respect network capacity and configuration-management discipline, integrate multi-vendor fleets through standards-based SCL exchange, and build commissioning plans that verify transport, data, and application layers independently.

From single-substation automation upgrades to utility-scale solar, storage, and data center interconnections where the SCADA architecture must satisfy both plant controls and utility master requirements, Keentel provides owner's engineer oversight, detailed design, and field commissioning support. Contact us at 813-389-7871 or contact@keentelengineering.com.

6. Engineering GOOSE That Behaves in Production

GOOSE deserves particular respect from the SCADA engineer because it lacks handshaking by design. A publisher broadcasts its dataset onto the LAN and never learns whether any subscriber received it. Speed is the point — but it means subscription health must be engineered deliberately, through supervision logic, quality-attribute monitoring, and time-to-live (TTL) diagnostics, rather than assumed.

A GOOSE scheme that works on the bench and a GOOSE scheme that behaves on a loaded substation LAN are separated by a handful of design decisions that the configuration screens make easy to skip.

6.1 Dataset Discipline: mag vs. instMag

Analog values expose multiple attributes a deadband-governed magnitude (mag), an instantaneous magnitude (instMag), quality, and timestamp. The distinction is not academic. The deadbanded magnitude updates only when the configured deadband is exceeded; the instantaneous magnitude updates on any change. Because every state change in a transmitted dataset triggers the fast retransmission burst, placing instantaneous magnitudes in a published dataset can turn one relay into a chattering traffic source that loads every device on the LAN. Production datasets should carry deadbanded values, with deadbands engineered deliberately in the project's dead-band configuration and quality and timestamp attributes included only where the subscriber actually consumes them.

6.2 Retransmission Behavior: Min Time and Max Time

GOOSE reliability comes from repetition. When a dataset member changes state, the publisher accelerates to the configured minimum interval, then decays back toward the maximum interval as the network quiets. These two settings define the scheme's latency floor and its steady-state background load, and they should be chosen as a system-level decision consistent across the project rather than left at whatever each device's dialog happened to default to.

6.3 Network Architecture: VLANs Are Not Optional at Scale

A two-device bench test runs happily on an unmanaged switch. A substation with dozens of publishers does not. Because GOOSE is Layer-2 multicast, every message reaches every port on a flat network, and every IED spends cycles discarding traffic it never subscribed to. Managed switches with VLAN segmentation confine each GOOSE application to the devices that actually consume it. VLAN IDs, priority tagging, MAC addressing, and APP IDs are assigned in the transmit message configuration and the project settings can enforce uniqueness of multicast MAC/APP ID combinations automatically, which should be enabled on any multi-engineer project.

6.4 Subscription Mapping on Both Ends

Publishing is half the scheme. On the subscribing side, binary points from remote devices map into the receiving relay's virtual bits, and analog points into its remote analog registers, where they become ordinary inputs to protection and control logic. On a gateway or automation controller, the same subscriptions arrive as single-point status and measured-value tags. The subscription map which remote attribute lands in which local variable is the true wiring diagram of a 61850 scheme, and it belongs in the project documentation with the same rigor as a terminal block drawing. Architect's printable subscription and report documentation exists precisely to make that deliverable auditable.

3. The SCL File Ecosystem: The Project's Single Source of Truth

SEL's engineering environment divides responsibility across three tools, and disciplined projects respect the boundary between them. Device settings protection elements, port parameters, front-panel logic live in ACSELERATOR QuickSet (SEL-5030). The IEC 61850 system configuration datasets, GOOSE transmit messages, subscriptions, server models, and addressing lives in Architect (SEL-5032). Gateway and automation-controller logic tag processing, protocol conversion, HMI data lives in ACSELERATOR RTAC (SEL-5033), which imports the Architect project so that every 61850 data point arrives as a named tag ready for user logic.

One architectural subtlety trips up newcomers: the real-time automation controller's data sources are mapped in its own software environment, not in Architect. For protective relays, Architect binds IEC 61850 data attributes directly to internal device points; for the automation controller, Architect defines the model and the controller's own software populates it. Knowing which tool owns which binding is the difference between a one-pass commissioning and a week of confusion.

Beyond core configuration, Architect carries a set of project-scale tools that matter on real systems: a Project Builder that reconstructs a lost SCD from surviving CID files a genuine field-recovery capability when documentation has drifted from reality; bulk CID deployment across every device in the project; dataset cloning to eliminate repetitive, error-prone re-entry across similar bays; an IED Upgrade tool for migrating old CID files to newer editions of the standard; a Semantic XML Compare for auditing exactly what changed between two SCL revisions; and a Schema Validator that checks files against the published IEC 61850 schema before a technician ever drives to site.

5. Server Model Engineering: The Data Model Is a Design Deliverable

A defensible commissioning plan for a 61850 SCADA integration verifies at three distinct layers, because each layer fails differently.

Transport

First: Is the message arriving at all? Device-level diagnostics report every GOOSE message published and subscribed, with time-to-live and error-code fields. A healthy subscription shows a live TTL countdown; an expired TTL signals the message stopped arriving before its validity window closed a network problem, not a data problem.

Data

Second: Are the right values landing in the right variables? Online tag inspection on the gateway confirms received values in real time, and force-value tools let the commissioning engineer drive a point from one end and watch it arrive at the other including forcing the quality attribute to prove that downstream logic honors data validity, not just data value.

Application

Third: Does the scheme do what the design intended? Toggling a test point and watching the mapped indication respond at the far end panel LED, HMI tag, or SCADA point closes the loop from model to mission. Terminal-level metering commands on the receiving relay provide independent confirmation that does not depend on the same software path being tested.

The transport-layer diagnostics confirm only that messages flow they cannot confirm that the correct values are being exchanged. Commissioning plans that stop at a live TTL have verified the network, not the scheme.

Case Studies

7. Commissioning: Verify at Three Layers

Substation communications are no longer an afterthought bolted onto a protection design. In modern substations, data center interconnections, and utility-scale renewable plants, the SCADA and automation architecture is a first-order design input — it determines how fast protection schemes coordinate, how cleanly a plant integrates with its utility master, and how much engineering rework a project absorbs when devices from multiple manufacturers must exchange data. This article examines how IEC 61850 changes the engineering discipline of SCADA integration, and how SEL's ACSELERATOR Architect (SEL-5032) software functions as the system configuration hub for GOOSE messaging, MMS reporting, and Sampled Values across SEL and multi-vendor fleets.

4. Where Architect Sits in the Toolchain

Keentel Engineering is a power systems and grid interconnection consulting firm headquartered in Tampa, Florida, with offices in Austin, Sacramento, and Baltimore. Our service lines include point-of-interconnection and grid interconnection engineering, power system studies , substation and transmission design, EMT modeling and power quality, renewables and battery energy storage engineering, NERC compliance, and owner's engineer services. Florida Engineering Firm Registry No. 36853.

Contact: 813-389-7871 · contact@keentelengineering.com · www.keentelengineering.com · 400 N Ashley Dr, STE #2600, Tampa, FL

Non-Affiliation Disclaimer

This document is original work product of Keentel Engineering, prepared for general informational and educational purposes. Keentel Engineering is not affiliated with, endorsed by, or sponsored by Schweitzer Engineering Laboratories, Inc. (SEL) or any other manufacturer referenced herein. ACSELERATOR, ACSELERATOR Architect, ACSELERATOR QuickSet, ACSELERATOR RTAC, SELOGIC, and related marks are trademarks or registered trademarks of Schweitzer Engineering Laboratories, Inc. IEC and IEC 61850 are marks of the International Electrotechnical Commission. All product names, trademarks, and registered trademarks are the property of their respective owners and are used for identification purposes only. Case studies are anonymized and generalized; figures are representative. Nothing in this document constitutes engineering advice for a specific installation consult a licensed professional engineer for project-specific application.

About Keentel Engineering

The following case studies are drawn from representative Keentel Engineering project experience. Client names, locations, and identifying details have been anonymized; figures are rounded and certain particulars generalized to protect confidentiality.

Case Study 1: Distribution Substation Automation Modernization for a Municipal Utility

Background

A municipal utility in the southeastern United States operated a 1980s-era 115/13 kV distribution substation with twelve feeder positions. Bus interlocking and breaker-failure schemes ran over hardwired copper between panels, and SCADA visibility was limited to a legacy RTU polling a partial point list. A feeder relay replacement program created the opening to modernize the communications architecture, but the utility's small engineering staff was wary of IEC 61850's configuration burden and of losing the ability to maintain the system in-house.

Challenge

Replace hardwired interlocking and breaker-failure signaling with GOOSE messaging across the new feeder relays and a substation automation controller, deliver a complete point list to the utility's SCADA master over its existing DNP3 WAN, and hand the utility a communications design its own staff could maintain with documentation that would survive personnel turnover.

Keentel's Approach

Keentel engineered the full IEC 61850 project in Architect: a substation SCD covering all twelve feeder relays and the automation controller, custom datasets sized to each scheme rather than factory defaults, and GOOSE transmit messages with project-standard Min/Max timing, VLAN assignments, and enforced-unique multicast addressing. Interlocking and breaker-failure signals were mapped relay-to-relay over GOOSE; the automation controller subscribed to every feeder's status and metering dataset and concentrated the plant into a single DNP3 outstation for the utility's master preserving the existing WAN and master-station configuration untouched.

Analog datasets were engineered with deadbanded magnitudes and deliberate dead-band settings to keep steady-state GOOSE traffic predictable on the new managed-switch LAN. Commissioning followed a three-layer verification plan: TTL-based transport checks on every subscription, forced-value data verification from each relay through the controller to the DNP3 master, and end-to-end application testing of every interlock permissive against the original hardwired logic before copper was lifted. Turnover documentation included the printed subscription map for every GOOSE connection and a change-control procedure built around configuration-revision tracking and semantic SCL comparison.

Outcome

The substation cut over with zero interlocking misoperations during parallel operation and retired approximately 1,400 feet of interpanel control cable. SCADA point coverage increased from a partial legacy list to full feeder status, metering, and event data roughly a fourfold increase in visible points with no changes to the utility's master station. The utility's own technicians executed a subsequent feeder addition using the documented workflow, adding the relay to the SCD, cloning an existing dataset, and deploying the CID without outside assistance the self-sufficiency outcome the utility had prioritized from the start.

Case Study 2: Multi-Vendor SCADA Integration for a Utility-Scale Solar and Storage Facility

Background

An independent power producer developing a 150 MW solar photovoltaic facility with co-located battery energy storage in the western United States faced a split-vendor reality common to utility-scale renewables: SEL protection and automation at the collector substation and point of interconnection, a third-party plant controller from the inverter OEM, and an interconnecting transmission owner whose facility requirements specified both hardwired and communications-based interfaces at the POI, with strict metering and status reporting obligations.

Challenge

Integrate the third-party plant controller with the SEL substation fleet over IEC 61850 despite the vendors' differing standard editions, satisfy the transmission owner's POI status, metering, and control requirements, and model plant-level quantities POI breaker status, aggregate MW/MVAR, curtailment and ride-through indications in a way that both the plant controller and the utility's systems could consume without custom protocol converters.

Keentel's Approach

Keentel built the integration around disciplined SCL exchange. The plant controller vendor's SCL files were imported into the Architect project so the SEL devices could subscribe directly to the controller's GOOSE publications; SEL device configurations were exported back to the vendor's system tool, with schema validation and semantic comparison run on every file crossing the vendor boundary. Edition differences between the vendors' implementations were resolved at the project level using Architect's multi-edition SCL support and the IED upgrade path for the devices that required it.

At the automation controller serving as the plant gateway, Keentel engineered a custom IEC 61850 server model organized around the plant's physical functions dedicated logical devices for the POI, the solar collector system, and the storage block so that plant-level aggregate quantities appeared as properly modeled, semantically named data objects rather than generic points. Binary and analog point counts were sized in the server model for the full build-out, avoiding a model rework when the facility's second phase added capacity. Where the third-party controller could publish analog values but subscribe only to binary data a capability asymmetry identified during design review of the vendor's documentation the signaling design was arranged so all analog flows ran toward the controller and command flows toward the plant ran as binary points, eliminating the mismatch before it reached the field.

Outcome

The facility completed communications commissioning ahead of its backfeed date, with every GOOSE subscription verified at transport, data, and application layers and the POI reporting package accepted by the transmission owner on first submission. The multi-vendor exchange required no custom middleware the SCL workflow carried the entire integration. When the second phase energized, the pre-sized server model absorbed the additional storage capacity with dataset edits only, and the semantic-comparison change record gave the transmission owner a complete, reviewable audit trail of exactly what changed in the communications system between phases.

Frequently Asked Questions: IEC 61850 SCADA Integration and ACSELERATOR Architect

Power System Protection and Relaying | SCADA Guide

Sat, 11 Jul 2026 09:58:49 GMT

Jul 11, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

POWER SYSTEM PROTECTION AND RELAYING IN THE SCADA-INTEGRATED GRID

Executive Overview

Technical Blog: Designing Power System Protection and Relaying for the SCADA-Integrated Grid

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

Design Philosophy, Relay Technology, Fault Analysis, Coordination, and Substation Automation

1. The Protection Design Philosophy: Five Properties in Tension

Protection engineering has one uncompromising mandate: detect an abnormal condition on the power system and isolate it in milliseconds, automatically, with the smallest possible portion of the network removed from service. Supervisory Control and Data Acquisition (SCADA) engineering has a complementary mandate: give operators real-time visibility and remote control over every breaker, relay, and measurement point in the system. For decades these two disciplines were designed by different teams, procured under different budgets, and commissioned on different schedules. That separation is no longer tenable. Modern numerical relays are simultaneously protection devices, fault recorders, revenue-grade meters, and SCADA data concentrators. A protection scheme that is not designed from day one as an integrated protection-and-SCADA architecture leaves reliability, compliance, and operational value on the table.

At Keentel Engineering, we design protection and relaying systems for utility substations, transmission lines, utility-scale renewable and BESS facilities, industrial plants, and large-load interconnections — and we treat the SCADA and communications layer as a first-order design input, exactly as we treat grid interconnection requirements. This article walks through the complete design chain: protection philosophy, instrument transformers, relay technology selection, fault analysis, coordination, equipment-specific schemes, high-impedance fault detection, grounding, and the SCADA architecture that binds it all together.

Every protection decision is a negotiated settlement among five properties. Reliability (dependability) demands the relay always operates for faults in its zone. Security demands it never operates when it should not. Selectivity demands that only the breaker(s) nearest the fault open, preserving service to the rest of the network. Speed demands fault clearing fast enough to protect equipment thermally and mechanically and to preserve transient stability. Sensitivity demands detection of the minimum credible fault — including high-impedance ground faults that may draw less current than normal load.

These properties conflict. Faster tripping erodes security margin. Higher sensitivity increases exposure to misoperation on load transients, inrush, and CT error. The protection engineer's job is not to maximize any single property but to document, for each zone, the deliberate trade-offs — and then to prove them with fault studies, coordination studies, and relay test plans. Systems are divided into overlapping zones of protection — generator, bus, transformer, transmission line, feeder, motor — with each zone bounded by circuit breakers and each relay assigned a primary zone plus remote or local backup duty. Overlap is achieved by lapping CT locations around breakers so that no point on the system is unprotected, and every primary scheme has a defined backup that operates with an intentional coordination delay if the primary fails.

Keentel design standard

Every protection single-line we issue carries a zone diagram with explicit CT overlap at each breaker, a primary/backup assignment matrix, and a documented fault-clearing time budget (relay operate time + breaker interrupting time + margin) checked against equipment withstand and stability limits.

Relays never see the primary system. They see the secondary output of current transformers (CTs) and voltage transformers (VTs or CVTs), and every error those transducers introduce propagates directly into relay reach, coordination margins, and differential stability. CT specification is therefore the first real engineering decision in any protection design.

2.1 Current transformer selection and saturation

A protection-class CT must reproduce fault current — including the fully offset asymmetrical waveform with DC component — without saturating before the relay makes its decision. The governing check is the CT's knee-point or accuracy-limit voltage against the voltage demanded by the maximum fault current flowing through the total secondary burden (relay input, lead resistance both ways, and CT winding resistance). Remanent flux from a prior fault clearance and the X/R-driven DC offset can multiply the effective flux demand several-fold; a CT that is marginally adequate for symmetrical current will saturate badly on an offset waveform, distorting the secondary current, delaying instantaneous elements, and destabilizing differential schemes through unequal saturation of paired CTs.

Our practice is to size protection CTs with explicit transient dimensioning: compute maximum through-fault current at the CT location from the short-circuit study, apply the system X/R ratio, include realistic lead burden from the physical cable schedule, and verify against the CT excitation curve or IEC/IEEE accuracy class rating. Metering-class accuracy and protection-class transient performance are different problems — a CT core optimized for 0.3-class revenue accuracy at load current is designed to saturate early, which is precisely what a protection core must not do. Dual-core or multi-ratio CTs resolve the conflict.

2.2 Voltage transformers and CVT transients

Magnetic (wound) VTs give faithful reproduction across the protection bandwidth and are standard at distribution and lower transmission voltages. At EHV, capacitive voltage transformers (CVTs) dominate on cost — but their tuned capacitive divider and compensating reactor produce a transient response of their own. When primary voltage collapses during a close-in fault, the CVT's stored energy discharges into the secondary as a decaying transient that can momentarily misrepresent the voltage magnitude and phase the distance relay uses to compute impedance. High-speed distance elements must therefore include CVT transient logic or accept a small intentional delay for close-in, low-voltage faults. This is a classic example of a component-level characteristic that becomes a scheme-level setting decision.

Four generations of relay technology remain in service across North American systems, and a large fraction of our relay modernization work involves migrating among them without breaking coordination with neighbors that have not migrated.

The numerical intelligent electronic device (IED) is not just a faster relay — it is the physical point where the protection system and the SCADA system merge. A single modern IED provides the protection elements, the disturbance recorder, the sequence-of-events log time-stamped to the microsecond via IRIG-B or PTP, the local metering the RTU used to require dedicated transducers for, and the communications interface that carries all of it to the control center. Designing the protection scheme and the SCADA points list as one exercise eliminates redundant hardware, wiring, and commissioning effort — and it is why we insist on a unified points list, protocol map, and network architecture drawing in every substation design package we deliver.

The self-supervision capability of numerical relays deserves particular emphasis. An electromechanical relay that fails does so silently — the failure is discovered when the relay fails to trip, which is the worst possible moment. A numerical IED continuously monitors its own power supply, memory, analog acquisition chain, and trip-coil continuity, and reports a failure to SCADA within seconds. This converts hidden protection failures from a statistical reliability risk into an ordinary maintenance work order, and it is the technical foundation for condition-based relay maintenance programs accepted under NERC PRC-005.

8. High-Impedance Faults: The Detection Problem Conventional Relaying Cannot See

Overcurrent protection remains the workhorse of distribution and industrial systems and the universal backup on transmission. Its elements are simple — a pickup threshold and a time characteristic — but coordinating dozens of them across a network so that every fault is cleared by the nearest device first, with backup always waiting behind it, is a genuinely difficult combinatorial problem.

5.1 Characteristic families and their physics

Inverse-definite-minimum-time (IDMT) characteristics operate faster as current increases, mirroring the thermal damage curves of the equipment they protect. The standard families — standard/moderately inverse, very inverse, and extremely inverse — differ in how steeply operating time falls with current. Extremely inverse curves approximate fuse and conductor damage characteristics and coordinate naturally with downstream fuses and reclosers on distribution feeders; very inverse curves suit systems where fault current falls off significantly with distance from the source; definite-time elements suit locations where fault current barely changes across the protected zone, such as systems with strong sources or short feeders. Pickup is set above maximum load with margin for cold-load pickup and transformer inrush, yet below minimum fault current with a dependable sensitivity margin; the time-dial (time multiplier) setting then positions each curve relative to its neighbors.

5.2 Coordination method

Discrimination can be achieved by current alone (where impedance between devices creates distinct fault levels), by time alone (fixed grading steps regardless of fault level), or — the standard modern practice — by both, using inverse-time curves graded with a coordination time interval (CTI). The CTI budget between a downstream and upstream device typically comprises breaker interrupting time, relay overtravel or reset allowance, CT and relay tolerance, and a safety margin; with numerical relays and modern breakers this commonly totals on the order of 0.2–0.3 seconds, versus 0.3–0.4 seconds in legacy electromechanical practice. Tightening CTI through relay modernization is often the single cheapest way to reduce arc-flash incident energy deep in an industrial system, because every upstream grading step compounds the clearing time at the worst-case bus.

5.3 Directional supervision

Wherever fault current can flow through a relay location in both directions — parallel feeders, ring buses, looped distribution, and any system with distributed generation — plain overcurrent elements cannot discriminate and must be supervised by directional elements (ANSI 67). Directional decision requires a polarizing quantity, classically the fault-loop voltage with a characteristic angle chosen for the system X/R, or negative- and zero-sequence quantities for ground directionality. The proliferation of inverter-based resources complicates this: inverters contribute limited fault current (often 1.1–1.5 per unit of rating) with unconventional sequence content, weakening both the pickup sensitivity and the polarizing signals traditional schemes rely on. Protection design for renewable-heavy networks must verify directional dependability against the actual inverter fault-current behavior documented in the manufacturer's EMT model — a direct link between interconnection modeling and protection settings that we routinely close for our clients.

A downed conductor lying on asphalt, dry sand, or gravel may draw only a few amperes to a few tens of amperes — far below feeder load, entirely invisible to overcurrent pickup, yet lethal to the public and a proven wildfire ignition mechanism. High-impedance fault (HIF) detection is therefore not an equipment-protection problem but a public-safety problem, and it demands fundamentally different signal processing.

HIF current is distinguished by signatures rather than magnitude: randomness and intermittency from arc extinction and re-strike, rich low-order and interharmonic content, cyclic asymmetry between half-cycles, and slow growth as the arc conditions the contact surface. Detection algorithms extract features — harmonic energy trajectories, wavelet coefficients, randomness metrics — and classify them against learned load behavior, trading detection probability against nuisance-alarm rate. Modern feeder relays ship with embedded HIF elements built on exactly this architecture, and utilities in wildfire-exposed territory increasingly pair them with downed-conductor logic, falling-conductor open-phase detection using synchrophasor or fast communications schemes, and REFCL/resonant-grounding strategies at the substation. Because HIF outputs are probabilistic, their disposition — alarm to SCADA for operator action versus automatic trip — is a policy decision each utility must make deliberately, and the SCADA alarm architecture must present these events with the context operators need to act within minutes.

6. Transmission Line Protection: Distance and Differential

2. Instrument Transformers: Where Every Protection Error Begins

6.1 Distance protection zones

Distance relays compute the apparent impedance from measured voltage and current; because line impedance is proportional to length, impedance is a proxy for fault position that is largely independent of source strength. The classical characteristics — plain impedance (a circle centered at the origin), reactance (a horizontal line, insensitive to arc resistance), and mho (a circle through the origin, inherently directional and load-tolerant) — remain the conceptual vocabulary, with modern numerical relays offering quadrilateral characteristics that combine reactance-like arc-resistance coverage with explicit resistive blinders for load encroachment security.

Standard zone philosophy: Zone 1 is set to underreach the remote terminal — commonly 80–85% of positive-sequence line impedance — and trips instantaneously; the deliberate underreach absorbs CT/VT error, line impedance uncertainty, and CVT transients so that Zone 1 can never overreach for a fault beyond the remote bus. Zone 2 overreaches the remote terminal (typically 120–150% of the line, checked against the shortest adjacent line) with a coordination delay on the order of 0.25–0.4 seconds, covering the last 15–20% of the line plus remote bus faults. Zone 3, where applied, provides remote backup with a longer delay and must be checked rigorously against maximum load encroachment — the misapplication of Zone 3 under heavy, depressed-voltage loading was a contributing mechanism in major blackout events and is the origin of today's NERC loadability requirements (PRC-023).

6.2 Pilot schemes and line current differential

Stepped-distance protection alone leaves the middle of every line dependent on a delayed zone for one terminal. Pilot (teleprotection) schemes close that gap by exchanging logic over a communications channel: permissive overreach transfer trip (POTT) requires both terminals' overreaching elements to agree; directional comparison blocking (DCB) trips unless the remote terminal declares the fault external; direct transfer trip (DTT) commands the remote breaker open for conditions the remote relays cannot see, such as transformer faults behind a line with no local breaker. Channel technology has migrated from power-line carrier and audio tones to SONET/SDH and MPLS-TP packet networks and dedicated fiber, which brings channel monitoring and deterministic latency into the protection engineer's scope.

Where dedicated fiber exists end-to-end, line current differential (87L) is the premier scheme: it compares actual current phasors between terminals, is inherently selective for the entire line with no reach settings to coordinate, tolerates weak infeed and series compensation better than distance elements, and is largely immune to power swings and load encroachment. Its engineering burden shifts to the communications layer — channel asymmetry and latency must be measured and compensated, and GPS/PTP time alignment or ping-pong synchronization must be engineered and monitored through SCADA.

3. From Electromechanical to Numerical: Choosing Relay Technology Deliberately

No relay setting is defensible without a fault study. The short-circuit analysis establishes, for every relay location, the maximum fault duty (which sets breaker interrupting ratings, CT dimensioning, and instantaneous element security) and the minimum fault current (which sets sensitivity requirements and proves that the relay will actually see the faults it is assigned to clear).

4.1 Balanced and unbalanced faults

The three-phase bolted fault is the classical balanced case and usually — though not always — the maximum-current condition. The overwhelming majority of real faults, however, are unbalanced: single line-to-ground faults dominate on overhead systems, followed by line-to-line and double line-to-ground events. Unbalanced conditions are analyzed with symmetrical components, which decompose any unbalanced set of three-phase phasors into positive-, negative-, and zero-sequence sets. Each sequence network has its own impedance model of the system, and each fault type corresponds to a specific interconnection of those networks: the single line-to-ground fault places all three networks in series through three times the fault resistance; the line-to-line fault connects positive and negative networks in parallel; the double line-to-ground fault parallels all three.

Sequence quantities are more than an analysis convenience — they are directly measurable operating quantities inside numerical relays. Negative-sequence current is a sensitive, load-independent indicator of unbalance used for generator rotor thermal protection (ANSI 46), for directional polarization, and for sensitive unbalanced-fault elements. Zero-sequence (residual) current exists only when ground is involved, which makes it the natural operating quantity for ground overcurrent (51N/51G) and restricted earth fault schemes. Transformer winding connections manipulate these quantities deliberately: a delta winding blocks zero-sequence current from passing through, creating a ground-source boundary, and a delta–wye transformation imposes a phase shift on positive- and negative-sequence quantities that every differential scheme across that transformer must compensate.

4.2 The DC offset and machine transients

Fault current is not a clean sinusoid. The instant of fault inception relative to the voltage wave superimposes a decaying DC component governed by the system X/R ratio, and contributions from nearby synchronous machines decay through subtransient and transient regimes as machine flux redistributes. Practical consequences: breakers must be rated for asymmetrical interrupting duty; instantaneous elements must be set with security margin against the offset peak; CTs must be dimensioned for the offset waveform; and time-overcurrent coordination must be checked at the current levels that actually persist beyond the subtransient window, not just the flashy first-cycle numbers.

Why this matters commercially

Interconnection studies, arc-flash studies, breaker duty assessments, and relay settings all consume the same short-circuit model. Building that model once, correctly, under configuration control — and keeping it synchronized with the as-built system — is one of the highest-leverage engineering investments an asset owner can make. Keentel builds and maintains these models in the owner's preferred platform and ties every relay setting file to a specific model revision.

5. Overcurrent Protection and Coordination: The Discipline of Time and Current

7.1 Transformer protection (87T and companions)

Percentage-restrained differential protection is the primary electrical scheme for power transformers above roughly 5–10 MVA. The operating principle — the vector sum of currents entering the zone should be zero — is complicated in transformers by ratio mismatch, on-load tap changer excursion, winding phase shift, and zero-sequence discontinuity, all of which numerical relays now compensate internally through settings rather than through auxiliary CT gymnastics. The restraint slope (commonly a dual-slope characteristic beginning around 20–40%) provides security against CT error that grows with through-current, with the second, steeper slope guarding against CT saturation on heavy external faults.

Magnetizing inrush is the defining security challenge: energizing a transformer draws an offset, harmonic-rich current on one side of the differential zone only, mimicking an internal fault. Inrush current is distinguished by its harmonic signature — a pronounced second-harmonic component — and by its waveform shape, so differential elements employ second-harmonic restraint or blocking (typical thresholds in the 15–20% range) and, in modern relays, waveform-gap recognition. Fifth-harmonic restraint similarly secures the element against overexcitation, which produces genuine excitation current the differential must tolerate while the volts-per-hertz element (24) takes coordinated action. Sympathetic inrush — offset flux driven in an already-energized transformer when a neighbor is switched — must also be reviewed anywhere transformers share a bus.

The electrical schemes are complemented by restricted earth fault protection (REF/87N, a sensitive zero-sequence differential covering the wye winding near the neutral where phase differential sensitivity collapses), by mechanical devices — the Buchholz gas-and-surge relay on conservator units, sudden-pressure relays on sealed units — and by thermal supervision of oil and winding temperature with staged alarm, cooling, and trip outputs. Every one of these devices is also a SCADA point: gas accumulation alarms, tap position, cooling stage status, and top-oil temperature trends feed the condition-monitoring dashboard that turns transformer protection into transformer asset management.

7.2 Generator protection

Generators accumulate the most demanding and most diverse protection portfolio of any single asset because their failure modes span electrical, thermal, and mechanical physics. The core electrical scheme is high-security phase differential (87G), backed by voltage-restrained or voltage-controlled overcurrent (51V) for system backup. Stator ground protection depends on the grounding method: high-impedance-grounded machines typically use fundamental neutral overvoltage (59N/64G) covering roughly 90–95% of the winding, extended to 100% coverage by third-harmonic differential methods or subharmonic injection so that faults near the neutral — invisible to fundamental schemes — are still detected. Rotor field grounds are detected by injection-based 64F schemes; the first ground is an alarm, but a second ground short-circuits part of the field, unbalances rotor flux and forces, and can be catastrophic, so tripping philosophy for the first ground must be an explicit owner decision.

The abnormal-operating-condition suite is equally important: loss-of-excitation (40, impedance elements watching the machine slide toward the steady-state stability limit), negative-sequence thermal duty (46, protecting the rotor surface against unbalance-induced heating per the machine's I2²t withstand), reverse power (32, anti-motoring for prime-mover protection), over/underfrequency (81O/U), volts-per-hertz overexcitation (24), inadvertent energization logic (50/27), and out-of-step (78). For synchronous machines connected to networks now governed by ride-through standards, every one of these elements must be checked against the required voltage and frequency ride-through envelopes so that protection does not trip the machine for conditions the interconnection agreement obligates it to survive — a coordination task between protection settings and NERC PRC-024/PRC-029-family compliance that we perform as a single integrated study.

7.3 Motor and busbar protection

Large motor protection is thermal-model protection first: the relay integrates a thermal replica (49) fed by stator current, tracks starting duty against the machine's hot and cold safe-stall limits, counts starts per hour, and supervises locked-rotor conditions (48/51LR) where the distinction between a prolonged start and a stalled rotor may require a speed switch. Instantaneous phase (50) and sensitive ground (50G on a core-balance CT) elements handle faults; undervoltage (27) and phase reversal/unbalance (46/47) elements handle supply-quality events that convert into rotor heating.

Busbar faults are rare but carry the highest fault energy and the widest outage footprint in the station, so bus protection prioritizes speed with extreme security. Low-impedance percentage differential schemes with per-terminal CT saturation detection are today's standard for complex, reconfigurable buses, with dynamic zone assignment following disconnector position; high-impedance differential remains a superbly secure and simple choice where CT ratios are uniform and dedicated cores are available. Breaker-failure protection (50BF) completes the station scheme: if a tripped breaker fails to interrupt within its timer window, 50BF trips the surrounding zone — and its correct, selective operation depends on accurate breaker status and initiate signals that are themselves part of the station's wiring-and-communications design.

9. System Grounding and the Ground Grid: The Silent Half of Protection Design

7. Equipment Protection: Transformers, Generators, Motors, and Buses

This content package presents Keentel Engineering's integrated perspective on power system protection and relaying design in the era of the SCADA-integrated, digitally instrumented grid. It is organized in three parts: a long-form technical blog covering the complete protection design chain — from protection philosophy and instrument transformer specification through relay technology selection, fault analysis, overcurrent and distance coordination, equipment-specific protection schemes, high-impedance fault detection, grounding, and substation SCADA/communications architecture; a practitioner-oriented FAQ addressing the questions asset owners, developers, and plant engineers most frequently raise; and three anonymized case studies demonstrating how this integrated approach performs on real project archetypes: a utility-scale solar-plus-storage interconnection, an industrial relay modernization, and a distribution utility's wildfire-driven ground-fault protection program.

The unifying thesis: protection design decisions are coupled — to the short-circuit model, to the grounding design, to the communications and SCADA architecture, to interconnection ride-through obligations, and to the NERC compliance program that must defend all of it. Engineering these as one configuration-controlled scope, rather than fragmented procurements, is the difference between a system that merely has relays and a system that is demonstrably protected, observable, and audit-ready.

4. Fault Analysis: The Quantitative Foundation

10.1 The four layers

A substation SCADA architecture resolves into four layers. Field instrumentation — the CTs, VTs, transducers, temperature probes, gas monitors, and breaker auxiliary contacts that originate every measurement and status. Data concentration and control — RTUs, PLCs, and increasingly the protection IEDs themselves, which acquire, time-stamp, and execute control.

Communications — the serial and Ethernet networks, protocols, and wide-area links that move data with the required latency, integrity, and availability. And the master layer — SCADA host/HMI, historian, alarm management, and the EMS/ADMS applications that convert telemetry into operating decisions. The historical distinction between the 'dumb telemetry' RTU and the logic-bearing PLC has effectively dissolved; modern platforms do both, and in a protection-centric substation the IEDs carry most of the acquisition burden, with the RTU or data concentrator serving as protocol gateway and point-of-demarcation to the control center.

10.2 Protocols and the IEC 61850 substation

DNP3 remains the dominant master-station protocol in North America, prized for its report-by-exception efficiency, time-stamped event buffering, and — in secure-authentication variants — cryptographic protection of control operations. Inside the substation fence, IEC 61850 changes the design language itself: devices publish self-describing logical nodes; GOOSE messaging replaces hardwired inter-relay trip, block, and interlock wiring with sub-4-millisecond peer-to-peer Ethernet messages; and Sampled Values streams from merging units allow process-bus architectures in which copper CT/VT wiring to every relay is replaced by fiber. The engineering payoff is enormous — interlocking and breaker-failure initiate schemes become configuration rather than construction — but it moves network engineering (VLANs, redundancy protocols such as PRP/HSR, time synchronization via PTP) squarely into the protection engineer's deliverables, and it makes the SCD configuration file a controlled engineering document on par with the relay settings files.

10.3 What SCADA integration returns to protection

Event analysis — Every relay operation arrives with a synchronized sequence-of-events record and oscillography, collapsing post-fault analysis from days of site visits to minutes at a desk.
Asset health — Relay self-supervision alarms, trip-circuit monitoring, breaker operation counters, and DC system telemetry convert hidden failures into scheduled work — the backbone of a PRC-005-compliant, condition-based maintenance program.
Operational flexibility — Settings group changes can be supervised remotely for planned abnormal topologies; adaptive schemes can respond to DER output, seasonal load, or storm posture.
Restoration speed — Fault-location estimates from relays, delivered through SCADA, dispatch crews to the right structure instead of the right circuit.
Compliance — Disturbance records and misoperation data organize evidence for NERC PRC-004 mis operation review and PRC-002 disturbance-monitoring obligations.

10.4 Cybersecurity as a protection-reliability requirement

Every communications path that makes protection observable also makes it attackable. A SCADA-integrated protection design must therefore carry a security architecture: network segmentation with an explicit electronic security perimeter, role-based access and centralized authentication for relay engineering access, secure protocol variants for control, logging and monitoring of engineering-access sessions, firmware and settings configuration management, and disciplined handling of transient devices such as test laptops. For BES facilities these are NERC CIP obligations with audit consequences; for every facility they are simply sound engineering. Our design packages treat the security architecture drawing set with the same rigor as the AC/DC schematics — because a protection scheme whose settings can be silently altered is not a protection scheme.

10. The SCADA Architecture: Turning Protection Into an Observable, Operable System

The system-neutral grounding method determines the entire character of ground-fault protection. Solidly grounded systems deliver high ground-fault current — easy to detect, harsh on equipment, and demanding on arc-flash energy. Low-resistance grounding (commonly a few hundred amperes) caps damage at the fault point while preserving straightforward selective relaying, and is the default for industrial medium-voltage systems. High-resistance grounding limits the first ground fault to a few amperes, permitting continued operation with alarm — invaluable for continuous processes — at the cost of pinpointing-and-locating discipline and careful insulation coordination for the sustained phase-to-phase voltage on the unfaulted phases. Ungrounded and resonant-grounded (Petersen coil) systems suppress fault current further still and, in the resonant case, actively extinguish arcing ground faults, which is why compensated grounding is central to several utilities' wildfire-mitigation programs; both demand sensitive wattmetric or admittance-based ground relaying because there is almost no zero-sequence current to measure.

Beneath every substation, the ground grid design closes the personnel-safety loop: grid resistance, ground potential rise, and touch and step voltages during the maximum ground fault must satisfy IEEE Std 80 tolerable-body-current criteria, which depend on soil resistivity structure measured in the field (Wenner four-pin traverses resolved into multi-layer soil models), fault-current split between the grid and overhead ground wires or cable sheaths, and fault-clearing time — which is set by the protection engineer. This is a genuinely coupled design problem: faster protection directly reduces the tolerable-voltage burden on the grid, and grid design assumptions must be revisited whenever a protection upgrade changes clearing times or system changes raise fault duty. Keentel performs grounding studies (including WinIGS-based grid modeling), protection studies and short-circuit studies as one coordinated package for precisely this reason.

11. The Keentel Perspective: Protection, SCADA, and Interconnection as One Design

The thread running through every section above is that protection design decisions are coupled — to the fault study, to the grounding design, to the communications architecture, to interconnection ride-through obligations, and to the compliance program that must defend all of it in an audit. Treating these as separate scopes procured from separate vendors is how systems accumulate miscoordination, hidden failure modes, and compliance findings. Keentel Engineering's practice is built on the opposite premise: the short-circuit model, the protection and coordination study, the relay settings and logic, the grounding study, the SCADA points list and network design, and the NERC compliance evidence are one integrated engineering product, developed under one configuration-control discipline, from the 30% design milestone through commissioning and into operations support.

Whether the asset is a transmission substation, a utility-scale solar or BESS plant working through its interconnection requirements, a data center campus negotiating large-load protection requirements, or an industrial system overdue for relay modernization, the design questions in this article are the ones we answer every day — quantitatively, with studies and settings files, not with generalities.

Case Studies

Frequently Asked Questions: Power System Protection, Relaying, and SCADA Integration

The following case studies are drawn from representative Keentel Engineering project experience. All client names, project names, locations, and identifying details have been anonymized or generalized; technical parameters have been rounded or adjusted where necessary to preserve confidentiality while retaining engineering fidelity.

Case Study 1: Protection and SCADA Integration for a Utility-Scale Solar-Plus-Storage Collector Substation

Background

An independent power producer was developing a utility-scale photovoltaic facility of approximately 200 MW, co-located with a four-hour battery energy storage system, interconnecting to a transmission utility at 230 kV through a new collector substation. The interconnection agreement imposed the transmission owner's full protection, telemetry, and disturbance-monitoring requirements at the point of interconnection, including redundant line protection on the gen-tie, remote terminal upgrades at the utility's end, and ride-through obligations that constrained every protective setting in the plant. The EPC schedule allowed no float between substation energization and back-feed for commissioning of the inverter blocks.

The Challenge

The gen-tie line was electrically short, making conventional stepped-distance Zone 1 reach margins unacceptable — CT/VT error and line-impedance uncertainty consumed most of the underreach allowance.
The plant's fault contribution was almost entirely inverter-based: roughly 1.2 per unit of aggregate inverter rating, with manufacturer-controlled negative-sequence behavior. Utility-side protection assumptions about infeed, and plant-side directional and sensitivity checks, both had to be validated against the certified EMT model rather than classical machine assumptions.
The transmission owner required protection settings that were demonstrably coordinated with mandated frequency and voltage ride-through envelopes — any plant protection element capable of tripping inside the ride-through region required written technical justification.
The owner's operations group required full visibility of the collector system — feeder relays, transformer protection, BESS interfaces, meteorological and plant-controller data — in both their own operations center and the transmission operator's EMS, over different protocols.

Keentel's Approach

Keentel served as protection and SCADA engineer of record for the collector substation and gen-tie, working from the same EMT and short-circuit models used in the interconnection studies — eliminating the model handoff gap that typically separates interconnection compliance from protection design.

Specified redundant line current differential (87L) over diverse fiber paths as primary gen-tie protection, with quadrilateral distance and directional ground overcurrent backup; POTT logic over the secondary channel provided a communications-degraded fallback. Reach and timing settings were validated against inverter fault-current signatures from the EMT model, not classical source assumptions.
Designed the 230/34.5 kV main power transformer protection with dual-slope percentage differential, second- and fifth-harmonic security, restricted earth fault on the wye winding, and sudden-pressure and thermal devices fully mapped into SCADA.
Engineered collector feeder protection with sensitivity checks at minimum inverter output, verifying that end-of-feeder faults remained detectable when the plant's fault contribution was at its floor — a check that failed under the vendor's template settings and required pickup revisions on three feeders.
Produced a protection-versus-ride-through coordination document mapping every plant protective element against the required voltage and frequency withstand envelopes, providing the transmission owner a single auditable artifact and accelerating their settings approval cycle.
Implemented IEC 61850 inside the substation — GOOSE for breaker failure initiate, transformer lockout distribution, and feeder interlocking on a PRP-redundant LAN — with a station gateway presenting DNP3 to both the owner's SCADA and the transmission operator's EMS from a single points database, and PTP-disciplined time synchronization for sequence-of-events alignment.

Results

Key takeaway

On inverter-based plants, protection design inherits its critical assumptions from the interconnection EMT model. Running interconnection modeling, protection settings, and SCADA architecture as one scope removed an entire class of late-stage rework — and turned the transmission owner's settings review from a schedule risk into a formality.

Case Study 2: Relay Modernization and Coordination Recovery at a Continuous-Process Industrial Facility

Background

A continuous-process manufacturing facility operating a medium-voltage distribution system — dual utility services at 69 kV, two main transformers feeding a 13.8 kV main-tie-main arrangement, and roughly forty downstream feeders, MV motors, and unit substations — had experienced two plant-wide outages in eighteen months from faults that should have been isolated locally. The protection fleet was predominantly electromechanical, some of it four decades old, with a settings record consisting of scanned test sheets of varying vintage. A parallel concern was arc-flash: the most recent incident-energy study had flagged several 13.8 kV and 480 V buses at hazard levels that constrained routine live work.

The Challenge

Forensic review of the two outages showed classic miscoordination: a feeder fault cleared by the main rather than the feeder relay in one event, and in the other, a failed electromechanical relay (a hidden failure with no self-supervision) allowed a transformer secondary fault to escalate to the utility's upstream protection.
The as-found system had drifted from every documented study: pickup settings had been raised over the years to ride through motor-starting trips, quietly consuming coordination margin; two added feeders had never entered any study at all.
Production constraints allowed only brief, scheduled bus outages — the modernization had to be sequenced around the plant's turnaround calendar, and the mixed old/new fleet had to remain coordinated at every intermediate stage.
The facility's engineering staff needed the end state to be maintainable in-house: a settings database of record, monitoring of relay health, and event records that did not require a consultant visit to interpret.

Keentel's Approach

Rebuilt the facility short-circuit and coordination model from field-verified data — nameplates, cable schedules, CT ratios confirmed by inspection — rather than inheriting the legacy model's assumptions; utility source impedances were refreshed from the serving utility's current published values.
Executed a staged migration to numerical multifunction relays across the mains, ties, transformer, and critical feeder positions, with interim coordination studies issued for each construction stage so the plant was never operating on undocumented settings.
Re-graded the entire time–current landscape with modern CTI budgets (0.2–0.25 s between numerical devices), applied instantaneous elements with proper security checks where the legacy design had omitted them, and introduced maintenance-mode arc-flash settings groups — selectable temporary sensitive/fast settings for periods of energized work — on all main and tie positions.
Addressed the motor-starting nuisance-trip history properly: thermal-model motor protection with accurate hot/cold safe-stall parameters replaced the raised-pickup workarounds, restoring both motor protection integrity and upstream coordination margin simultaneously.
Deployed a plant power-monitoring SCADA layer over the relay fleet: relay self-supervision and trip-circuit monitor alarms, breaker operation counters, event and oscillography retrieval to a central historian, and a settings database of record with as-left verification procedures — turned over with training to plant staff.

Results

Key takeaway

The facility's outages were not caused by old relays; they were caused by decades of uncontrolled settings drift and hidden failures that electromechanical technology cannot report. Modernization delivered its value less through faster elements than through restored coordination discipline, self-supervision, and a settings management system the owner can actually maintain.

Case Study 3: Ground-Fault Protection Strategy and Wildfire-Risk Mitigation for a Distribution Utility

Background

A distribution utility serving a mixed suburban and high-fire-risk rural territory initiated a program to improve detection of downed conductors and high-impedance ground faults after an internal review found that a broken conductor on dry soil could persist undetected on several of its feeders. The utility operated a conventional four-wire multigrounded system with legacy feeder relaying set on phase and residual overcurrent alone. Concurrently, several substations were due for ground-grid reassessment: fault duties had grown with system reinforcement, and the original grids had been designed decades earlier to superseded assumptions.

The Challenge

Measured and simulated high-impedance fault currents on the utility's soil types fell in the range of a few amperes to a few tens of amperes — below feeder residual pickups by an order of magnitude, and below load imbalance on several circuits, making conventional threshold-based detection impossible.
Any increase in ground-fault sensitivity risked nuisance operations from load unbalance, capacitor switching, and single-phase recloser events — on rural feeders where each unnecessary sustained interruption carries real customer impact.
The trip-versus-alarm policy for probabilistic HIF detection had never been formalized; operations, engineering, and the utility's risk function held different implicit assumptions.
Ground-grid compliance at three substations was uncertain under present fault duties and clearing times, and the protection upgrade itself would change those clearing times — coupling the two scopes.

Keentel's Approach

Deployed modern feeder relays with embedded high-impedance arc-detection elements on the prioritized fire-risk feeders, tuned through a supervised learning period in alarm-only mode against each feeder's recorded load signature before any tripping was enabled.
Layered the detection strategy: sensitive wattmetric/directional ground elements for the detectable band, HIF signature elements for the arcing band, and a fast open-phase/falling-conductor scheme — using loss-of-voltage and current-signature logic at reclosers with peer-to-peer communications — aimed at de-energizing a breaking conductor before ground contact on the highest-risk segments.
Facilitated a structured trip/alarm policy workshop with operations, protection engineering, and risk management, producing a documented decision matrix: automatic tripping on confirmed HIF plus fire-weather condition flags, operator-confirmed action otherwise, with seasonal settings groups switched through SCADA on declared fire-risk days.
Performed soil resistivity surveys (Wenner traverses resolved to two-layer models) and full grid analyses at the three flagged substations, evaluating touch and step voltages under present maximum ground-fault duty and the new protection clearing times; two grids passed with the faster clearing credited, one required a targeted perimeter enhancement rather than the full rebuild the utility had budgeted.
Integrated all new elements into the utility's DNP3 SCADA with dedicated HIF alarm classes, fault-location context for dispatch, and settings-group status telemetry, plus an event-review workflow so every HIF alarm — actionable or not — fed back into detector tuning.

Results

Key takeaway

High-impedance fault protection is a system design problem — detection algorithms, layered schemes, operating policy, SCADA presentation, and grounding analysis together — not a relay feature to be switched on. Treating the clearing time as a shared variable between the protection study and the ground-grid study converted a coupled compliance risk into a program-funding opportunity.

Disclaimer

This document is published by Keentel Engineering for general informational and educational purposes. It does not constitute engineering services, and no engineering decisions should be made on the basis of this document without a project-specific study performed by a licensed professional engineer. Standards, regulatory requirements, and grid codes referenced herein (including IEEE, IEC, ANSI, and NERC materials) evolve; readers must consult the current editions applicable to their jurisdiction and facility.

All product names, standards designations, organization names, and trademarks referenced in this document are the property of their respective owners. Keentel Engineering is not affiliated with, endorsed by, or sponsored by any standards body, software vendor, equipment manufacturer, utility, or other organization referenced herein. References are made solely for identification and educational purposes.

SPP HILL Process Guide for AI Data Centers & Large Loads

Thu, 09 Jul 2026 18:32:06 GMT

Jul 09, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

SPP's High Impact Large Load (HILL) Process Explained

AI data centers, crypto mining facilities, hydrogen electrolyzers, and large industrial plants are connecting to the grid at a scale the power industry has never seen. A single AI campus can now demand more power than a mid-size city — and unlike a city, it can appear on the system in 18 months and change its consumption by hundreds of megawatts in seconds.

The Southwest Power Pool (SPP) responded with one of the most comprehensive large-load interconnection frameworks in North America: the High Impact Large Load (HILL) process, established through Revision Request 696 (effective January 15, 2026) and unanimously approved by FERC in January 2026. If you are developing a data center or any large load in the SPP footprint — Kansas, Oklahoma, Nebraska, the Dakotas, and beyond — this process now governs how your project gets studied, what your equipment must be capable of, and what models you must deliver before SPP will run a single simulation.

This guide walks through the entire process step by step, explains the engineering behind each requirement, and answers the technical questions we hear most often from developers, owner's engineers, and EPCs.

Part 1: What Is a HILL?

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

By Keentel Engineering — Power System Studies | EMT & Dynamic Modeling | Interconnection Support

Part 2: Choosing Your Pathway Attachment AQ, Attachment AX, or CHILL

A HILL is a classification of load defined in the SPP Open Access Transmission Tariff (OATT). Your project is a HILL if it is a new commercial or industrial load — or an increase in load at a single site that meets either threshold:

10 MW or more connected at a voltage of 69 kV or below, or
50 MW or more connected above 69 kV.

Electric storage resources are excluded from the definition. Virtually every hyperscale or co-location data center, and most AI training campuses, will exceed these thresholds by an order of magnitude.

Why the special treatment?

Because large power-electronic loads behave nothing like the aggregate residential and commercial load the grid was planned around. A 245 MW data center is, electrically, a massive rectifier plant: UPS front ends, server power supplies, and variable-speed cooling drives. During a transmission fault, that entire load can disconnect in a few cycles — and either stay off or slam back on. Uncontrolled tripping of gigawatts of load during a voltage sag can turn a routine fault into a regional frequency and voltage event. SPP's HILL framework exists to make sure that doesn't happen.

Before any study begins, the Transmission Customer (typically the load-serving entity or network customer sponsoring your project) must choose a study pathway. As of RR720 (effective July 1, 2026), customers with sufficient Designated Resources have a choice.

Attachment AQ — the Delivery Point Assessment path.

This is the traditional route, used when the Transmission Customer already has enough Designated Resources (contracted generation) to serve the new load. The HILL is studied through a HILL Delivery Point Study (HDPS) alongside the host Transmission Owner's Load Connection Study.

Attachment AX — the Provisional Load Process.

This newer path addresses the reality that many large loads want service faster than new generation can clear the traditional interconnection queue. Under AX, the HDPS incorporates planned generation identified to serve the load, including a Generator Outlet Facility (GOF) analysis. The load receives provisional service tied to that generation; once the resource obtains firm service through the Aggregate Transmission Service Study, both move to a standard Network Integration Transmission Service Agreement (NITSA). Network upgrade costs are directly assigned to the customer during the provisional period, then migrate to Base Plan funding.

CHILL — Conditional HILL service.

For projects with the most aggressive energization timelines, SPP offers conditional access: quick study results and interconnection in exchange for accepting potential curtailment during periods of system stress.

Co-located generation: HILLGA and HILLGIA.

If your project pairs on-site or near-site generation with the load behind a common point of interconnection (POI) — increasingly the default for AI campuses — SPP's HILL Generation Assessment (HILLGA) process provides an interconnection agreement (HILLGIA) path, with requirements including a POI no more than two substations from the HILL and commercial operation within five years of the study agreement.

How Keentel helps:

Pathway selection is a commercial decision with deep technical consequences. Keentel Engineering supports developers and owner's engineers in evaluating AQ vs. AX vs. CHILL against Designated Resource positions, generation development timelines, and curtailment risk tolerance — before the application is filed, when the decision is still cheap to change.

There is no cluster window or filing deadline. HILL requests are processed on a rolling basis, which means the real "deadline" is self-imposed: the 90-day study clock does not start until your package is complete and validated. An incomplete submittal doesn't get rejected — it just sits, burning schedule.

For an Attachment AQ request, the submittal (sent to SPP's load studies group) must include:

A fully completed Delivery Point Assessment (DPA) Request Form (Addendum 1 to Attachment AQ).
A fully executed HDPS Agreement (template on SPP's OASIS studies page).
A ten-year load forecast with summer, winter, and light-load values for the delivery point, plus any associated changes at other delivery points.
A one-line diagram showing the anticipated load and changes to local delivery facilities.
Associated IDEV files (SPP's power-flow model change files).
The completed Additional HILL Characteristics Form (Business Practice 7850) — the detailed load-behavior questionnaire introduced by RR724.
Load modeling data — a CMLD and/or PERC1 model, with PERC1 preferred by SPP.
Any useful supplemental information — and SPP explicitly recommends including a PSCAD model at the time of submission (more on why below).

The Attachment AX list is similar but adds details of the planned generation — location, capacity, and the associated IDEV and DYR (dynamics) files. Under AX, the customer has 30 calendar days after submitting the request form to deliver the executed HDPS Agreement and study deposit.

The item that trips up most projects is #7.

Ten-year forecasts and one-lines are familiar territory. A validated PERC1 or CMLD parameter set that actually represents your UPS topology, cooling plant, and IT load ramp behavior is specialized power-systems work — and it is now a gating item for the study clock.

Once the package is complete:

Step 1 — Scoping call (within 10 days).

SPP, the Transmission Customer, and the host Transmission Owner meet to validate the request data, agree on study scope and assumptions, and set the study due date. Data validation at this call is what officially starts the 90-day clock.

Step 2 — The 90-day base study stage.

SPP performs the HDPS while the host TO simultaneously performs its Load Connection Study, and affected-system impacts are screened. The base stage comprises five analyses:

Thermal overload study. Power-flow analysis under specified contingencies and operating conditions, comparing pre- and post-project loading on transmission equipment to identify overloads and required mitigations.
Voltage stability study (steady state). Power-flow analysis across dispatch scenarios and contingencies to confirm the load does not cause voltage violations or non-convergent (voltage-collapse) conditions, identifying reactive support needs.
Short circuit study. Fault-duty analysis per NERC TPL-001-5.1, comparing pre- and post-project fault levels to determine whether breaker or equipment mitigation is needed.
RMS dynamic performance study. Transient stability analysis using SPP's network model and your RMS load model (this is where your CMLD/PERC1 gets exercised). SPP evaluates frequency response, voltage recovery, transient stability, load tripping and reconnection, ramping, energization, and fault ride-through against the SPP Disturbance Performance Requirements and NERC TPL-001-5.1.
EMT screening — the SCRCCT gate. SPP screens the POI's system strength using Short Circuit Ratio (SCR), Weighted SCR (WSCR), Composite SCR (CSCR), and Critical Clearing Time (CCT). The thresholds: SCR/WSCR/CSCR ≥ 6.0 and CCT ≥ 0.15 seconds. Fall below either, and your project proceeds to the supplemental stage.

Step 3 — The supplemental study stage (if screening fails).

Here is the schedule trap: supplemental studies are performed outside the 90-day window, and they require a detailed EMT (PSCAD) model of your facility. If you didn't submit one with the application, you now build one while your project waits. This is precisely why SPP recommends — and we insist — that the PSCAD model be part of the original package.

The supplemental stage can include:

EMT dynamic performance study — detailed electromagnetic-transient simulation of the facility against the transmission network.
Sub-synchronous oscillation (SSO) screening and detailed studies — impedance frequency scans at the POI per CIGRE Technical Brochure 909, under contingencies up to N-5, checking for sub-synchronous resonance (SSR) and sub-synchronous control interactions (SSCI) between your converters and the network.
Converter-driven stability studies — screening via SCR and weighted Multi-Infeed Interaction Factor (wMIIF), with detailed EMT analysis of control interactions between your facility, nearby converters, and third-party equipment.
Emergency power control (EPC) study — verification that the facility's emergency power controls and ramp-rate capability perform as designed, tested against a Thévenin network equivalent.
Active/reactive power capability and controller studies — confirmation of P/Q capability at the POI and step-response verification of the facility's power controllers.
Power quality study — harmonic compliance, energization voltage-step limits, and operation under unbalanced conditions.
Fault ride-through EMT study — direct demonstration of the facility riding through the required voltage/frequency envelopes.
Network model verification and HILL model verification — SPP checks its own RMS-vs-EMT network consistency, and critically, checks that your RMS and EMT models agree with each other.

Step 4 — Study report and next steps.

SPP delivers the HDPS report within the 90 days (base stage). Results are valid for one year, and the Transmission Customer has one year from posting to notify SPP of intent to add the load to its NITSA (or Provisional Service Agreement under AX).

Part 5: The Fault Ride Through Requirements What Your Facility Must Actually Do

How Keentel Engineering Supports HILL Projects

Three models sit at the heart of every HILL submittal, and SPP will formally verify that they agree with each other.

CMLD (Composite Load Model – Dynamic).

The WECC composite load model (CMPLDW) represents the facility as a mix of motor types, electronic load, and static load behind a feeder/transformer equivalent, with protection-based tripping and reconnection. For a data center, the parameterization must capture the electronic (UPS-served) fraction, the VSD-driven cooling fraction, any direct-connected motors, and the trip/reconnect thresholds that mirror the FRT envelope.

PERC1 — SPP's preferred model.

PERC1 is the newer performance-based electronic load model developed through industry load-modeling efforts specifically for power-electronic-dominant facilities. Rather than approximating a data center as a motor mix, PERC1 directly represents the behaviors SPP's FRT requirements are written around: proportional load reduction during sags, constant-current disturbance behavior, deep-sag transfer/trip logic, and the timed recovery ramp. SPP's submittal requirements list load modeling data as "CMLD and/or PERC1 (preferred)" — submitting a well-parameterized PERC1 signals to SPP's study engineers that your team knows the current framework.

PSCAD (EMT).

The electromagnetic-transient model represents the facility at the converter level: UPS rectifier front-ends with their ride-through, transfer, and recovery controls; VSD drive stages; transformers and the collector system; and facility protection logic. Because the supplemental studies exercise this model against Thévenin equivalents, impedance scans, unbalanced faults, energization events, and multi-reclose sequences, the model must be built to a much higher standard than "it runs": stable initialization, flat start, numerical robustness, and validated ride-through behavior.

Model verification is not optional.

SPP's HILL Model Verification study compares the responses of your RMS and EMT models to confirm consistency. If your CMLD/PERC1 and PSCAD models were developed by different parties at different times from different data, discrepancies surface during SPP's study — the most expensive possible moment. The models should be developed as a matched pair and benchmarked before submittal.

A note on PNNL reference models.

Pacific Northwest National Laboratory has published excellent reference models and parameterizations for data center loads, and many projects rightly start there. But reference models represent a generic facility. SPP's process — particularly the FRT validation and model verification steps — requires parameters traceable to your equipment: your UPS topology and transfer settings, your chiller and DLC plant, your block-load ramp rates, your protection settings. Refinement from PNNL baselines to site-specific, SPP-format models is exactly the specialized scope most project teams need help with.

How Keentel helps:

Keentel Engineering develops CMLD and PERC1 parameter sets and PSCAD facility models as a matched, mutually benchmarked pair — validated against the full SPP FRT envelope (voltage sag tables, IEEE 2800 TOV cumulative-duration logic, frequency profiles, six-reclose sequences, and one-second recovery) and packaged in SPP-submittal formats, including IDEV-compatible records and Business Practice 7850 documentation. We build every PSCAD model supplemental-stage-ready, so a failed SCRCCT screen costs your project analysis time, not model-development time.

Keentel Engineering provides end-to-end dynamic modeling and study support for large load interconnections in SPP and across North American markets:

CMLD and PERC1 dynamic load model development — site-specific parameterization from your equipment data, benchmarked against reference models (including PNNL baselines), encoded to the SPP FRT envelope, and delivered in SPP submittal formats with Business Practice 7850 documentation.
PSCAD EMT model development and validation — supplemental-stage-ready facility models covering UPS front-ends, VSD cooling plant, protection and reconnection logic; validated against the full SPP ride-through requirements including reclose sequences, TOV cumulative-duration logic, and recovery performance.
RMS-vs-EMT model verification — matched-pair benchmarking that pre-satisfies SPP's HILL Model Verification study.
FRT gap assessments and VSD exemption studies — equipment-versus-requirement gap analysis before model build, and HDPS-basis technical justification for VSD exemptions.
Submittal package support — Additional HILL Characteristics Form, IDEV files, and coordination of the load forecast data set.
Study-cycle coordination — scoping call support, SPP/TO comment response, model resubmittals, and on-call engineering through base and supplemental study stages.
Pathway and strategy advisory — AQ vs. AX vs. CHILL evaluation, co-located generation (HILLGA/HILLGIA) coordination, and weak-grid mitigation planning.

Every engagement is performed under the responsible charge of a licensed Professional Engineer, with mobilization typically within two business days of notice to proceed.

Ready to start — or stuck in a comment cycle?

Contact Keentel Engineering at contact@keentelengineering.com or visit keentelengineering.com to schedule a technical consultation. Bring your one-line and your UPS datasheets; we'll bring the roadmap.

Keentel Engineering — Power System Studies | EMT & Dynamic Modeling | Interconnection Support

This article reflects the SPP HILL framework as of mid-2026, including RR696, RR720, and RR724, the HDPS Overview, and SPP's HILL Fault Ride-Through Requirements (V2.0). SPP's requirements continue to evolve through the stakeholder process; always confirm current requirements against SPP's OASIS studies page and governing documents before submittal.

Part 7: A Realistic Timeline

Part 3: The Application Package What You Must Submit

For a well-prepared project, the sequence looks like this:

The two levers a developer actually controls are at the top of the table: how fast the package gets complete, and whether the PSCAD model is in it. Everything after that runs on SPP's clock.

Part 4: The HDPS Step by Step

SPP's HILL Fault Ride-Through Requirements are where the framework moves from paperwork to physics. These are binding performance expectations, evaluated through modeling, commissioning tests, and operational monitoring. The highlights:

Voltage ride-through (measured at the POI).

The facility must operate continuously between 0.90 and 1.10 per-unit voltage. It must ride through 0.80–0.90 pu for at least 2.0 seconds, 0.50–0.80 pu for 0.5 seconds, voltages below 0.50 pu for 0.15 seconds, and overvoltage of 1.10–1.20 pu for 0.5 seconds. The envelope is aligned with the ITIC (formerly CBEMA) curve and ERCOT's large electronic load proposal.

Behavior during sags — not just survival.

During shallow sags (0.5–0.8 pu), the load must keep drawing power from the grid, with load reduction proportional to the voltage dip. For deep sags below 0.5 pu, the load may transfer to UPS or trip — but it must return to at least 90% of pre-disturbance consumption within one second of voltage recovering to 0.9 pu. In weak-grid conditions, that recovery time can be extended in consultation with SPP and the TO to avoid post-fault oscillations.

Constant current, not constant power.

Power-electronic loads must use constant-current control during disturbances. Constant-power control is prohibited — because a constant-power load increases its current draw as voltage falls, which is exactly the behavior that deepens sags and drives voltage collapse.

Reclosing ride-through.

Transmission faults are often cleared by breakers that automatically reclose — sometimes into a still-faulted line. Your facility must withstand up to six voltage fault-clearing attempts within a 90-second period before protection-driven disconnection is acceptable.

Transient overvoltage and frequency.

Instantaneous overvoltage withstand follows IEEE 2800-2022 Table 14 (cumulative durations over a one-minute window — e.g., 3 ms above 1.40 pu, 15 ms above 1.20 pu). Frequency ride-through follows a PRC-029-1-based profile: continuous operation from 58.8 to 61.2 Hz, and 299 seconds of ride-through out to 57.0 and 61.8 Hz.

The VSD exemption opportunity.

The requirements include a pragmatic carve-out: variable-speed drives (a large fraction of any data center's cooling plant) may be exempted from the constant-current requirement, the deep-sag minimum duration, and the transient overvoltage requirements — but only if the HDPS demonstrates the exemption will not adversely affect SPP grid reliability. That demonstration is a modeling exercise. Done well, it can meaningfully reduce your equipment specification and compliance cost.

Co-located generation has its own rules.

Inverter-based resources behind the same POI must meet IEEE 2800-2022 Clause 7 (with SPP-specified decision points, including reactive-current priority during faults); synchronous machines follow NERC PRC-024-4.

How Keentel helps:

Keentel performs facility-level FRT gap assessments — comparing your actual UPS transfer settings, VSD behavior, and protection philosophy against the SPP tables before models are built — and prepares the technical justification for VSD exemptions where the physics supports them. Finding a ride-through gap in Week 1 costs a design memo; finding it in the SPP study report costs a restudy.

Part 6: The Models CMLD, PERC1, and PSCAD

Submitting without load models and assuming SPP will "work with you." The clock simply doesn't start.
Submitting CMLD only, when SPP prefers PERC1. At best a data-validation comment cycle; at worst, a model that can't represent the required disturbance behaviors.
Skipping the PSCAD model to save budget. If SCRCCT screening fails at a weak POI — common in load pockets — the project waits months while a model is built.
Developing RMS and EMT models independently. SPP's model verification study will find the inconsistencies for you.
Modeling a facility the equipment can't deliver. If the UPS transfer settings or VSD ride-through don't actually meet the FRT tables, a beautiful model just documents non-compliance.
Ignoring the VSD exemption pathway. Teams over-specify equipment to meet requirements the tariff would have exempted them from — if anyone had run the demonstration.
Treating the Additional HILL Characteristics Form as paperwork. The form's answers must match the models. Inconsistencies between the form, the forecast, and the model parameters are the fastest route to a deficiency notice.

Frequently Asked Questions

Part 8: The Seven Most Common Mistakes We See

EHV GIS Substation Interlocking Guide | One-and-a-Half Breaker

Thu, 09 Jul 2026 15:33:54 GMT

Jul 09, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Interlocking Engineering for EHV GIS Substations

In an extra-high-voltage gas-insulated substation, the interlocking scheme is the last line of defense between a routine switching order and a catastrophic event. A disconnector opened under load, a breaker closed onto a grounded bus, or an earth switch applied to an energized section can destroy primary equipment, collapse a diameter, and put personnel at severe risk. Interlocking logic exists to make those events mechanically and electrically impossible — not merely procedurally discouraged.

At Keentel Engineering, we treat interlocking design and verification the same way we treat grid interconnection: as a first-order engineering input that must be resolved early, not a documentation task cleaned up at the end of a project. The interlocking philosophy shapes the bay layout, the control architecture, the SCADA point list, the maintenance procedures, and ultimately the operational flexibility of the substation for its entire service life.

This article walks through interlocking engineering for a 380–400 kV class GIS diameter arranged in the one-and-a-half breaker scheme. We cover the roles of the bus-side, middle, and line-side circuit breakers; the isolators and earth switches that surround them; the distinction between the service position and the test position; SF6 gas density interlocks; and the specific permissive conditions that govern each closing and opening operation. A detailed FAQ follows the main article.

2. The One-and-a-Half Breaker Diameter: Anatomy of the Bay

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

1. Why Interlocking Is a First-Order Design Input, Not an Afterthought

3. The "Complete Position" Operating Philosophy

In a one-and-a-half breaker arrangement, two circuits share three circuit breakers connected in series between two main busbars. Each diameter contains a bus-side breaker tied to Bus 1, a second bus-side breaker tied to Bus 2, and a middle breaker shared between the two circuits. Each circuit taps the diameter between a bus-side breaker and the middle breaker.

Every circuit breaker in the diameter is flanked by two motor-operated isolators (disconnectors), and each isolator section carries its own maintenance earth switches. Line-side isolators connect each circuit to the diameter, line earth switches provide grounding for outgoing circuits, and dedicated busbar earth switches allow each main bus to be grounded for maintenance. Current transformer cores are distributed through the diameter to support protection zones that overlap across each breaker.

The value of this topology is redundancy: either circuit can remain in service with any single breaker out for maintenance, and a bus outage does not interrupt either circuit. But that same flexibility multiplies the number of possible switching states which is exactly why the interlocking matrix for a one-and-a-half breaker diameter is substantially more complex than for a single-bus or double-bus arrangement, and why it deserves rigorous engineering review at the design stage.

Keentel Perspective

The interlocking matrix is where the single-line diagram meets operational reality. Every cell in that matrix is a claim about what can never happen in the field. We recommend a formal, line-by-line interlocking review with the utility’s operations staff in the room before control wiring design is released for construction.

Under normal operating conditions, a well-run one-and-a-half breaker diameter is maintained in what operators call the complete position: all three circuit breakers and all series isolators in the diameter are kept closed, even if one of the two circuits is de-energized at its remote end. If a circuit is out of service, its line isolator remains closed along with everything else in the string.

This practice delivers two operational benefits. First, restoration is fast returning the idle circuit to service requires only remote-end action rather than a local switching sequence. Second, it keeps the diameter’s protection zones intact and simplifies the operator’s mental model: the default state of the bay is "everything closed," and any departure from that state is a deliberate, logged exception associated with maintenance or a fault.

Interlocking logic does not act on simple open/closed indications alone. GIS switchgear reports device positions with a richer vocabulary, and the distinction between an in-travel device and a device that has completed its stroke is fundamental to safe permissive logic.

The engineering significance of FO and FCL is straightforward: an isolator caught mid-stroke has neither the dielectric withstand of the open position nor the current-carrying capability of the closed position. Permissive logic built on full-travel confirmation rather than momentary auxiliary contacts ensures that no downstream operation is enabled while any interlocked device is still in motion.

Consider the bus-side circuit breaker connected to Bus 1. It sits between two series isolators — a bus-side isolator toward the busbar and a diameter-side isolator toward the circuit tap — and each isolator section carries its own earth switches. The breaker is governed by two separate interlocking tables: one for closing and one for opening. The asymmetry between them is deliberate, and understanding it is essential for anyone writing or reviewing switching orders.

5.1 Closing in the Service Position

Closing a breaker in the service position means connecting it into a live network, so the permissive chain is the most restrictive in the entire scheme. Four independent conditions must be satisfied:

Isolator status — Both series isolators flanking the breaker must be confirmed Fully Closed (FCL). This guarantees the electrical path through the breaker is continuous and secure before any current is asked to flow. The line isolator further out in the circuit may be either fully open or fully closed depending on the circuit’s operating state it is not part of this breaker’s closing permissive.
Synchronization The synchro-check relay must confirm that voltage magnitude, phase angle, and frequency match across the open breaker contacts. Closing across a mismatch drives severe transient torques into rotating plant, stresses transformer windings, and can destabilize the local network. The synchro-check permissive is bypassed only under a controlled maintenance condition.
SF6 gas density The breaker cannot be closed if gas density has fallen to the Stage 1 alarm threshold, and it is absolutely locked out at Stage 2. Closing is deliberately held to a stricter standard than opening: a breaker that closes with marginal gas may be called on to interrupt a fault an instant later with degraded arc-quenching capability.
Control authority The close command is accepted from the control hierarchy defined in the interlocking table local, remote, or SCADA with the selector-switch position determining which path is live at any moment.

5.2 Opening: Fewer Conditions, One Critical Restriction

Opening a circuit breaker is intentionally easier than closing it. A breaker is the one device in the bay designed to interrupt load and fault current, so the logic never wants to obstruct a trip. Opening does not depend on synchronization, on the position of surrounding earth switches, or on maintenance status. The breaker can be opened with SF6 gas at the Stage 1 alarm level; only the Stage 2 lockout condition blocks it.

There is, however, one restriction that matters enormously: the breaker must not be operated while either of its series isolators is in motion. If the breaker changes state mid-stroke of an isolator, load current can be made or broken across the isolator’s contacts — contacts that have no arc-control capability whatsoever. The result is contact destruction, internal flashover in the GIS enclosure, and a forced outage of the diameter. The interlocking logic enforces this by blocking breaker commands whenever the flanking isolators report an in-travel condition.

Open commands are accepted from four control points: the local control cubicle at the equipment, the remote operator station, SCADA, and a dedicated emergency control path that overrides standard hierarchy restrictions in critical situations.

6. SF6 Gas Density Supervision: The Two-Stage Interlock

Design Note

The asymmetry between Stage 1 and Stage 2 is intentional and worth preserving in every scheme review: at alarm level the system still allows the breaker to remove itself from service, but refuses to let it take on new duty. Lockout at Stage 2 recognizes that operating a breaker with critically low gas density is more dangerous than leaving it in its current state while the gas system is attended to.

Modern gas monitoring extends beyond the hardwired interlock: networked density sensors, trend analysis, leak-rate forecasting, and alarm escalation through station automation systems give maintenance teams days or weeks of warning before an interlock threshold is ever reached. Keentel routinely specifies communicating density monitoring in new GIS designs so that Stage 1 alarms become rare events rather than routine ones.

About Keentel Engineering

Every withdrawable or GIS-class breaker installation distinguishes between two fundamental operating states, and the interlocking logic changes completely between them.

7.1 The Service Position

In the service position, the breaker is fully connected to the high-voltage power system and to its control and auxiliary circuits. Its primary contacts are engaged with the busbars and primary conductors; it conducts load current, interrupts fault current, and responds to protection and control commands. This is the breaker doing its job.

Three requirements define a healthy service position:

Proper primary engagement — The primary contacts must be fully engaged with the high-voltage circuit. Partial engagement causes localized heating, arcing, and progressive contact damage.
Control connectivity — Control and auxiliary circuits must be live and proven, so open/close commands and status feedback (position, alarms, trip events) flow reliably between the breaker and the control system or SCADA.
Position interlocking — Interlocks must prevent the breaker from moving between service and any other position except through the correct, deliberate sequence.

Supporting safety measures include unambiguous position indication (mechanical and electronic), locking of the breaker in the service position during operation, and continuous monitoring of contact resistance, temperature, and gas density.

7.2 The Test Position

In the test position, the breaker is electrically and mechanically disconnected from the high-voltage system but remains connected to its control and auxiliary circuits. The breaker cannot conduct primary current in this state — and that is precisely the point. The test position exists so that the full secondary ecosystem around the breaker can be exercised without touching the live network.

The test position supports four categories of work:

Routine operational testing — Exercising open/close operations, trip mechanisms, and auxiliary contacts with no primary consequence.
Protection verification — Injecting into protection schemes — overcurrent, earth fault, differential — and proving trip paths end-to-end without disturbing the in-service system.
Maintenance and inspection — Servicing and inspecting the breaker while it is positively isolated from the high-voltage circuit.
Pre-commissioning tests — Proving every system on a newly installed or freshly serviced breaker before it is returned to service.

7.3 Side-by-Side Comparison

Keentel Engineering is a power systems and grid interconnection consulting firm serving utilities, developers, and independent power producers nationwide. Our services span EMT and PSS/E modeling, POI interconnection engineering, substation and transmission line design, utility-scale renewables and BESS engineering, owner’s engineer services, and NERC O&P compliance. We believe grid interconnection — and the interlocking, protection, and control engineering that supports it — is a first-order design input, not a late-stage administrative step.

8. Closing the Breaker in the Test Position

4. Status Signaling: Why "Fully Open" and "Fully Closed" Matter

When a breaker is placed in the test position, the interlocking logic inverts several of its service-position rules and the inversions are as instructive as the rules themselves.

Isolation is mandatory — Both series isolators must be confirmed Fully Open. In service, closed isolators guarantee a continuous path; in test, open isolators guarantee complete separation from system voltage so the breaker can be cycled with zero risk of accidental energization.
No synchro-check — With no system voltage on either side of the breaker, there is nothing to synchronize to. The synchro-check permissive is removed from the closing chain.
SCADA closing disabled — Test-position operation is permitted from the local panel, the remote operator station, and the emergency path — but closing from SCADA is deliberately disabled. A dispatcher looking at a system-wide display should never be able to cycle a breaker that field personnel may be working around.
SF6 interlock remains — Gas density supervision does not relax in the test position. If density falls below the Stage 1 or Stage 2 thresholds, operation is inhibited — a breaker with inadequate insulation should not be cycled even for test purposes.

Field Safety Insight

Disabling SCADA control in the test position is one of the quiet, high-value details of a well-designed interlocking scheme. It converts an administrative rule ("do not operate remotely during maintenance") into an engineered impossibility — which is the entire philosophy of interlocking in one sentence.

9. Isolator Interlocking: The Bus-Side Isolator

5. Circuit Breaker Interlocking: The Bus-Side Breaker

SF6 provides both the dielectric insulation and the arc-quenching medium for the breaker, and its effectiveness depends on gas density, not merely pressure. GIS installations therefore use temperature-compensated density monitors with two supervised thresholds feeding directly into the interlocking logic:

7. Service Position vs. Test Position: Two States, Two Logics

10. Isolator Interlocking: The Diameter-Side Isolator

Isolators provide visible, verifiable isolation, but they have no capability to make or break load current. Every permissive in an isolator’s interlocking table exists to guarantee the isolator only ever operates on a dead, current-free section. The conditions differ sharply between service and test regimes.

9.1 Service Position Conditions

For the bus-side isolator — the device connecting the diameter to the main busbar — three conditions must be proven before operation is permitted in the service position:

Breaker open — The associated circuit breaker must be open. With the breaker open, no current can flow through the isolator during its stroke.
Adjacent earth switches open — The earth switches on both isolator sections adjacent to the breaker must be open. This prevents the isolator from connecting an energized busbar into a grounded section — a bolted fault by design.
Busbar earth switch open — The busbar earth switch on the associated main bus must be open, proving the busbar itself is not grounded while the isolator connects to it.

9.2 Test Position Conditions

In the test (maintenance) regime the logic changes character entirely:

Breaker status becomes irrelevant — The isolator may be operated regardless of whether the circuit breaker is open or closed. During maintenance the primary circuit is already dead and grounded, so the breaker’s position is no longer a safety constraint — and maintenance activities frequently require cycling isolators with the breaker in either state.
Earth switches must be closed — The adjacent earth switch and the busbar earth switch must be closed. This is the exact inversion of the service-position rule: in service, grounding anywhere near the isolator is forbidden; in test, positive grounding everywhere around the work area is mandatory. The grounded envelope is what protects personnel from induced voltages and inadvertent energization.

11. How Keentel Engineering Supports GIS Interlocking Projects

The isolator on the other flank of the breaker — between the breaker and the circuit tap — follows the same philosophy with its own set of interlocked devices.

10.1 Service Position Conditions

Breaker open — The associated circuit breaker must be open, for the same current-free reason as before.
All adjacent earth switches open — All earth switches electrically adjacent to the isolator — the earth switches on both breaker flanks and the earth switch on the circuit-tap section — must be open, preventing any inadvertent grounding of the operating path.

10.2 Test Position Conditions

Breaker status irrelevant — The isolator may be operated regardless of breaker status, providing the flexibility maint enance work requires.
Grounded envelope required — The earth switch on the breaker-side section and the earth switch on the circuit-tap section must be closed, keeping the entire working envelope positively grounded throughout the operation.

The Underlying Pattern

Read enough interlocking tables and a single pattern emerges. Service position: prove the path is dead (breaker open) and ungrounded (earth switches open) before an isolator moves. Test position: prove the path is dead and grounded before anything moves. The earth-switch conditions flip polarity between the two regimes and that flip is the signature of a correctly engineered scheme.

Interlocking engineering sits at the intersection of primary design, protection and control, and operations — which is exactly where Keentel Engineering works every day. Our support for utility and developer clients on EHV GIS projects includes:

Interlocking design & review — Development and independent review of interlocking matrices and permissive tables for one-and-a-half breaker, double-bus, and breaker-and-a-half hybrid arrangements.
Factory & site acceptance testing — Verification that vendor GIS interlocking logic, hardwired and IEC 61850 GOOSE-based, matches the approved scheme — including witness testing of service/test position transfers, SF6 threshold responses, and control-hierarchy behavior.
Operations documentation — Switching-order development, standard operating procedures for line isolation and restoration, and operator training material aligned to the as-built interlocking scheme.
Full-lifecycle engineering — Substation and interconnection engineering across the full project lifecycle — from POI studies and EMT modeling through detailed substation design commissioning support, and NERC compliance.

If your project involves a new GIS bay, a retrofit into an existing diameter, or an interlocking scheme that has never had an independent line-by-line review, our team can help. Contact Keentel Engineering at 813-389-7871 or contact@keentelengineering.c

Frequently Asked Questions

The questions below expand on the concepts in the article and reflect the questions our engineers most often field from operations, maintenance, and protection teams working on EHV GIS installations.

Disclaimer:

This document is published by Keentel Engineering for general informational and educational purposes. It does not constitute engineering advice for any specific installation, and switching or interlocking decisions for any facility must be based on that facility’s approved drawings, interlocking tables, and operating procedures. All product names, trademarks, and registered trademarks referenced are the property of their respective owners. Keentel Engineering is not affiliated with, endorsed by, or sponsored by any equipment manufacturer or organization mentioned or implied herein.

Data Center EMT Modeling & Large Load Interconnection Guide

Wed, 08 Jul 2026 15:43:50 GMT

Jul 08, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Data Center Model Quality, Hardware Validation & Load-Fluctuation Mitigation

Modern AI data centers are packed with power electronics: server power supplies, double-conversion UPS systems, static transfer switches, VFD-driven cooling, and increasingly on-site BESS and STATCOM devices. Electrically, a gigawatt-class AI campus behaves less like a traditional industrial load and more like an inverter-based resource (IBR) operating in reverse. It is fast, non-linear, sensitive to weak-grid conditions, and capable of tripping or refusing to reconnect in ways conventional composite load models simply cannot represent.

The consequences are no longer hypothetical. In one widely discussed 2024 event, a routine 230 kV feeder fault near a 1.5 GW data center facility triggered repeated failed reclosure attempts: the facility's protection and controls judged the grid 'not stable enough' to reconnect, while the utility urgently needed that load back to preserve system stability. Roughly half a dozen reclosure attempts followed. The root cause was not a single device failure it was the absence of coordinated control, protection, and grid-interface behavior that was never studied, never modeled, and never tested before energization.

Grid operators have taken notice. ERCOT's interconnection queue now contains more than 400 GW of large-load requests, the vast majority being data centers and computational loads at the 1–2 GW scale. European TSOs report tens of gigawatts of new demand connection inquiries within a two-year window figures comparable to their entire existing peak load. The response has been a wave of new grid codes, modeling requirements, and model quality test (MQT) frameworks aimed squarely at large computational loads.

Keentel Insight

If your interconnection strategy still treats the EMT model as a deliverable to be 'handled later by the vendor,' you are carrying schedule risk measured in quarters, not weeks. Model quality checkpoints are now embedded in interconnection milestones a rejected model can freeze your energization ramp.

2. The New Regulatory Landscape: Ride-Through Is Now a Requirement, Not a Courtesy

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

1. Why Large Loads Suddenly Look Like Inverter-Based Resources

How EMT Model Quality Testing, Hardware-in-the-Loop Validation, and Megawatt-Scale Testing De-Risk Large-Load Interconnection

3. Why Generic Models Fail and What 'Model Quality' Actually Means

The interconnection conversation for AI data centers has fundamentally changed. Two years ago, most utilities treated a large load as a passive megawatt number on a one-line diagram. Today, transmission operators on both sides of the Atlantic are demanding vendor-specific EMT models, model quality tests, hardware benchmarking, and ride-through performance guarantees before a single GPU rack energizes. A recent industry panel on hardware testing and model quality assessment featuring transmission system operators, a U.S. national laboratory, and specialist EMT consultancies made one thing unmistakably clear: the model is now part of the interconnection agreement, and a bad model can stall gigawatts of load.

At Keentel Engineering, we treat grid interconnection requirements FERC LGIP procedures, NERC MOD/PRC/FAC obligations, and ISO-specific large-load rules as first-order design inputs, not late-stage administrative steps. This article distills what the industry's leading testing and modeling practitioners are now expecting from data center developers, and what it means for your project schedule, your equipment selection, and your interconnection risk.

2.1 ERCOT: NOGRR282 and PGRR144

ERCOT has moved decisively. Nodal Operating Guide Revision Request 282 (NOGRR282), approved by the ERCOT Board in mid-2025, defines voltage and frequency ride-through performance requirements for large computational loads (LCLs). Key expectations include continued power consumption through defined low-voltage and high-voltage bands, and restoration of load to pre-disturbance levels within two seconds of voltage recovery after a cleared fault.

In parallel, Planning Guide Revision Request 144 (PGRR144) establishes the data submission and model review machinery: a large-load data survey (on the order of 69 questions spanning fundamentals through protection settings and ride-through characteristics), a formal Model Quality Test framework, and defined model checkpoints throughout the interconnection process. Models must be delivered and cross-validated across three platforms PSS/E for positive-sequence dynamics, PSCAD for high-fidelity EMT, and TSAT for real-time operational security assessment with the PSS/E response benchmarked against the PSCAD response as an explicit quality gate.

PGRR144 defines three principal model checkpoints: before stability study entry (demonstrating voltage ride-through capability), before the Quarterly Stability Assessment (full MQT and multi-platform model submission), and for large computational loads only a final as-studied versus as-built model reconciliation before initial energization. Loads co-located behind existing or new generation face an additional, more rigorous layer of review through the generation interconnection process.

2.2 European TSO Practice: Models as a Condition of Energization

European transmission operators are converging on similar principles, in some respects moving even faster. One Nordic TSO facing roughly 54 GW of new demand connection inquiries across 2024- 2025 against a national peak load of only 16 GW has issued PSS/E and PSCAD modeling requirements for demand facilities and released a new demand grid code introducing low-voltage ride-through, over-voltage ride-through, phase-jump performance, sub-synchronous active power oscillation limits, and post-fault active power recovery requirements (recovery within one second for facilities above 30 MW).

Two features of this approach deserve particular attention from developers:

Staged model acceptance tied to energization blocks. The first plant model is due six months before the interim operational notice. The facility may then ramp in 30 MW increments but each additional block is only released after the updated model passes validation. The full-plant model must be accepted before the final operational notice and unrestricted operation.
Vendor-specific 'real-code' EMT models for UPS systems. Generic library representations are accepted only as a bridge. UPS and power-supply equipment must ultimately be represented by manufacturer-specific models embedding actual controller code, validated against factory test results.

Keentel Insight

The pattern is unmistakable across jurisdictions: model quality is being tied directly to your revenue ramp. Under staged-acceptance regimes, every week a model languishes in review is a week your capacity and your customers' compute sits idle.

The industry learned this lesson the hard way with inverter-based generation. Major solar-loss events in Texas in 2021 and 2022 with 1.1 GW and 1.7 GW of unintentional curtailment respectively occurred at facilities that had completed conventional PSCAD and PSS/E studies. Root-cause analyses pointed to phenomena that generic or poorly parameterized models could not capture: PLL loss of synchronism, inverter AC overvoltage response, momentary cessation, and unmodeled control-loop and communication behavior.

Large computational loads inherit every one of these vulnerabilities, plus several of their own. Based on the practices now emerging from TSOs, national laboratories, and EMT specialists, a defensible data center model must capture:

Real-code power supply and UPS behavior. A 100 MW step at the GPU level may appear at the medium-voltage bus as a shaped ramp or as a 120 MW transient with overshoot, meaning the mitigation system must compensate more than the nominal load change. Only manufacturer real-code models capture this reliably.
Control and communication latency. In active compensation schemes, a single millisecond of additional measurement, processing, filtering, or communication delay can swing a facility from compliant to non-compliant. Latency values in models should be deliberately conservative (worst-case) rather than tuned to pass.
Impedance fidelity for passive mitigation paths. For double-conversion UPS and grid-forming BESS solutions, battery internal impedance and lead impedances determine whether the DC bus truly decouples load transients from the grid. Modeling the DC bus as an ideal voltage source hides ripple that appears on the AC side in reality.
Correct measurement and monitoring practice. Instantaneous active power in EMT tools must be evaluated with minimal smoothing — a typical 20 ms filter constant suppresses everything above roughly 8 Hz and can mask the very fluctuations a grid requirement targets. A ~1 ms filter preserves the phenomena of interest.
Realistic load ramp profiles. A load profile can be chosen to pass or chosen to fail; neither is acceptable. Mitigation studies must be run against realistic AI training and inference ramp characteristics, including synchronized load-crash scenarios.
Aggregation and functional description. The connecting entity should propose and justify how thousands of racks, UPS branches, and mitigation devices aggregate into a study model, and document plant behavior in normal, disturbed, and load-sharing/parallel-facility conditions.

A recurring theme among practitioners is that offline simulation is necessary but not sufficient. The emerging best practice is a three-stage validation continuum, with each stage retiring risk the previous stage cannot see.

4.1 Stage 1 Offline EMT and Positive-Sequence Studies

PSCAD/EMT studies remain the foundation: ride-through screening, phase-jump response, load-fluctuation mitigation sizing, and sub-synchronous behavior. For fluctuation-mitigation studies, a compact single-machine-infinite-bus–style EMT model — analogous to an MQT setup — is generally sufficient; a wide-area model is not required. System strength (SCR and X/R) must be represented faithfully, since grid-forming devices' initial phase-jump power injection depends directly on it. Converter models should be average-value models for simulation performance, with vendor real-code models for all UPS systems.

4.2 Stage 2 Controller Hardware-in-the-Loop (CHIL)

In offline simulation, time freezes while the solver exchanges data with the control model; in the real world, the clock keeps ticking. Hardware-in-the-loop testing closes that gap by connecting the physical controllers power conversion system controllers, plant/power management systems, protection relays, and even metering with synchrophasor interfaces to a real-time simulator representing the facility and the grid.

The payoff is discovering integration defects before commissioning. In one 20 MW data center project in the Middle East, CHIL testing exposed control 'jittering' during grid-forming to grid-following transitions of the power conversion system behavior invisible in offline models and initially denied by the OEM until the hardware evidence was on the bench. The same testbed validated protection control coordination, legacy utility communication protocols, unintentional islanding response against IEEE 1547 criteria, and damping of sub-synchronous torsional interactions between on-site gensets and fluctuating AI load.

4.3 Stage 3 Megawatt-Scale Grid Simulator Testing

At the top of the continuum sits full-power hardware validation. A leading U.S. national laboratory now operates a grid integration testbed with two megawatt-class grid simulators (7 MVA and 20 MVA, 13.2–34.5 kV, response times down to 100 microseconds), a medium-voltage impedance network capable of emulating short-circuit ratios down to 1, and a site-wide GPS-synchronized 24-bit, 50 kHz data acquisition network. Critically, a device under test can be connected between the two grid simulators one emulating the fluctuating AI load, the other emulating grid disturbances which is precisely the configuration needed to validate double-conversion and medium-voltage UPS architectures.

Demonstrated capabilities include AI load-profile emulation with zero-to-full-power ramps of several megawatts in four to five milliseconds, low-voltage ride-through testing benchmarked directly against PSCAD models, and frequency-domain impedance scans comparing hardware admittance against model admittance. Practitioners report that initial vendor models frequently required multiple correction iterations before hardware and model responses aligned a finding that should give every developer pause about unvalidated models in their interconnection submittals. ERCOT now accepts hardware-in-the-loop converter model validation performed once per converter family in an OEM environment, making this a leverageable investment across a fleet.

Keentel Insight

Think of validation as insurance priced in milliseconds. Every phenomenon caught at the CHIL bench or the grid simulator is a commissioning delay, a compliance finding, or a reclosure standoff that never happens in the field.

5. The 'Million-Dollar Slide': Comparing Load-Fluctuation Mitigation Technologies

Synthesizing the regulatory trajectory and the validation practices above, Keentel Engineering recommends data center developers and their EPC partners adopt the following sequence:

Engage the interconnecting utility's modeling requirements before equipment selection. UPS and mitigation vendor choices constrain what models you can deliver; real-code model availability should be a procurement criterion, not an afterthought.
Build the EMT model quality test into the project schedule as a critical-path activity with explicit checkpoints mirroring the ISO's (pre-stability-study, pre-assessment, pre-energization).
Run mitigation-selection EMT studies early against realistic load profiles and the actual point-of-interconnection limits in absolute megawatts to size BESS, E-STATCOM, or UPS-based solutions before layouts freeze.
Validate controllers with CHIL before commissioning, including protection–control coordination, utility communication protocols, islanding response, and mode-transition behavior.
Leverage once-per-family hardware benchmarking. Where the ISO accepts OEM-environment hardware-in-the-loop converter validation, coordinate with vendors so a single campaign covers your fleet.
Plan for staged model acceptance where applicable align energization block schedules, model update deliveries, and validation reviews so the revenue ramp is never hostage to a model queue.

The direction of travel is clear: within a few years, delivering a validated, hardware-benchmarked EMT model will be as routine an interconnection requirement for a gigawatt data center as a protection coordination study is today. Developers who internalize that now will interconnect faster, at lower risk, and with mitigation systems sized by engineering rather than by guesswork.

Work With Keentel Engineering

Keentel Engineering provides EMT/PSCAD model development and model quality testing, PSS/E dynamic modeling, large-load interconnection engineering ride-through compliance assessment, mitigation-selection studies for BESS/E-STATCOM/UPS architectures, and owner's engineer support for hardware-in-the-loop and factory validation campaigns. Contact our team to de-risk your large-load interconnection from the first inquiry to final energization.

The most valuable synthesis presented in the panel described by its presenter as the 'million dollar slide' condensed a large body of EMT mitigation studies into a qualitative comparison of eight mitigation approaches for AI load fluctuation. Keentel's recreation of that comparison is presented below.

Table 1 Qualitative Comparison of AI Load-Fluctuation Mitigation Methods (synthesized from industry EMT study results)

6. What This Means for Your Project Keentel's Recommended Playbook

Several strategic conclusions flow from this comparison:

There is no universal winner. Grid-forming BESS is mature, efficient, and fast, but its physics-based (phase-jump-driven) compensation is weakest exactly where AI loads are most aggressive high ramp rates. Double-conversion architectures deliver very high compensation across ramp rates but pay for it in losses, since the full facility power flows through two back-to-back conversion stages; shunt devices need only be sized for the fluctuating component.
Hybrid solutions dominate the balanced scorecard. E-STATCOM plus GFM BESS pairs supercapacitor speed (seconds-scale storage) with battery depth (minutes-scale storage), covering both fast phase-jump response and low-frequency oscillation. Similarly, LV double-conversion UPS supplemented by GFM BESS overcomes the UPS's limited internal energy storage.
Grid limits are absolute megawatts not per-unit. Emerging fluctuation requirements are hard-coded numbers (for example, 10–20 MW peak-to-peak residual at the point of interconnection) regardless of facility size. A 1,000 MW campus meeting a 10 MW limit needs roughly 0.1% residual demanding the highest-efficiency mitigation while a 200 MW campus may comply comfortably with GFM BESS or an E-STATCOM hybrid. Mitigation selection is therefore inseparable from facility sizing.
Active schemes live and die on latency. For actively controlled compensation (rows 2–4), communication, measurement, processing, and filtering delays are decisive; a one-millisecond change can flip a pass to a fail. Models must use defensibly conservative delay assumptions, and closed-loop response times should be confirmed with manufacturers.
GPU-level mitigation is coming, but unproven. Smoothing load at the silicon or rack power-management level could shrink the external mitigation burden dramatically — but maturity, losses, and verified performance data remain open questions today.

4. The Validation Continuum: From Offline EMT to Megawatt-Scale Hardware

Case Studies

Disclaimer

The following case studies are drawn from representative large-load interconnection and validation engagements. Client names, locations, and identifying details have been anonymized; figures are rounded and illustrative of project scale.

Case Study 1 Model Quality Test Readiness for a 700 MW AI Campus in ERCOT

Background

A hyperscale developer was advancing a 700 MW AI training campus through the ERCOT large-load interconnection process just as NOGRR282 ride-through requirements and the PGRR144 model review framework took effect. The project's original modeling plan a composite load model plus a vendor-supplied generic UPS block predated the new rules and would not survive the first model checkpoint.

Challenge

Models were required across three platforms (PSS/E, PSCAD, and TSAT), with the PSS/E dynamic response benchmarked against PSCAD as an explicit quality gate.
The facility had to demonstrate NOGRR282 voltage ride-through performance continued consumption through the defined voltage bands and load restoration within two seconds of recovery before entering the stability study.
The transmission service provider's 69-item large-load data survey exposed gaps in protection settings, ride-through characteristics, and aggregation documentation that the developer's vendor data packages did not cover.

Keentel's Approach

Built a site-specific PSCAD model with average-value converter representations and vendor real-code UPS models secured through procurement-stage engagement with the UPS OEM.
Developed the load aggregation methodology and functional description — normal operation, fault response, and load-sharing behavior — as a formal submittal, rather than leaving aggregation implicit.
Executed the full site-specific MQT: flat-start (no-disturbance quiescence) and large-load disturbance tests sweeping the ride-through envelope from gradual voltage variations through shallow and deep fault dips and high-voltage excursions.
Cross-validated the PSS/E model against the PSCAD reference, iterating parameters until dynamic responses aligned within review tolerances, then confirmed TSAT consistency for operations use.

Outcome

The project cleared the pre-stability-study model checkpoint on first submission and entered its Quarterly Stability Assessment window on schedule. The as-studied versus as-built reconciliation before initial energization required only minor parameter updates. The developer avoided an estimated one-to-two-quarter re-study cycle that a rejected first submission would have triggered schedule value far exceeding the entire modeling budget.

Key Takeaway

Under PGRR144-style frameworks, the cheapest model checkpoint to pass is the first one. Front-loading real-code model procurement and aggregation documentation converts model review from a schedule risk into a formality.

Case Study 2 Mitigation Technology Selection for a 250 MW Campus Facing a Hard 10 MW Fluctuation Limit

Background

A colocation operator planning a 250 MW AI-ready campus received point-of-interconnection requirements from its transmission provider that included a hard limit on residual power fluctuation: peak-to-peak variation at the POI was not to exceed 10 MW an absolute figure, not a percentage of facility size. The operator's initial assumption was that a low-voltage double-conversion UPS fleet, already planned for reliability, would inherently satisfy the requirement.

Challenge

The absolute limit equated to a 4% residual on facility rating achievable by several technologies, making over-specification (and its capital and efficiency penalties) the real risk.
LV double-conversion UPS units offered excellent high-ramp-rate compensation but limited internal energy storage, high through-losses on the full facility power, and weak performance against low-frequency load oscillation.
Anticipated AI training profiles included both millisecond-scale ramps and multi-second synchronized swings, plus a credible full load-crash scenario.

Keentel's Approach

Constructed a compact EMT study model a single-machine-infinite-bus-style representation consistent with model quality test practice at the utility-specified SCR and X/R, with vendor real-code UPS models and realistic load profiles spanning step, second-order ramp, and load-crash cases.
Screened four candidate architectures from the mitigation comparison matrix (Table 1): GFM BESS alone; E-STATCOM plus GFM BESS; LV double-conversion UPS alone; and LV double-conversion UPS plus GFM BESS.
Applied disciplined monitoring practice approximately 1 ms instantaneous-power filtering so high-frequency residuals the requirement targets were not masked by measurement smoothing.
Stress-tested active-compensation variants against conservative worst-case communication and processing latencies, demonstrating that a one-millisecond latency increase moved one candidate from compliant to marginal.

Outcome

The screening showed the facility did not require a double-conversion mitigation path for grid compliance: an E-STATCOM plus GFM BESS hybrid met the 10 MW peak-to-peak limit across all realistic profiles, with very low losses and roughly full coverage of both fast phase-jump events and low-frequency swings. The operator retained LV UPS purely for IT reliability at reduced ratings, and avoided routing the entire 250 MW through back-to-back conversion saving multi-megawatt continuous losses and material capital. The EMT study package was accepted by the transmission provider without revision.

Key Takeaway

Because fluctuation limits are hard megawatt numbers, mitigation selection scales inversely with facility size. Engineering the study realistic profiles, correct filtering, conservative latencies is what separates a defensible selection from an expensive guess.

Case Study 3 Controller Hardware-in-the-Loop Validation Rescues a 20 MW Data Center Commissioning

Background

An international developer was delivering a 20 MW data center with on-site gensets, a central BESS, UPS-fed IT load, and a plant-level power management system, interconnecting to a national utility that mandated a legacy IEC 60870-5-104 communication interface and IEEE 1547-aligned islanding behavior. Keentel served in an owner's engineer capacity for pre-commissioning validation strategy.

Challenge

Offline PSCAD studies had been completed, but the physical power conversion system (PCS) controller, protection relays, revenue/synchrophasor metering, and the power management system had never operated together as a coordinated system.
The facility required seamless grid-forming to grid-following mode transitions of the PCS behavior governed by proprietary controller code that offline models represented only approximately.
Preliminary analysis flagged a sub-synchronous torsional interaction risk between the genset shafts and fast-fluctuating AI load.

Keentel's Approach

Specified a comprehensive controller hardware-in-the-loop campaign: the facility and grid modeled on a real-time simulator, exchanging analog voltage/current signals and digital firing commands with the physical PCS controller, and testing the physical protection relays and meters in the loop so protection control coordination and the C37.118 synchrophasor path were exercised end-to-end.
Validated the mandated IEC 104 utility interface against the real-time model, retiring a protocol-integration risk that could not be tested any other way before site arrival.
Executed unintentional-islanding scenarios to demonstrate IEEE 1547-consistent voltage/frequency behavior, and swept mode-transition cases across load and grid-strength conditions.
Evaluated a generator-centric damping solution a dedicated battery-based damper at the genset rather than a facility-scale battery against the sub-synchronous torsional risk across multiple oscillation amplitudes and frequencies.

Outcome

The CHIL campaign exposed control jittering during grid-forming to grid-following transitions of the PCS a defect invisible in offline simulation. Presented with the bench evidence, the OEM implemented a hardware fix (output capacitance at the affected terminal) before shipment. Protection settings were re-coordinated after in-loop testing showed premature tripping would have prevented full facility utilization. The torsional damper demonstrated effective oscillation damping across the tested profiles. The facility commissioned without a single controller-related field defect, and the utility accepted the validation dossier as evidence of interconnection readiness.

Key Takeaway

In offline simulation, time freezes while models exchange data; on real hardware, the clock keeps ticking. CHIL testing is the only pre-commissioning stage where latency, protocol, coordination, and mode-transition defects reveal themselves at bench cost rather than commissioning cost.

This document was prepared by Keentel Engineering for general informational and educational purposes. Keentel Engineering is an independent consulting engineering firm and is not affiliated with, endorsed by, or sponsored by ERCOT, NERC, FERC, IEEE, any transmission system operator, national laboratory, equipment manufacturer, software vendor, or any other organization referenced herein. All trademarks, product names, and standards designations are the property of their respective owners and are used solely for identification. Case studies are anonymized and illustrative; figures are representative and rounded. Regulatory requirements including NOGRR282, PGRR144, and TSO grid codes evolve; readers should consult the current official documents and their interconnecting entity before making project decisions. This document does not constitute engineering services, and no professional engineering relationship is created by its distribution. Project-specific engineering should be performed under the responsible charge of a licensed Professional Engineer.

Frequently Asked Questions

NERC Large Loads Action Plan (LLAP) Guide for AI Data Centers

Wed, 08 Jul 2026 13:52:14 GMT

Jul 08, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

NERC's Large Loads Action Plan

A single steel mill draws 100 to 200 megawatts. The computational load centers now requesting interconnection to the North American bulk power system (BPS) are asking for thousands of megawatts at a single point of interconnection — the equivalent of adding a city the size of San Francisco or Washington, D.C. to the grid, behind one substation fence. Some individual facilities now exceed 1 gigawatt of demand, roughly equal to 250,000 homes at peak consumption.

This is not a hypothetical planning scenario. It is happening now, it is concentrated in geographic pockets with favorable land and power availability, and it has already produced real reliability events. In response, the North American Electric Reliability Corporation (NERC) has moved from studying the problem to regulating it. The Large Loads Action Plan (LLAP) Q2 2026 Update, published in July 2026, confirms that NERC has issued a Level 3 Essential Action Alert with a mandatory response deadline of August 3, 2026, published a Reliability Guideline, opened the door to registering data centers as NERC entities, and put new Reliability Standards on track for Board adoption in December 2026.

For data center developers, utilities, transmission owners, and grid operators, the message is unambiguous: large-load interconnection is no longer just a transmission service and energy procurement exercise — it is now a reliability engineering and compliance discipline. This article breaks down what NERC has done, what is coming next, and how the engineering requirements translate into design, modeling, and commissioning obligations at the project level.

Why NERC Is Acting: The Reliability Problem in Plain Terms

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

NERC's Large Loads Action Plan: The Compliance Clock Is Now Running for Gigawatt-Scale Data Centers

Integrating Gigawatt-Scale Data Centers and Computational Loads Reliably onto the Bulk Power System Q2 2026 Update, Compliance Deadlines, and What They Mean for Developers, Utilities, and Grid Operators

From Task Force to Action Plan: How We Got Here

Modern computational loads AI training campuses, hyperscale data centers , and cryptocurrency mining facilities behave differently from any load class the grid has served before. Three characteristics drive the risk:

Unprecedented scale and concentration. Individual facilities exceeding 1 GW, clustered in a handful of regions, mean a single site can dominate the load profile of its transmission zone.
Voltage-sensitive, self-protective behavior. The power electronics inside these facilities — server power supplies, UPS systems, and switchgear protection — are designed to protect the computational equipment, not the grid. When a routine transmission fault causes a voltage sag, these facilities can shed hundreds or thousands of megawatts in a fraction of a second, without any operator command.
Behind-the-meter opacity. These dynamics occur inside the facility, beyond the visibility of the transmission operator. The result is a growing discrepancy between predicted and actual load response during disturbances.

NERC's incident reviews from January 2025 (simultaneous voltage-sensitive load reductions) and January 2026 (voltage-sensitive cryptocurrency load reductions) documented exactly this failure mode: grid disturbances triggering sudden, uncommanded large-load reductions that produced voltage and frequency excursions operators had to correct rapidly to preserve system stability. NERC has further observed customer-initiated load reductions and significant oscillations that develop within seconds — leaving little or no room for real-time operator response.

The core engineering issue:

when 1,500 MW of load disconnects in under a second, the grid experiences it much like the sudden loss of a large generator local voltage rises, frequency swings, and nearby generation and protection systems must absorb a transient that was never studied in the interconnection process. The fix is not operational heroics; it is accurate load modeling, disturbance ride-through design, and protection coordination engineered before energization.

NERC established the Large Loads Task Force in August 2024 since evolved into the Large Loads Working Group (LLWG) and developed a structured work plan. Through 2025, NERC published its Characteristics and Risks of Emerging Loads report, issued a Level 2 Alert urging specific industry actions, hosted workshops and technical conferences, released a draft Reliability Guideline, and filed comments in FERC's Advance Notice of Proposed Rulemaking (ANOPR) on the reliable interconnection of large loads.

In Q1 2026, the tempo accelerated: NERC published its incident review on voltage-sensitive crypto load reductions, an assessment of gaps in existing practices and Reliability Standards, an aggregated report on the Level 2 Alert responses, appointed a Standards Drafting Team, and submitted a supplemental letter to FERC detailing an accelerated action plan. That acceleration culminated in the four major Q2 2026 actions below.

1. Level 3 Essential Action Alert Responses Due August 3, 2026

On May 4, 2026, NERC issued a Level 3 Essential Action Alert: Computational Load Modeling, Studies, Instrumentation, Commissioning, Operations, Protection, and Control. Level 3 is NERC's highest alert tier — reserved for essential actions — and this is one of the rare instances NERC has invoked it. The alert outlines seven actions that registered entities must implement to address the immediate risks posed by computational loads interfacing with the BPS.

The scope embedded in the alert's title is effectively a checklist of the engineering lifecycle: dynamic load modeling, interconnection and system impact studies, disturbance instrumentation and monitoring, commissioning verification, operating procedures, and protection and control coordination. Registered entities — transmission operators, planners, and others serving these facilities — must submit their responses by August 3, 2026. An informational webinar on July 8 addresses frequently asked questions.

2. Reliability Guideline: Risk Mitigation for Emerging Large Loads

Published May 1, 2026, the Reliability Guideline recommends concrete actions not only for utilities and grid operators, but also for the companies behind emerging large loads — including equipment manufacturers. The intent is that industrial-scale consumers actively participate in practices that protect grid stability: sharing accurate load composition data, designing for voltage ride-through, coordinating protection settings, and supporting model validation.

Although the guideline is non-binding, NERC strongly urges adoption — and there is a commercial incentive to comply: NERC's own framing is that proactive planning and participation can enable more of these facilities to come online reliably and quickly. Developers who show up to the interconnection queue with validated dynamic models and ride-through-capable designs will move faster than those who treat these requirements as friction.

3. Registration Criteria: Data Centers as NERC-Registered Entities

This is the structural change with the longest-term consequences. NERC is revising its Rules of Procedure — specifically Appendix 5B, the Statement of Compliance Registry Criteria — to create a new registration category for Computational Load Entities. Entities meeting defined physical and electrical criteria would be required to register with NERC and comply with applicable Reliability Standards, just as generators, transmission owners, and balancing authorities do today.

The initial draft was posted for a 45-day stakeholder comment period on April 1, 2026. Initial comments produced meaningful insights that NERC will incorporate before the next comment period, anticipated in August 2026. A July 13 webinar will cover the registration process for entities likely to fall under the new criteria. For large data center owners and operators, this means an entirely new compliance program — registration, standards applicability, audits, and enforcement exposure — is on the horizon.

4. New Reliability Standards on a December 2026 Fast Track

In parallel, NERC is drafting Reliability Standards to define computational loads and establish measurable requirements for newly registered entities. The sequence so far: a Standard Authorization Request (SAR) with new computational-load definitions and a foundational standard was posted for comment in April 2026; the Standards Committee approved a final SAR and authorized drafting in May. The proposed standards will be posted for a 45-day comment period in August 2026, with an additional comment period in Q4, and the Reliability Standard is targeted for final Board approval and adoption in December 2026.

By NERC standards-development norms, this is extraordinarily fast — a reflection of how seriously NERC views the risk. Entities that wait for the final standard before adjusting their engineering practices will find themselves retrofitting compliance into projects already in the queue.

The Complete LLAP Timeline: 2024 Through 2027

The table below consolidates the full arc of NERC's actions, from task force formation through the standards filing targets.

What This Means in Practice — By Stakeholder

What This Means in Practice By Stakeholder

Disclaimer

Data Center Developers and Owners

The era of showing up to a utility with only a load letter and an energization date is over. Expect interconnecting utilities and planning coordinators to require validated dynamic load models (increasingly aligned to the forthcoming Data Center Load Modeling Technical Reference and the PERC2 model specification), documented voltage ride-through behavior for the facility's power electronics and protection, disturbance monitoring instrumentation at the POI, and commissioning test plans that verify as-built behavior against the studied model. Developers should also track the registration criteria closely — facilities above the eventual physical/electrical thresholds will inherit direct NERC compliance obligations.

Utilities, Transmission Owners, and Planning Coordinators

The Level 3 Alert puts the immediate burden here: seven essential actions spanning modeling, studies, instrumentation, commissioning, operations, protection, and control, with responses due August 3, 2026. Practically, that means auditing which large loads are on your system and in your queue, assessing whether current interconnection study practices capture voltage-sensitive load loss as a contingency, verifying protection coordination at large-load POIs, and closing the gap between forecasted and actual load dynamic behavior through measurement and model validation.

Equipment Manufacturers

The Reliability Guideline explicitly reaches OEMs — server power supply, UPS, VFD, and switchgear vendors whose default protection and control settings determine how facilities respond to grid disturbances. Ride-through capability, configurable trip settings, and transparent equipment response characteristics are becoming procurement differentiators.

This document was prepared by Keentel Engineering for general informational and educational purposes only. Keentel Engineering is an independent engineering consulting firm and is not affiliated with, endorsed by, or sponsored by the North American Electric Reliability Corporation (NERC), the Federal Energy Regulatory Commission (FERC), or any regional transmission organization, independent system operator, or regulatory authority referenced herein. All trademarks, program names, and publication titles referenced — including the Large Loads Action Plan (LLAP) — are the property of their respective owners and are used solely for identification and commentary.

The regulatory milestones, deadlines, and requirements summarized in this document are based on publicly available NERC materials current as of July 2026, including the Large Loads Action Plan Q2 2026 Update. Regulatory schedules, comment periods, and standard content are subject to change; several milestones are expressly conditional (e.g., subject to RSTC or Board approval). Readers should consult NERC's official publications and their own legal and compliance advisors before making decisions based on this material. This document does not constitute legal, regulatory, or compliance advice.

For project-specific engineering support, contact Keentel Engineering — Tampa, FL (HQ) | Austin, TX | Sacramento, CA | Baltimore, MD.

The Engineering Response: How Keentel Engineering Supports LLAP Readiness

The Four Pillars of the Q2 2026 Update

Keentel Engineering has long advocated a principle that NERC's action plan now codifies: grid interconnection and reliability compliance are first-order design inputs, not late-stage administrative steps. Our power systems engineering practice maps directly onto the obligations the LLAP creates:

Dynamic load modeling and EMT studies. Development and validation of dynamic and electromagnetic transient (EMT/PSCAD) models for large computational facilities, including composite load representation, power-electronic interface behavior, and alignment with emerging references such as the Data Center Load Modeling Technical Reference and PERC2 specification.
Interconnection and system impact studies. POI selection, load flow, short circuit, stability, and voltage-sensitivity studies that treat sudden large-load loss as a studied contingency — for both the developer's application and the utility's system assessment.
Voltage ride-through and protection coordination. Facility-level ride-through design reviews, protection setting coordination between the utility POI and facility switchgear, and mitigation of uncommanded load-shed behavior — the exact failure mode documented in NERC's 2025 and 2026 incident reviews.
Substation and transmission design. 30%/60%/90%/IFC design of the substations, transmission interconnections, and metering/instrumentation infrastructure that gigawatt-scale campuses require — with disturbance monitoring designed in from the start.
Commissioning and energization support. Commissioning test plans and field verification that as-built facility behavior matches the models submitted in the interconnection process — a core theme of the Level 3 Alert.
NERC compliance program support. Level 3 Alert response preparation, gap assessments against the Reliability Guideline, comment-period technical support, and readiness planning for prospective Computational Load Entity registration and the December 2026 standards.

The bottom line: NERC's timeline is compressed by design. Level 3 Alert responses are due August 3, 2026; proposed standards post in August; Board adoption is targeted for December. Whether you are a developer with gigawatts in the queue or a utility serving them, the entities that engineer for these requirements now will interconnect faster, operate more reliably, and avoid retrofitting compliance later.

Keentel Engineering provides power system studies, grid interconnection engineering , substation and transmission design, EMT/PSCAD modeling, and NERC compliance support from offices in Tampa, Austin, Sacramento, and Baltimore. Contact our team to discuss LLAP readiness for your facility or system.

Frequently Asked Questions: NERC's Large Loads Action Plan

AI Data Center Power Infrastructure & Grid Engineering Guide

Wed, 08 Jul 2026 10:20:25 GMT

Jul 08, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

The Five-Layer Cake Starts with the Grid

At the 2026 Global Conference, NVIDIA CEO Jensen Huang sat for a wide-ranging conversation about where artificial intelligence is heading. Most of the coverage focused on models, market caps, and the “boomer versus doomer” debate. But buried in that conversation was a message aimed squarely at our industry: the binding constraint on AI is no longer software or even silicon. It is energy, land, and grid infrastructure — the physical foundation that every megawatt of “intelligence production” sits on.

Huang described AI as a “five-layer cake”: energy at the base, then chips, then infrastructure (land, power, and shell — the data center buildings, cloud services, and neoclouds), then models, and finally applications. His point was blunt: without the underlying layers, no model is useful. For power systems engineers, utilities, transmission planners, and large-load developers, that means the AI build-out is now fundamentally an electrical infrastructure program — arguably the largest sustained load-growth event the U.S. grid has faced since post-war industrialization.

In this article, Keentel Engineering breaks down the four themes from Huang's remarks that matter most to the power sector — energy, chips, infrastructure, and U.S. re-industrialization — and translates each into concrete engineering and interconnection implications.

2. From Generative to Agentic AI: Why Compute Demand Jumped ~1,000×

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

1. Why a Power Engineer Should Read This Transcript

What NVIDIA CEO Jensen Huang's 2026 Global Conference Remarks Mean for Energy, Chips, Infrastructure — and the U.S. Power System

An Engineering Perspective for Utilities, Developers, and Large-Load Customers

3. The Five-Layer Cake and Why Energy Is the Foundation

Huang framed the last two years as a phase change. Generative AI the ChatGPT moment taught machines to generate text, images, and video. But the industry quickly realized that generation enables two deeper capabilities: reasoning (generating internal chains of thought) and tool use (generating commands to control browsers, spreadsheets, and software). The result is agentic AI: systems that understand intent, reason, plan, and act.

The electrical punchline is in the arithmetic. Huang estimated that the computation required for agentic reasoning is on the order of 1,000 times more than simple generative AI and that demand is then multiplied again by rapid growth in the number of users. Older GPUs sold four or five years ago are appreciating in price because there is not enough compute to go around. He also noted that the computing paradigm itself is shifting from retrieval (pre-recorded content fetched from the cloud) to generation (every response computed fresh, in context). Nothing can be pre-computed and cached; everything must be generated on demand.

For grid planners, this is why data center load forecasts keep being revised upward. Retrieval-era data centers were storage-heavy and comparatively power-light. Generation-era AI factories are computation-dense, thermally intense, and power-hungry and their duty cycle looks less like an office building and more like a continuous industrial process with fast, large power swings driven by synchronized GPU training steps.

Keentel Engineering Insight

Agentic AI workloads change the electrical character of the load, not just its magnitude. Training clusters can exhibit coordinated MW-scale power steps in milliseconds as thousands of GPUs synchronize. Interconnection studies for AI campuses should evaluate voltage flicker, ride-through behavior, harmonic injection from high-density power electronics, and dynamic load-shedding schemes not just peak MW and thermal limits.

Huang's five-layer framing is a useful mental model for anyone planning capital deployment in this cycle. Each layer is a distinct industry with its own bottlenecks, and he was explicit that the bottleneck migrates over time: two years ago the constraint was chips while energy was adequate; today, the constraint has shifted decisively toward the bottom of the stack land, power, and shell.

Huang gave the audience a physical picture of NVIDIA's current-generation systems: a single rack-scale computer roughly twice the width of a stage, weighing about three tons, containing on the order of 1.5 million parts, costing $4–5 million — with silicon photonics, advanced 3D-packaged memory, and liquid cooling throughout. A single AI factory contains a football field of these racks.

Translate that into electrical engineering terms and the design challenge becomes clear:

Extreme power density. Rack densities of 100–600+ kW push facilities to medium-voltage distribution deep into the white space, direct-to-chip liquid cooling, and in emerging designs, +/-800 VDC or high-voltage DC distribution architectures.
Transmission-level interconnection. Campuses of 300 MW to multi-GW connect at transmission voltage (115–500 kV) and require dedicated substations, often with on-site generation, BESS buffering, and NERC-jurisdictional facilities.
Dynamic behavior. Synchronized training workloads create fast MW-scale load steps; combined with 100% power-electronic interfaces, these loads raise real questions for frequency response, voltage stability, and protection coordination on the bulk system.
Compliance and study requirements. NERC's work on large-load reliability, ERCOT's large-load interconnection framework (including NOGRR282-era ride-through requirements), and evolving RTO/ISO study practices mean large loads are increasingly studied with the same rigor as generation including EMT-level modeling where dynamics matter.

Design Reality Check

An AI factory is best treated as a hybrid facility: part industrial load, part power plant (when co-located generation and BESS are included), and part power-electronics laboratory. The engineering disciplines that utility-scale solar and BESS developers learned over the past decade — EMT/PSCAD modeling, ride-through design, reactive capability planning, protection and control integration at the POI — now apply directly to the load side.

5. Chips and Re-Industrialization: Three Kinds of Plants, Trillions in Capital

Perhaps the most striking statement in the entire conversation, from a grid perspective, was Huang's direct assessment: the U.S. power grid is “a little antiquated,” and AI represents the first opportunity in decades to use market forces — rather than mandates — to fund grid modernization and new generation, including nuclear and other sustainable resources. His logic: for the first time, there are abundant creditworthy customers willing to pay for new power at scale.

He also disclosed where NVIDIA's own capital attention sits: closer to home and nearer on the horizon ensuring that “land, power, and shell” projects in the United States are sufficiently funded, potentially backstopping financings to get “power inserted.” When the world's most valuable chip company says its strategic investment focus is helping energize sites, the message to our industry could not be clearer.

Keentel Engineering sees this playing out across five fronts:

Huang made a forceful economic argument: AI is the United States' best opportunity to re-industrialize. He identified three categories of manufacturing plants driving the build-out — chip plants, computer plants, and AI factories — and estimated the opportunity at several trillion dollars of domestic capital formation, creating hundreds of thousands of jobs over the next four to five years. He credited market forces (anchored by roughly half a trillion dollars of purchase orders steered to suppliers willing to build in the U.S.) with accomplishing what subsidy programs alone had struggled to do.

Every one of those three plant types is an electrical infrastructure project first:

Chip plants. Semiconductor fabs draw 100–500 MW continuously with the tightest power-quality tolerances of any industrial customer. Voltage sags measured in cycles can scrap wafer lots worth millions. Fab interconnections demand redundant transmission supply, static transfer and ride-through design, and rigorous harmonic and flicker studies.
Computer plants. Server and system assembly plants are more conventional industrial loads, but they cluster around fabs and AI factories, compounding regional load growth on distribution and sub-transmission systems that were never planned for it.
AI factories. The AI factories themselves are the anchor loads the gigawatt-class campuses reshaping resource plans in ERCOT, PJM, MISO, the Southeast, and the Desert Southwest.

The strategic corollary Huang drew that America should export energy, chips, infrastructure, models, and applications at every layer reinforces the same conclusion: U.S. competitiveness in AI is now inseparable from U.S. competitiveness in building power infrastructure quickly, safely, and to standard.

6. “AI Is the World's Best Opportunity to Modernize the Power Grid”

One remark deserves special attention from the finance and utility community: Huang observed that in the last three to six months, the gross margins of leading AI companies turned strongly positive. When a product is profitable, the rational response is to make more of it which is why AI companies are now “racing for capacity.” In power-system terms: the load growth is no longer speculative venture-funded demand. It is margin-funded industrial expansion, which historically is the most durable kind.

7. The U.S. Outlook: Speed Is the New Reliability Metric

8. How Keentel Engineering Can Help

Huang closed with an observation about national posture: the United States won the last industrial revolution not because it invented the technology, but because it applied it. His stated fear is not foreign competition but domestic hesitation — that fear of AI slows adoption and forfeits the lead. Whatever one's view of the broader AI debate, the infrastructure version of that argument is already settled fact: the countries and regions that can interconnect large loads and new generation fastest will capture the investment.

That puts uncomfortable but useful pressure on our industry's timelines. Multi-year interconnection queues, serial study processes, and late-stage discovery of network upgrade costs are now national-competitiveness issues, not just developer frustrations. The practical response is the one Keentel has advocated across every market we work in: treat grid interconnection — FERC LGIP processes, NERC MOD/PRC/FAC compliance, RTO-specific large-load rules — as a first-order design input from day one, not a late-stage administrative step. Projects that enter the queue with credible EMT models realistic reactive and ride-through capability, and study-ready data packages move faster, re-study less, and reach commercial operation sooner.

Key Takeaways

(1) Agentic AI multiplied compute demand ~1,000× and shifted computing from retrieval to generation — load growth is structural, not cyclical. (2) The bottleneck has moved to the bottom of the five-layer cake: energy, land, and shell. (3) AI factories, chip fabs, and computer plants are electrical infrastructure projects with generation-grade study requirements. (4) Profitable AI economics mean margin-funded, durable load growth. (5) The U.S. grid modernization opportunity is real — but only for projects engineered for interconnection speed from day one.

Keentel Engineering is a power systems and grid interconnection consulting firm serving utilities, developers, data center operators, and industrial customers nationwide. Our services span the exact intersection of the AI build-out and the grid:

Grid interconnection engineering . Load and generation interconnection studies, POI engineering, feasibility through facilities studies, and queue strategy across ERCOT, PJM, MISO, CAISO, NYISO, and the Southeast.
EMT modeling and studies. PSCAD/EMT model development and benchmarking for inverter-based resources, BESS, and dynamic large loads, including weak-grid screening and ride-through (PRC-029) assessments.
Substation and T-line design. Substation and transmission line design from 30% through IFC, protection and control, grounding system analysis (WinIGS), and arc-flash studies for AI campuses, fabs, and co-located generation.
NERC compliance. NERC O&P compliance support across the MOD, PRC, and FAC standard families for both generation and emerging large-load obligations.
Owner's engineer. Independent ow ner's engineer services for utility-scale solar, BESS, gas, and data center power infrastructure — from due diligence through commissioning.

If your organization is planning an AI factory, a chip fab, co-located generation, or the transmission to serve them, our team can help you engineer the interconnection path before it becomes the critical path. Contact Keentel Engineering at www.keentelengineering.com.

9. Frequently Asked Questions

4. The Validation Continuum: From Offline EMT to Megawatt-Scale Hardware

Disclaimer

This article is published by Keentel Engineering for general informational and educational purposes only. Keentel Engineering is an independent engineering consulting firm and is not affiliated with, endorsed by, or sponsored by NVIDIA Corporation, the Global Conference or its organizers, or any other company, organization, or individual referenced herein. Quotations and paraphrases from the referenced conversation are provided for commentary and analysis. All trademarks, product names, and company names are the property of their respective owners. The content herein does not constitute engineering advice for any specific project; readers should consult a licensed professional engineer regarding their particular facts and circumstances.

WinIGS Grounding Analysis for Utility-Scale Solar Projects

Tue, 07 Jul 2026 13:12:31 GMT

Jul 07, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Grounding Analysis for Utility-Scale Solar Projects with WinIGS: An Integrated Engineering Approach

Our methodology organizes the study around three IEEE standards, and each maps onto specific WinIGS capability.

IEEE Std 80 — design criteria, computed natively.

WinIGS computes permissible touch, step, and metal-to-metal voltages directly from the study inputs: electric shock duration (from fault clearing time — we routinely evaluate backup clearing as well as primary), the soil resistivity model, the insulating surface layer thickness and resistivity, and the X/R ratio at the worst-fault location, from which the program derives the decrement factor for DC offset. The surface-layer reduction factor — crushed rock or asphalt derating — is computed to the IEEE Std 80-2000 formulation (the program retains older-edition options only for legacy-study compatibility). Because these inputs flow from the same integrated model that found the worst fault, the criteria and the stresses are always self-consistent.

IEEE Std 81 — measurement in, verification out.

On the front end, WinIGS ingests field soil data in all the standard forms — Wenner four-pin, Schlumberger, and driven-rod — and automatically estimates a two-layer soil model with statistical rigor: parameter tolerances at a user-set confidence level (90% default), a validity depth telling you how deep the survey can actually see, and graphical bad-data screening where outlier measurements are marked and excluded, visibly tightening the model tolerances. On the back end, WinIGS supports the verification workflow: its Point-to-Point Impedance analysis (Model B or D) predicts what continuity testing of the as-built system should measure, and its Smart Ground Meter integration allows field measurement locations to be imported and interface points generated automatically — closing the loop between model and measurement that IEEE 81 exists to enforce.

IEEE Std 998 — shielding in the same model.

The WinIGS Lightning Shielding Analysis tool implements all three IEEE 998 methods — the full electrogeometric model, the rolling sphere method, and the shielding angle method — operating directly on the same 3-D geometric model used for grounding: buildings, buswork, outdoor equipment, shield wires, and masts are modeled once and analyzed for both fault safety and stroke interception.

Design to 80, verify to 81, shield to 998 — one model, one tool, no re-entry of data between siloed programs.

Step 1: Soil Characterization Across a Heterogeneous Site

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

The Three-Standards Framework in WinIGS: Design, Verify, Shield

A Keentel Engineering Technical Publication

Why Solar Plant Grounding Is Not a Bigger Substation Grid and Why the Software Choice Matters

Step 2: The Integrated System Model in WinIGS

A 100 MW solar plant occupies 500 to 900 acres. A transmission substation occupies five. Yet many grounding studies for utility-scale PV facilities still apply the mental model of the fenced substation yard — design a copper grid, check touch and step voltages inside the fence, stamp it, done. That model fails on solar plants, and it fails in ways that trace directly back to the analysis tool used.

Most grounding software analyzes a grounding system in isolation: the user draws the grid, enters a soil model, assumes a grid current, and the program computes potentials. WinIGS — the Windows-based Integrated Grounding System design program — is built on a fundamentally different premise: the grounding system is modeled together with the power system it belongs to. The network model (sources, transmission lines, transformers, collector circuits) and the geometric grounding model (conductors, rods, piles, fences, foundations) live in one integrated study case, and fault currents, split factors, and ground potential rise emerge from physics rather than from an input field. At Keentel Engineering, this is why WinIGS anchors our grounding practice for utility-scale solar: on a plant this electrically extensive, the assumed-grid-current shortcut is where studies go wrong before the first potential is ever computed.

Three realities of solar plants demand this integrated treatment.

First, the geographic extent changes the physics. A PV plant is not one grounding system; it is thousands of driven steel piles, hundreds of inverter and transformer pads, dozens of miles of medium-voltage collector circuits, a collector substation grid, perimeter fencing, and often a switchyard at the point of interconnection — all electrically interconnected and all embedded in soil that can vary by an order of magnitude across the site. The performance of any one element depends on all the others, and WinIGS's grounding model — built in a full 3-D geometric editor from electrode objects including ground rods, mats, bare and coated horizontal conductors, concrete foundations, and metallic fences — is designed to represent exactly that interconnected reality.

Second, the fault current picture is different. The worst fault for ground potential rise (GPR) at a solar collector substation is very often not a fault at the substation itself but a line-to-ground fault miles away, where the division of fault current between metallic return paths and the earth — the split factor — is least favorable. WinIGS's Maximum GPR analysis automates precisely this: it sweeps fault locations and types across the entire modeled network and identifies the worst fault, reporting its location, type, current, X/R ratio, and the resulting GPR. A study that begins with "assume the grid current is X kA" has already made its most consequential decision by assumption; a WinIGS study makes it by search.

Third, the design objectives multiply: personnel safety per IEEE Std 80 (or IEC 479-1 — both criteria sets are built into WinIGS), a stable reference for protection and communications, lightning protection per IEEE Std 998 via the program's integrated Lightning Shielding Analysis tool, and control of transferred potentials onto fences, pipes, and neighboring infrastructure.

Keentel Engineering's position is that grounding for utility-scale solar must be treated as an integrated, plant-wide power system study — grounding as a first-order design input, not a civil detail closed out at 90% design. WinIGS is the platform on which we execute that philosophy.

Everything rests on the soil model, and on a multi-hundred-acre site the biggest error we encounter in third-party studies is a single soil model extrapolated across the whole plant.

Our practice is to specify Wenner traverses at a density matched to site geology — several distributed across the array area plus dedicated sets at the collector substation and switchyard — with probe spacings pushed as far as logistics allow, because WinIGS's reported validity depth makes the limitation explicit: a survey with spacings to 100 feet supports a model to roughly 150 feet of depth and no further, and for an electrode the size of a solar plant, deep-layer resistivity governs.

Each traverse's data — either measured resistance or apparent resistivity; WinIGS accepts both and cross-computes the other, along with probe length and meter frequency — is processed through the program's soil interpretation engine, which fits the two-layer model and superimposes the fitted curve over the measured points. That plot is where field data quality becomes visible: measurements contaminated by buried metal or instrument error sit conspicuously off the curve, are marked as bad data with the Mark/Unmark function, and the re-run fit reports dramatically tighter parameter tolerances. A defensible soil model is one whose uncertainty is quantified, and this workflow quantifies it.

On sites with genuinely distinct soil regions — river terrace against upland clay, landfill cell against native ground — we model multiple conductor groups and soil regions rather than averaging site geology into a fiction that exists nowhere.

Step 3: Worst-Fault Search and Split Factor the Maximum GPR Analysis

A WinIGS study case has two tightly coupled halves, edited in the program's two specialized environments.

The Network Editor

holds the power system as a single-line diagram: equivalent sources, transmission lines with their physical construction (conductor types and positions, tower geometry from the program's tower library, span length, structure footing resistance, shield wires), transformers including autotransformers with tertiaries, collector circuits, cable systems built from the cable library, and the grounding system elements that tie it all to earth. Our standard scope for a utility-scale PV model: the collector substation grid in full detail; the interconnecting transmission line modeled physically for the first several miles; equivalent sources at interface buses beyond; the 34.5 kV collector system with actual cable construction; representative inverter/transformer pads; the pile and racking network; and the perimeter fence.

Two modeling disciplines deserve emphasis. Equivalents must be verified, not assumed. The external network is reduced to equivalent sources and inter-bus equivalent ci rcuits at the interface buses, with parameters taken from the utility's short-circuit model (ASPEN, CAPE, or similar). The acceptance test is non-negotiable: the reduced WinIGS model must reproduce the utility's fault duties at the POI before any grounding conclusion is drawn. Exact match is not expected — WinIGS's physically based three-phase models capture asymmetries that classical short-circuit programs ignore — but the discrepancy must be quantified and small. Everything needs a path to remote earth, and everything conductive is part of the answer. WinIGS enforces the first as a modeling rule (a floating delta tertiary will stop the solver, for good reason); the second is engineering judgment the tool supports: reinforced-concrete foundations are modeled as steel mat electrodes on a dedicated Foundations layer — capturing their Ufer-effect contribution while WinIGS's layer system keeps them out of the auto-generated Bill of Materials, since they're structure, not purchased copper.

The Ground Editor

holds the physical grounding model in true 3-D — top, side, perspective, and rendered views — with the site's foundation drawing imported as a scaled background image (DXF import/export is supported for CAD interchange), so grounding conductors are placed against real equipment locations, not idealized rectangles. Analysis model fidelity is selectable: Model A for rapid equipotential-conductor screening through Models B, C, and D, which represent voltage drop along conductor lengths — with the enhanced Model C in current releases capturing ground-conductor self-inductance at near-Model-D accuracy for a fraction of the runtime, a meaningful advantage on plant-scale models with thousands of electrode segments.

With the model built, WinIGS's Maximum GPR analysis runs a large automated series of fault simulations — every fault type, swept along every line and bus within a user-set radius (or the whole system) — and identifies the fault producing maximum GPR at the monitored node. The report is the design basis in one screen: worst fault type and location, fault current magnitude and phase, X/R at the fault, and the GPR.

Three phenomena make this search essential on solar plants. The split factor — reported directly in WinIGS's Ground Resistance report as the ratio of earth current to fault current — varies strongly with fault location, and the worst case is routinely a fault out on the line, not at the station. Transformer winding effects are captured physically: WinIGS's Internal I/O reporting shows the circulating current in a delta tertiary during ground faults — a current that can exceed 10 kA while the winding's terminal currents read zero, and that raises GPR for remote faults while lowering it for local ones. And on the collector side, the grounding-source configuration for the 34.5 kV system shapes touch voltage at every inverter pad in the plant.

Typical well-designed collector substations on effectively grounded transmission systems see split factors in the 20–40% range. An isolated-grid study assuming 100% of fault current enters the earth over-designs by a factor of three to five; one that guesses low gambles with compliance. WinIGS replaces the guess with a computation.

Step 4: Safety Assessment — Touch, Step, and Transferred Potentials

Step 5: Design Enhancement — Where the Money Is

Against the worst-fault condition, WinIGS computes earth-surface potentials and evaluates touch and step voltages over user-defined assessment regions rectangular or polygonal plot frames drawn directly on the grounding model. Region definition is engineering judgment the tool makes explicit: touch-voltage frames cover everywhere a person can stand and reach grounded metal — substation interior plus a band extending three feet outside the fence, every inverter pad and enclosure row; step-voltage frames emphasize the areas just outside perimeter fences where gradients peak; and each frame is assigned its own permissible basis crushed-rock derating inside the station, native-soil criteria in the arrays and beyond the fence.

The results read as color-coded equipotential contour plots with the maximum location flagged and the permissible value printed alongside compliance or violation at a glance plus a rendered 3-D touch-voltage surface where violations rise as literal red peaks above the plant. Those peaks are a design instruction: they show precisely where conductor must be added, which is a far more economical process than uniformly densifying the grid. Two referencing details matter for correctness at plant scale, and WinIGS handles both: touch voltage referenced to the nearest grounding point (the recommended setting for Models B/C/D, where conductor voltage drop is represented) rather than a single remote reference; and step voltage computed across the standard 3-foot stride distance.

Transferred potential closes the assessment. WinIGS's user-defined contour feature draws the zone-of-influence boundary commonly the 300 V contour directly on the site plot, making visible which pipelines, fences, communication routes, and neighboring facilities fall inside it during the worst fault and therefore require isolation, gradient control, or dedicated mitigation.

Frequently Asked Questions

The study is complete when the as-built system is measured and reconciled. Our commissioning scope: fall-of-potential impedance measurement with leads long enough to escape the zone of influence WinIGS plotted; plant-wide continuity verification, with WinIGS's Point-to-Point Impedance report generating the predicted value for every test-point pair — the complete pairwise report exports to file for direct comparison against field data; and investigation of every divergence before energization. A measured point-to-point impedance that disagrees with the model is not noise — it is a missing bond, a broken conductor, or a soil surprise, found while it is still cheap to fix.

WinIGS is a proprietary software product of its respective developer and owner. Keentel Engineering is an independent consulting engineering firm and is not affiliated with, endorsed by, or sponsored by the developers of WinIGS or any other software vendor or standards organization referenced herein. IEEE Std 80, IEEE Std 81, and IEEE Std 998 are publications of the Institute of Electrical and Electronics Engineers. All product names and trademarks are the property of their respective owners.

Case Study 1: 120 MW PV Plant — Three Bonding Strategies Settled by One WinIGS Model

Situation

A utility-scale PV facility in the southeastern U.S. — roughly 120 MW across approximately 700 acres, 34.5 kV collection, 115 kV interconnection — engaged Keentel to resolve a design dispute between the EPC and the owner's independent engineer over array bonding philosophy: (A) isolated array grounding, each inverter block grounded locally and deliberately not bonded to the collector substation; (B) fully integrated bonding tying racking, pads, and station into one continuous system; (C) a hybrid integrating the collector grounding while insulating fence sections remote from the substation.

Approach

Rather than argue philosophy, we built one WinIGS integrated study case and ran all three schemes as variants of it. The Network Editor model carried the 115 kV interconnection with shield wires and structure grounds, equivalent sources tuned until the reduced model reproduced the utility's fault duties at the POI, the full collector system with actual cable construction, and the BESS-ready station arrangement; the Ground Editor model carried the station grid, pads, pile network, and fencing over the imported site drawing. Fourteen Wenner traverses were processed through the soil interpretation engine — site resistivity ranged from about 90 Ω·m over 30 Ω·m in the bottomland to over 400 Ω·m in the upland third, so two soil regions were modeled with separate conductor groups rather than averaged. The Maximum GPR analysis was run for each variant against both transmission and collector systems.

Findings

The isolated scheme (A) — defended as preventing station GPR from "exporting" into the arrays — performed worst. The model showed the isolation to be largely fictional (cable neutrals and incidental paths defeat it), while depriving inverter pads of the gradient control the integrated network provides: during the worst collector-side fault, the touch-voltage plot frames at three pads showed exceedances over 40% above the native-soil permissible. The integrated scheme (B) cleared the pad violations and cut station impedance by roughly a third — the pile network's contribution, now visible in the Ground Resistance report — but the transmission-side worst fault transferred elevated potential onto remote fence runs, and the fence-line frames lit up nearly half a mile from the station. The hybrid (C) met criteria everywhere: integrated bonding delivered the pad safety and impedance benefit, insulated fence breaks with local gradient loops handled the remote transfer.

Outcome

Scheme C became the grounding design basis. The quantified pile contribution supported a reduction of the station grid from the EPC's preliminary layout, offsetting most of the fence-mitigation cost. At commissioning, point-to-point continuity results matched WinIGS predictions across the test-point set, and the fall-of-potential measurement of the completed system landed within 12% of the model's computed impedance.

Case Study 2: High-Resistivity Mountain Site — WinIGS Split-Factor Analysis Beats a Copper Crisis

Situation

A 150 MW solar project in mountainous western-U.S. terrain hit a grounding crisis at 60% design: shallow soil over rock, survey models showing 600–1,100 Ω·m upper layers, and an EPC study — an isolated-grid analysis with an assumed grid current — concluding that compliance required massive grid densification plus deep wells at a cost the project could not absorb.

Approach

Keentel rebuilt the problem in WinIGS as an integrated study. The 230 kV interconnection was modeled physically for the first several miles — both shield wires, tower geometry from the tower library, measured footing resistances — with the external network reduced to equivalent sources and inter-bus equivalents verified against the utility's short-circuit duties at the POI. The Maximum GPR analysis, searching all fault locations, placed the governing fault on the 230 kV line roughly a mile and a half from the plant — and the Ground Resistance report put the split factor near 25%: only about a quarter of the fault current actually entered the earth at the station. The EPC study's assumed grid current had been more than triple physical reality.

Findings

With the computed earth current, the crisis shrank but did not vanish: touch-voltage frames along one switchyard edge still showed exceedances around 20%, and the station GPR kept a fiber route and a pipeline crossing inside the 300 V zone-of-influence contour WinIGS plotted on the site plan. Two mitigations were engineered in the same model. Grid enhancement needed on the order of 1,300 additional feet of 4/0 in rock trench — and the re-run showed GPR essentially untouched, leaving the transferred-potential exposures alive. The counterpoise alternative used the Model Conversion tool to convert the first two spans to the mutually coupled multiphase line model, adding a buried counterpoise conductor bonded to structure grounds and the station grid through a corridor of deeper soil. Result: station GPR down approximately 25%, every touch frame under permissibles with better than 10% margin, the 300 V contour pulled inside the pipeline crossing — at under two-thirds the conductor length of the copper-only option, in trenchable soil.

Outcome

The counterpoise design was built. Total grounding cost came in below half the EPC's original mitigation estimate, and the study package — worst-fault basis, split-factor derivation, soil model with confidence tolerances, IEEE 81 verification plan — cleared utility review on first submittal. On hard-soil sites, the cheapest copper is the copper the split-factor computation proves you don't need.

Case Study 3: Solar-Plus-Storage Collector Station — the Worst Fault Wasn't Where Anyone Looked

Situation

A solar-plus-storage facility — approximately 200 MW PV with a 100 MW / 400 MWh BESS sharing the collector substation — engaged Keentel after the owner's engineer flagged that the preliminary grounding study had evaluated only station bus faults. The station was electrically busy: two 34.5 kV collector buses, a BESS medium-voltage system with its own grounding transformer, a main transformer with a buried delta tertiary, and a 138 kV interconnection.

Approach. The WinIGS model carried the full station, the 138 kV line with shield wires, and both feeder systems with their distinct grounding sources; the delta tertiary was modeled explicitly (with its required path to remote earth). The Maximum GPR analysis was configured with the search radius at zero — every circuit, transmission and medium-voltage alike — and the tertiary's behavior was examined via Internal I/O reports under both remote and local fault cases.

Findings

The governing fault was a line-to-ground fault on a 34.5 kV BESS feeder a short distance from the station — not the 138 kV bus fault the preliminary study assumed conservative. Two effects an isolated study cannot see drove it. The BESS grounding transformer fed ground-fault current returning through station earth for close-in feeder faults, over a cable system whose metallic return was weaker than the transmission side's shield wires — an unfavorable split, visible directly in the split-factor output. And the delta tertiary, inert at its terminals, circulated over 10 kA in the Internal I/O report during the feeder fault, adding to the earth current for that case while reducing GPR for the bus-fault case — the exact inversion of the preliminary study's assumption. Touch-voltage plot frames around the BESS enclosure rows, evaluated against the feeder fault with the decrement factor from that location's X/R, showed exceedances at eleven enclosure positions.

Outcome

Mitigation was surgical: gradient-control loops around the affected rows, a revised neutral-grounding-resistor specification for the BESS grounding transformer that trimmed fault contribution without degrading protection sensitivity, and one relocated feeder ground bond roughly 900 feet of added conductor per the final Bill of Materials, versus the several-thousand-foot blanket densification a bus-fault-only reading would have driven. At commissioning, WinIGS's exported point-to-point impedance predictions exposed one unbonded BESS enclosure row construction had missed — the measured value disagreed with the model, and the model was right. The station has since cleared two real collector-system ground faults without incident.

Case Study

Power System Resilience Metrics: Engineering Guide & Methods

Tue, 07 Jul 2026 11:28:26 GMT

Jul 07, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Measuring What Matters

1. Why Resilience Metrics, and Why Now

For most of the past century, the power industry measured itself with a small, stable family of reliability indices. SAIDI, SAIFI, and CAIDI told regulators how the average customer fared over the average year. LOLE and EENS told planners whether the resource fleet could cover expected demand. These indices did their job well for the grid they were designed to describe: a grid dominated by dispatchable generation, statistically independent component failures, and weather that behaved roughly the way it always had.

That grid no longer exists. Extreme weather events are striking with greater frequency and severity, cyber and cyber-physical threats against utilities and grid assets are escalating, infrastructure is aging, and the resource mix is shifting rapidly toward weather-dependent, inverter-based generation. At the same time, load itself is transforming data centers, electrification, and large industrial additions are concentrating enormous demand behind single points of interconnection. The events that now define grid performance are not the routine, uncorrelated faults that reliability indices average away. They are high-impact, low-probability (HILP) events: hurricanes, ice storms, wildfires, extended heat domes, coordinated attacks, and cascading failures that take out hundreds of components at once and keep customers dark for days or weeks.

Reliability metrics were never built to see these events clearly. In fact, standard practice often explicitly excludes them major event days are routinely removed from SAIDI reporting precisely because they distort year-to-year comparisons. The result is a measurement blind spot exactly where the stakes are highest. Resilience metrics exist to close that blind spot: to quantify how deeply a system degrades under extreme stress, how quickly it recovers, what the disruption actually costs customers and the economy, and critically whether a proposed hardening or flexibility investment is worth the money.

This article lays out a practical, engineering-grade framework for resilience measurement: what resilience means, how it differs from reliability, how metrics should be classified and selected, how real utility outage data is converted into defensible resilience numbers, how heavy-tailed blackout risk changes the math, and how resilience value is monetized for regulatory filings and investment decisions. Throughout, we approach the problem the way Keentel Engineering approaches every grid problem: resilience, like interconnection, is a first-order design input — not a downstream compliance formality.

Keentel Perspective:

A resilience metric is only useful if it can be computed from data you actually have, moves when you invest in the system, and survives cross-examination in front of a regulator. Metrics that fail any of those three tests are decoration. Everything in this framework is filtered through that standard.

2. Where Traditional Reliability Metrics Fall Short

Reliability indices remain indispensable for day-to-day planning and performance regulation, and nothing in a resilience program replaces them. But they are structurally incapable of characterizing performance during extreme events, for seven distinct reasons:

Built for expected conditions. Reliability standards assume failures are independent, statistically consistent, and well represented by historical averages. Extreme events violate every one of those assumptions, producing correlated, cascading, large-scale failures that exceed design limits.
No view of degradation or recovery. Reliability indices are long-term averages. They say nothing about how fast a system degrades, how deep the service loss goes, or how quickly restoration proceeds during and after a specific event.
Human and organizational factors invisible. Extreme events strain crews, communications, and decision-making. System performance during a hurricane is a function of organizational resilience and staffing as much as hardware — dimensions reliability indices simply do not carry.
Blind to cyber-physical and multi-infrastructure threats. Standard indices focus on physical component outages and ignore cyberattacks, communication failures, fuel supply disruption, interdependent infrastructure failures, and distributed resource variability.
Not designed to guide hardening. SAIDI improvements do not tell you where to underground, which substations to elevate, where microgrids belong, or how much flexible capacity to hold. Utilities need metrics that connect directly to actionable investments.
Equity and critical loads averaged away. Reliability indices treat all customers identically, masking impacts on hospitals, emergency services, water systems, and vulnerable populations exactly the loads a resilience program exists to protect.
Misaligned with non-stationary threats. Reliability frameworks rest on stationarity the assumption that the future statistically resembles the past. Climate-driven weather and evolving cyber threats are fundamentally non-stationary, so backward-looking indices cannot support forward-looking resilience planning.

The conclusion is not that reliability metrics should be discarded they remain the baseline expectation of performance under normal conditions. The conclusion is that a second, complementary class of metrics is required to characterize robustness, adaptability, and recovery under abnormal conditions. That is the entire project of resilience measurement.

3. Defining Resilience: The Event Lifecycle

Across the definitions advanced by standards bodies, system operators, national laboratories, and regulators, a consistent core emerges: resilience is the ability of a power system to anticipate, absorb, adapt to, and rapidly recover from extraordinary, high-impact events including natural hazards, equipment failures, accidents, and deliberate physical or cyber attacks while limiting the extent, severity, and duration of degradation and sustaining critical services.

Two features of this definition deserve emphasis, because they drive everything about metric design. First, resilience is explicitly multi-phase. A resilience event unfolds through anticipation and preparation (risk identification, hardening, contingency planning before the event), absorption (withstanding the initial shock while maintaining at least partial operation), adaptation and sustainment (dynamic response, reconfiguration, and critical-load prioritization during prolonged disruption), and recovery (efficient restoration of service and infrastructure to pre-event or improved condition). A meaningful metric set must cover the full lifecycle — a system that absorbs well but recovers slowly, and a system that fails hard but restores fast, are different engineering problems requiring different investments.

Second, resilience admits degraded operation. Reliability compliance is binary: criteria are satisfied or violated. Resilience explicitly accepts that an extreme event may temporarily compromise normal operation, and instead asks whether degradation was controlled, proportionate to the event, and reversed quickly. This is a fundamentally different acceptability philosophy, and it is why resilience cannot simply be bolted onto existing reliability criteria.

4. Resilience vs. Reliability: A Working Comparison

The practical distinctions between adequacy, security, and resilience determine which analytical tools, contingency sets, and acceptance criteria apply to a given study. The comparison below summarizes the differences that matter in practice.

One doctrinal point is worth carrying into every study: security requires the integrity of conventional load service, while resilience does not. Loads under interruption or demand-response contracts are a security resource; involuntary loss of conventional load during an extreme event is a resilience impact to be measured, bounded, and priced. Keeping those categories straight prevents a common analytical error counting contracted flexibility as a resilience failure, or worse, counting involuntary shedding as acceptable flexibility.

There is also a productive middle path: existing reliability metrics can be conditioned to capture resilience attributes. Expected load curtailment, for example, becomes a resilience metric when the expectation is restricted to events exceeding a severity threshold — a duration beyond 24 hours, an event with more than a set number of simultaneous outages, or costs above a defined level. Threshold-conditioned reliability metrics are often the fastest path to a defensible resilience baseline, because they reuse data pipelines and institutional familiarity that already exist.

5. A Taxonomy of Resilience Metrics

Resilience cannot be condensed to a single number. A workable program classifies metrics along several axes and selects deliberately from each, aligned with the decisions the metrics are meant to inform.

5.1 Vulnerability Modeling: Model-Based vs. Statistical

Every resilience assessment rests on a model of how components and systems fail under stress. Two complementary approaches exist. Th e model-based (analytical) approach builds fragility relationships for individual components the conditional probability that a line , structure, or insulator fails as a function of wind speed, ice accretion, flood depth, or other hazard intensity and composes them into system response. Its strength is direct investment traceability: once the model is calibrated, a proposed hardening measure changes a fragility curve, and the resilience benefit can be computed. Its weakness is the sheer heterogeneity of real systems ages, designs, exposures, and failure modes vary so widely that system-level calibration and validatio n are genuinely hard.

The statistical approach works from the other direction: it relates recorded utility outage data to recorded hazard intensity, producing area outage rate curves that describe the system's observed average outage rate as a function of measured stress such as wind speed. These curves capture the system's overall response all component types, ages, and exposures blended together which is exactly what makes them powerful for benchmarking and exactly what limits them for prediction: because they encode the past, they cannot by themselves describe a future that differs from past experience, whether through climate change or through the very hardening investment being evaluated. Mature programs use both: statistical curves to establish the empirical baseline and validate models, analytical fragility models to project the effect of proposed changes.

5.2 Planning vs. Operational Metrics

Planning metrics evaluate a system's inherent capability before an event: performance under predefined damage scenarios, redundancy and robustness attributes, multi-criteria scores weighting robustness, redundancy, and recoverability, and risk-based cost-benefit indicators used to select portfolios of hardening and operational measures. Operational metrics evaluate performance in and after real events: real-time degradation tracking, restoration progress, and post-event assessment against planning assumptions. The most effective frameworks close the loop event data validates planning models, exposes gaps, and recalibrates the next planning cycle. A resilience program that only plans, or only reports, is half a program.

5.3 Transmission vs. Distribution

Transmission and distribution demand different metrics because they differ in topology, control hierarchy, failure physics, and data. Transmission resilience is characterized by energy not served in MWh, stability margins, N-k withstand capability, topological criticality, and cascading-failure exposure, measured largely through SCADA and synchrophasor data. Distribution resilience is customer-facing: customers interrupted, customer-hours lost, restoration rates, critical-load continuity, feeder reconfiguration and islanding capability, measured through outage management systems and AMI. Because impacts propagate across the interface transmission events strand distribution restoration, and distribution DERs increasingly support transmission recovery metric frameworks should align both levels on shared dimensions such as restoration duration, robustness, and recovery slope so that results remain comparable across the hierarchy.

5.4 Grid vs. Community and End-User Perspectives

System-level metrics total customers affected, aggregate energy not supplied, time to restore a defined fraction of regional load describe the interconnected grid. Community-scale metrics describe microgrids and islandable systems: sustainable islanding duration, fraction of community demand met by local resources, voltage and frequency stability in islanded mode, and the speed and reliability of transitions between grid-connected and islanded operation. And beneath both sits the end-user perspective, which reframes resilience in terms customers actually experience: individual interruption duration, continuity of critical services, predictability of restoration estimates, and economic losses by customer segment. Regulators are increasingly explicit that resilience value must ultimately be expressed in customer terms one practical formulation indexes each candidate grid intervention by the ratio of disconnected users to the return period of the initiating event, a compact way to rank investments by customer risk reduction per event likelihood.

6. From Curves to Numbers: Practical Metrics from Real Outage Data

The most defensible resilience metrics are computed directly from data utilities already record: time-stamped outage and restoration events with minute-level resolution, cause codes, and on distribution systems customer counts. Converting that raw data into resilience metrics follows a disciplined pipeline.

6.1 Grouping Outages into Resilience Events

Resilience is a property of events, not of isolated outages. The first processing step groups individual forced outages into events based on temporal clustering of outage start times and overlap of outage durations. Event sizes span an enormous range most events are single outages; the rare extreme events involve hundreds of simultaneous outages from severe weather or cascading phenomena. It is precisely this event-level structure, invisible to annual averages, that resilience analysis exists to capture.

6.2 The Outage, Restore, and Performance Processes

For each event, define the outage process O(t) as the cumulative number of outages by time t and the restore process R(t) as the cumulative number of restorations. The performance curve P(t) = R(t) − O(t) tracks the negative of unrestored outages, decrementing at each outage and incrementing at each restore. The idealized resilience trapezoid degradation phase, sustained nadir, then recovery phase is the textbook picture, but real utility data almost never separates so cleanly: outage and restore processes overlap in time, heavily so on distribution systems. The decomposition of the performance curve into its outage and restore processes generalizes the trapezoid, works on real data, and preserves every trapezoid metric without assuming restoration waits for damage to finish.

The vertical axis is a design choice. Track component counts for asset-centric analysis, MVA ratings for transmission capability, or customer counts for distribution impact in which case the area between the performance curve and the time axis is exactly the customer-hours lost in the event, and the nadir is the maximum simultaneous customers out.

6.3 The Core Metric Set

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

PART I — TECHNICAL BLOG: Measuring What Matters in Grid Resilience

A Practical Framework for Power System Resilience Metrics, Evaluation Methods, and Investment Valuation

Keentel Engineering Technical Blog • Engineering FAQ • Anonymized Case Studies • Service Capabilities

Notice and Disclaimer

Case Study

This publication is original technical content developed and copyrighted by Keentel Engineering. It reflects Keentel Engineering's independent engineering perspective on power system resilience metrics, evaluation methods, and investment valuation, informed by our project experience in transmission and distribution planning, grid interconnection, utility-scale renewables and energy storage, and reliability compliance engineering.

Keentel Engineering is an independent consulting engineering firm. This document is not affiliated with, sponsored by, or endorsed by any standards development organization, regional transmission organization, independent system operator, reliability coordinator, government agency, national laboratory, or equipment manufacturer. All trademarks, registered marks, program names, and product names referenced in this publication, if any, are the property of their respective owners and are used solely for identification purposes.

The case studies presented in this document are anonymized and generalized composites drawn from engineering practice. Client names, project names, locations, counterparties, and commercially sensitive parameters have been altered or omitted. Any resemblance to a specific project or organization is coincidental.

This document is provided for general technical information only and does not constitute engineering services, legal advice, or a professional engineering deliverable for any specific project. Resilience metric selection, valuation, and investment decisions must be evaluated by qualified professionals against project-specific data, applicable reliability standards, regulatory requirements, and contractual obligations.

The crew metrics deserve special mention because they close the gap between engineering and operations. Tracking deployed crews C(t) hour-by-hour through a restoration yields total crew-hours; dividing customer-hours lost by crew-hours gives restoration efficiency, and normalizing customer-hours by customers affected gives the restoration experience of the average outaged customer. A logarithmic composite of average customer restoration duration and crew-hours per outage the emergency response efficiency compresses these into a single comparable score across events of very different size. Lower is better, and trends in this metric across storm seasons are a direct, data-driven readout of whether emergency response investments are working.

7. The Heavy-Tail Problem: Why Blackout Risk Breaks Ordinary Statistics

Convert each event's customer-hours to direct customer cost using a cost-per-customer-hour factor appropriate to the service territory's load mix, and assemble every event over several years into an empirical exceedance curve the probability that an event's cost exceeds a given value, plotted on log-log axes. The small-cost region of that curve describes routine risk. The large-cost region describes resilience risk, and it routinely exhibits a property that changes the mathematics: heavy tails, with log-log slope magnitudes below one observed on multiple real distribution systems.

A heavy tail has three brutal implications. First, there is no representative large blackout the tail keeps producing larger events, so planning to a single design-basis event systematically understates risk. Second, the risk contributed by large events dominates the risk from all the moderate events combined. Third and this is the one that invalidates familiar tools sample means over the tail do not converge with any realistic amount of data, so metrics and methods that depend on evaluating a mean, including conditional value-at-risk and expected-value optimization, simply cannot be applied to large-blackout cost.

The practical remedy is to work in log space. Taking the logarithm of large-event costs converts heavy-tailed data into light-tailed data on which ordinary statistics behave. Averaging the log-costs of large events yields a stable severity metric; multiplying by the annual frequency of large-cost events yields an annualized log-cost resilience index that is stable year over year, sensitive to genuine changes in system resilience, and unlike a tail expectation actually estimable from the data a utility has. For risk statements, complement this with exceedance-based metrics: the probability that an event exceeds a defined large-cost threshold, the annual frequency of such events, and their recurrence interval.

Engineering Judgment:

If a consultant hands you a single expected-value dollar figure for large-blackout risk, ask how it was estimated. On heavy-tailed systems that number is an artifact of the sample, not a property of the system and an investment case built on it will not survive technical review. Exceedance metrics and log-domain indices are the defensible alternative.

8. Putting a Dollar Value on Resilience

Resilience investments compete for capital against everything else a utility or developer could build. Winning that competition in front of a board, a commission, or an interconnection counterparty requires converting resilience performance into money. Three methodological families carry most of the weight.

8.1 Probabilistic Risk Assessment

PRA follows the fundamental equation risk = probability × consequence: characterize hazards and their frequencies, assess component vulnerabilities, analyze consequences across failure modes, and rank risks. Executed across thousands of simulated scenarios, PRA gives decision-makers a quantitative map of failure modes and mitigation effectiveness. Its cost is data intensity especially for rare events and the specialized expertise needed to build and defend the scenario set.

8.2 Cost-Benefit Analysis and the Value of Lost Load

CBA compares the present value of resilience investment costs capital, O&M, opportunity cost against the present value of avoided costs: avoided infrastructure damage and repair, avoided utility revenue loss, avoided customer outage cost, and avoided broader economic and social impact. The customer term is typically monetized through the Value of Lost Load, and here a critical caution applies: VoLL parameters calibrated for reliability analysis describe interruptions of minutes to a day. Resilience events last days to weeks, and outage cost is strongly nonlinear in duration spoilage, equipment damage, business failure, and public-safety costs accumulate in regimes short-duration VoLL never sampled. Customer damage functions that express outage cost as a function of duration, by facility type, are the better instrument: they reveal, for example, that a battery covering short outages delivers major value to a facility whose costs front-load at interruption onset, and almost none to a facility whose costs begin when refrigeration fails four hours in.

8.3 Resilience-Adjusted Resource Adequacy

Traditional effective load carrying capability measures a resource's contribution to adequacy under normal conditions and misses how resources perform when it matters most. The resilience-adjusted extension evaluates a resource across a probability-weighted set of disruptive event scenarios, each characterized by intensity, duration, and restoration trajectory. For each scenario, compute expected energy not served with and without the resource; the probability-weighted fractional reduction in unserved energy is the resource's resilience-adjusted capacity value. Multiplying that value by the applicable VoLL and representative event duration converts it directly into avoided outage cost a single, auditable bridge from technical adequacy modeling to monetary benefit that lets grid hardening, dispatchable generation, storage, and DERs compete in one framework.

As a scale illustration: a distribution-connected battery energy storage system that reduces modeled storm-event unserved energy by thirty percent, valued at a defensible multi-day VoLL over a representative event duration, produces an avoided-cost figure that can anchor a rate case or an interconnection benefit study. The arithmetic is simple; the engineering content lives in the scenario set, the degradation and restoration modeling, and the VoLL selection which is exactly where these analyses are won or lost.

8.4 Macroeconomic Impact and Composite Indices

For regional and national decisions, input-output and computable general equilibrium models trace outage impacts through inter-sector dependencies to GDP, employment, and sectoral output capturing cascading economic effects that customer-level VoLL misses, at the price of heavy data and modeling requirements and limited ability to disaggregate to specific customer classes. Composite resilience indices resilience triangles and trapezoids, multi-dimensional scores blending performance, consequence, and attribute measures condense multidimensional resilience into single figures useful for tracking and comparison. Use them for communication and trend monitoring; do not let a composite substitute for the underlying independent metrics, and where aggregation is unavoidable, weight components dynamically by event type and contingency level rather than freezing a single weighting for all hazards.

9. Data Foundations

Every metric above is only as good as its data. A serious resilience program inventories and integrates several data classes: utility outage records (time-stamped outage and restoration events with cause codes for transmission and distribution), national and regional outage aggregation platforms and regulatory event reporting, standardized transmission and generation availability databases, detailed weather station observations, numerical weather prediction and extreme-event reanalysis products, and critical-infrastructure geospatial layers. Distribution programs add outage management system and AMI data; transmission programs add SCADA and synchrophasor archives.

Data quality determines credibility. Cleansing workflows must reconcile timestamps across systems, filter planned outages from forced outages, de-duplicate cascaded records, validate customer counts, and align hazard observations spatially and temporally with outage records. Privacy and security constraints customer-level data, critical infrastructure information must be engineered into the pipeline from the start, not patched in before a filing. In our experience, data preparation is sixty to seventy percent of the effort in a first-time resilience benchmarking engagement, and it is effort that pays permanent dividends: once the event-extraction pipeline exists, every subsequent storm season updates the metrics nearly for free.

10. What a Credible Resilience Framework Must Include

Surveying the existing landscape of resilience tools and frameworks community-resilience GIS platforms, adaptation toolkits, investment rating systems, federal grant metrics frameworks, and laboratory assessment methodologies reveals consistent gaps when they are applied to power systems: limited treatment of cross-sector interdependencies, weak quantitative handling of equity and vulnerable populations, static indicators that cannot track evolving threats, insufficient representation of power system technical complexity, and a general shortage of quantitative, event-driven metrics grounded in operational data. A framework fit for purpose must therefore encompass:

Grid performance evaluation. Dynamic performance characterization across the full event lifecycle pre-event preparedness indicators, event-phase degradation profiles, and post-event recovery metrics supporting both pre-event estimation and post-event assessment.
Dynamic and adaptive metrics. Metrics that update with system conditions and threat evolution, scale from feeders to bulk systems, and normalize across urban and rural contexts for meaningful comparison.
Cross-sector interdependencies. Explicit quantification of cascading effects between the power system and water, transportation, communications, and fuel infrastructure.
Socio-equity dimensions. Metrics capturing disproportionate outage impacts on disadvantaged and vulnerable populations, integrated into planning and investment prioritization rather than appended to it.
Power-system technical complexity. Renewable and inverter-based resource behavior, protection performance, cyber-physic al threat surfaces, and hardening measure effectiveness represented at engineering depth.

Institutionally, resilience must be integrated across three time horizons: long-term planning (climate-informed vulnerability assessment of future grid configurations, flexible and DER-ready designs, redundant communications), operational planning (dynamic contingency plans, forecasting and early-warning systems, scenario stress-testing, pre-positioned resources), and real-time operation (situational awareness, adaptive protection, coordinated transmission-distribution restoration). Utilities, large industrial and data-center loads, governments, and standards bodies each carry distinct responsibilities in that structure and the interconnection interface between them is where resilience requirements most often get lost.

11. The Keentel View: Resilience Is an Interconnection-Grade Design Input

Keentel Engineering's core conviction is that grid interconnection is a first-order design input, not a downstream utility formality. Resilience belongs in exactly the same category. The moment a project is conceived a utility-scale solar plant, a battery system, a data center, a transmission line, a substation its contribution to and dependence on system resilience is being determined by design choices: point of interconnection selection, redundancy and sectionalizing architecture, ride-through and protection settings, islanding capability, black-start and restoration roles, structural hardening class, and the contractual definition of critical load.

Treated early, these choices cost little and compound. Treated late, they surface as interconnection restudies, compliance findings, storm-season failures, and rate-case rejections. The metrics framework in this article is how the early treatment becomes rigorous: it gives developers, utilities, and large loads a shared quantitative language for what resilience is worth, what it costs, and which investments actually move it.

Case Study 2 — Transmission Corridor Hardening: A Regulator-Ready Cost-Benefit Case Built from Recorded Outage Data

Background

A transmission owner operating several hundred miles of 230 kV and 138 kV lines through terrain exposed to severe wind and wet-snow icing faced a familiar dilemma: engineering judgment said certain corridors needed structural hardening and anti-icing measures, but the capital request had twice failed to clear internal investment review because the benefit case rested on qualitative risk language rather than quantified resilience improvement.

The Challenge

The owner held fifteen years of minute-resolution forced-outage records and regional weather station data, but had never converted them into event-level resilience metrics. Candidate investments — structure replacement on two corridors, anti-torsional devices to prevent wet-snow accretion, and an expanded emergency spare-structure program competed for a fixed budget with no common quantitative yardstick. And because the heaviest historical events dominated total impact, ordinary expected-value analysis was unstable: the benefit estimate swung wildly depending on whether one or two extreme winters were included in the sample.

Keentel's Approach

Event metrics from recorded data. Grouped fifteen years of forced outages into resilience events by start-time clustering and duration overlap; computed event size, MVA-tracked performance curve areas, nadirs, and time-to-95%-restoration for every event, deliberately avoiding full-restoration times whose final stragglers made statistics unusable.
Statistical vulnerability baseline. Constructed area outage rate curves relating observed corridor outage rates to measured wind speed and icing-proxy conditions, establishing the empirical stress-response baseline of the system as built.
Analytical fragility modeling. Developed component fragility models for existing and hardened structure classes and for conductors with and without anti-torsional devices, expressing each candidate investment as a quantified change in failure probability versus hazard intensity — including the increase in failure return period delivered by each measure.
Rerun-history benefit quantification. Applied the engineering-predicted outage-rate reductions retrospectively to the recorded event history — rerunning history — to compute how each historical event's MVA-hours, restoration percentile times, and customer impact would have changed had each investment been in place. Presented corridor-by-corridor, this became the emotional and analytical core of the filing: benefits expressed against storms the service territory remembered.
Heavy-tail-aware portfolio optimization. Characterized large-event cost risk with exceedance curves; the observed tail was heavy enough that expected-value optimization was formally invalid, so the portfolio was optimized against exceedance-frequency reduction and log-domain severity metrics instead, with a risk-based selection of the investment mix across hardening and restoration-improvement measures.

Outcome

The recommended portfolio full hardening on one corridor, anti-icing devices on two others, and a doubled emergency spare-structure and crew pre-staging program in place of the second corridor rebuild delivered approximately 85% of the maximum modeled resilience benefit at 60% of the original capital request. Rerun-history analysis showed the portfolio would have reduced the worst recorded winter event's unsupplied MVA-hours by roughly half and cut time-to-95%-restoration from eleven days to five. The investment cleared internal review on first submission and anchored the owner's subsequent resilience plan filing, with the event-metrics pipeline retained as the standing measurement basis for post-investment verification.

Case Study 1 Resilience Valuation of a Utility-Scale BESS for a Storm-Exposed Coastal Utility

Background

A coastal investor-owned utility serving a mixed residential, commercial, and light-industrial territory faced recurring multi-day distribution outages from named-storm events. The utility had contracted a 120 MW / 480 MWh battery energy storage system connected at a 115 kV substation serving several vulnerable feeders, justified initially on energy arbitrage and capacity value. Ahead of a rate proceeding, the utility needed to quantify defensibly the resilience value the storage asset contributed, so that resilience benefits could be represented alongside conventional adequacy value in the regulatory record.

The Challenge

Conventional effective load carrying capability analysis credited the battery under normal peak conditions but said nothing about its performance during the storm events that motivated ratepayer concern. Standard-issue VoLL figures — calibrated on short interruptions — were indefensible for the three-to-six-day restoration windows in the utility's own storm history. And the utility's prior resilience narrative was qualitative, which the commission staff had explicitly flagged as insufficient in the previous cycle.

Keentel's Approach

Event baseline. Extracted resilience events from nine years of the utility's outage management records, grouping forced outages by temporal clustering and duration overlap, and constructed customer-tracked performance curves for every major event.

Scenario construction. Built a probability-weighted scenario set of storm events characterized by intensity, footprint, duration, and restoration trajectory, calibrated against the utility's own area outage rate behavior versus recorded wind speed.
Resilience-adjusted capacity analysis. Simulated each scenario with and without the BESS providing localized supply to sectionalized feeder segments, critical-facility support, and accelerated cold-load pickup during restoration. Baseline expected energy not served across the weighted scenario set was 940 MWh per representative severe event; with the BESS dispatching under the developed resilience operating strategy, this fell to 610 MWh a probability-weighted fractional reduction of 0.35, the asset's resilience-adjusted capacity value.
Duration-aware monetization. Replaced flat VoLL with duration-dependent customer damage functions by customer class, reflecting the strongly nonlinear cost growth of multi-day outages refrigeration losses, business interruption, and critical-facility backup fuel exhaustion and computed avoided outage cost per scenario and in probability-weighted annual terms.
Regulatory-grade documentation. Documented every assumption scenario probabilities, dispatch strategy, damage-function sources, discounting in a traceable engineering basis suitable for discovery and cross-examination.

Outcome

The analysis established an annualized resilience avoided-cost value for the BESS equal to roughly 1.8 times its previously credited adequacy-only value, transforming the asset's regulatory narrative from an arbitrage project with a resilience footnote into a quantified resilience resource. The commission accepted the methodology framework, and the utility adopted the event-extraction pipeline as a standing capability every subsequent storm season now updates the metrics automatically. The resilience operating strategy developed for the study was also folded into the plant's actual emergency dispatch procedures.

Case Study 3 — Islandable Community Microgrid: Quantifying Resilience for Critical Infrastructure

Background

A geographically exposed community served by a single sub-transmission supply path — hosting a regional hospital, water treatment plant, emergency operations center, and communications hub — pursued an islandable microgrid combining solar PV, a battery energy storage system, and existing backup generation. The sponsoring authority required a quantified resilience assessment to support grant funding, utility coordination, and the design basis itself: how long must the island sustain, what must it carry, and how would success be measured after commissioning.

The Challenge

Grid-level metrics were useless at this scale, and the community's requirements were heterogeneous: the hospital and water plant needed continuity measured in days, while general load could tolerate managed curtailment. Historical supply-path outage data existed but was sparse — a handful of multi-day events over two decades — so the assessment had to combine the thin event record with scenario modeling rather than pure statistics. The funding application also demanded metrics a non-specialist review panel could interpret.

Keentel's Approach

Requirement definition from customer damage. Established the community's resilience requirement from the critical-facility side: duration-dependent damage functions for the hospital, water plant, and communications loads defined the cost of inaction as a function of outage duration, revealing that the dominant risk inflection occurred once outages exceeded roughly eight hours — when backup fuel logistics and water storage margins began to bind.
Tiered critical-load design. Defined a tiered critical-load architecture with quantified continuity targets per tier, and designed the sectionalizing and load-management scheme around those tiers rather than treating the community as a single block.
Probabilistic islanding-duration metrics. Simulated islanded operation across seasonal solar profiles, storage states of charge, generator availability draws, and multi-day event durations, computing sustainable islanding duration distributions per tier — not a single advertised number, but a probability statement: the fraction of modeled events in which each tier remains served for one, three, and seven days.
Islanded stability assessment. Assessed islanded power quality and stability — voltage and frequency performance under load steps, motor starts at the water plant, and inverter-generator interaction — establishing the stability component of the resilience metric set and driving control-settings requirements into the design.
Transition and acceptance metrics. Specified transition metrics — planned and unplanned islanding success rate and transfer time, and resynchronization performance — as commissioning acceptance criteria, so the resilience claims made in the funding application became testable contractual requirements.

Outcome

The final design sustained Tier-1 critical loads for seven days in over 95% of modeled events and full community load for more than 24 hours in the majority of scenarios, at a capital cost roughly 20% below the initial single-block concept — the tiered architecture eliminated storage capacity that had been sized to carry non-critical load through worst-case durations. The funding application succeeded, with reviewers specifically citing the probabilistic islanding-duration presentation. Commissioning tests validated the transition metrics, and the community now maintains the metric set as a living dashboard, re-evaluated as loads grow and resources are added.

PART IV KEENTEL ENGINEERING SERVICES: Grid Resilience Metrics and Valuation

Keentel Engineering provides end-to-end resilience measurement, valuation, and enhancement engineering for utilities, developers, independent power producers, large loads, and communities. Our resilience practice is built on the same foundation as our interconnection practice: quantitative, standards-aware, data-driven engineering delivered to a regulatory-grade standard of documentation.

Resilience Metrics Development and Baselining

Extraction of resilience events from utility outage records; construction of outage, restore, and performance processes; computation of the full event metric set including customer-hours, MVA-hours, nadir, and restoration-percentile times.
Area outage rate curve development relating observed outage rates to measured hazard intensity; cost exceedance and heavy-tail risk characterization with exceedance-frequency and log-domain severity indices.
Standing metric pipelines that update automatically each storm season, giving utilities a permanent, auditable resilience measurement capability.

Hardening and Investment Cost-Benefit Engineering

Component fragility modeling for structures, conductors, insulators, and substations against wind, ice, flood, wildfire, and seismic hazards; quantification of failure-return-period improvement per candidate measure.
Rerun-history benefit quantification — expressing investment benefits against the recorded events a service territory actually experienced — and probability-weighted forward scenario analysis.
Portfolio optimization across hardening, operational, and restoration-improvement measures; preparation of resilience plan filings, rate-case exhibits, and grant applications with fully traceable engineering bases.

Resilience Valuation for Storage, Renewables, and DERs

Resilience-adjusted capacity value analysis for BESS, hybrid plants, and DER portfolios across probability-weighted extreme-event scenario sets.
Duration-dependent customer damage function development by customer class and facility type; defensible multi-day VoLL selection for regulatory proceedings.
Resilience-informed dispatch and emergency operating strategy development, integrating storage assets into restoration and critical-load support roles.

Microgrid and Critical-Infrastructure Resilience Engineering

Tiered critical-load architecture and sectionalizing design; probabilistic islanding-duration analysis; islanded stability, protection, and power-quality assessment.
Transition metrics and commissioning acceptance criteria that convert resilience claims into testable requirements; ongoing resilience dashboards for communities and campuses.

Extreme-Event Simulation and Grid Studies

EMT (PSCAD) and positive-sequence (PSS®E) simulation of extreme-event scenarios, cascading outage analysis, ride-through and protection performance under degraded conditions, and restoration/black-start studies.
Interconnection studies that treat resilience as a first-order design input — POI selection, redundancy architecture, ride-through settings, and restoration roles engineered from day one.

Compliance, Planning Integration, and Advisory

NERC operations and planning compliance engineering aligned with resilience objectives; integration of resilience metrics into long-term planning, operational planning, and real-time operating practice.
Owner's engineer and independent review services for resilience programs, storm-hardening plans, and post-event assessments.

Work With Keentel Engineering

Whether you are baselining resilience for the first time, defending a hardening investment, valuing a storage asset's resilience contribution, or designing an islandable system around critical loads — Keentel Engineering brings the metrics, the models, and the regulatory-grade documentation to make the case. Offices in Tampa, FL and Austin, TX. Visit keentel.com to start the conversation

PART II ENGINEERING FAQ: Power System Resilience Metrics

CCPP Electrical Design Engineering | 30%, 60%, 90% & IFC

Tue, 07 Jul 2026 07:04:40 GMT

Jul 07, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

COMBINED CYCLE POWER PLANT (CCPP) ELECTRICAL DESIGN ENGINEERING

A combined cycle power plant pairs a gas turbine with a steam turbine and recovers what a simple-cycle unit throws away. The gas turbine burns natural gas with compressed air; its hot exhaust — instead of venting to atmosphere — passes through a heat recovery steam generator (HRSG) that raises high-pressure steam to drive a second turbine and a second generator. The result is thermal efficiency in the 55–65% range against roughly 35–40% for a simple-cycle machine, with lower fuel burn per megawatt-hour, lower emissions intensity, fast start capability, and strong operating flexibility.

From an electrical engineering standpoint, a CCPP is really three interlocking systems: a generation and interconnection system at high voltage (two or more generators, generator step-up transformers, and a switchyard tied to the transmission grid); a plant auxiliary power system at medium and low voltage (13.8 kV switchgear feeding large motors, stepping down to 480 V for motor control centers and balance-of-plant loads); and an emergency and control power system (125 VDC batteries and UPS) that must ride through a total blackout to protect rotating equipment.

The Keentel Position

Grid interconnection is a first-order design input — not a downstream utility formality. On a CCPP, the point-of-interconnection requirements (short circuit duty, reactive capability, ride-through, protection interfaces, NERC modeling obligations) cascade backward into GSU impedance selection, generator capability curves, switchyard breaker ratings, and even the auxiliary system's ability to survive grid disturbances. Plants that treat interconnection as a late-stage checkbox routinely discover at 90% design that equipment already procured cannot satisfy POI requirements. We anchor the interconnection basis at 30%.

This document lays out the complete electrical engineering scope for a CCPP across the industry-standard design milestone progression — 30%, 60%, 90%, and Issued for Construction (IFC) — under the U.S. IEEE/ANSI framework, with a site basis of 13.8 kV medium voltage and 480 V low voltage. It closes with an engineering FAQ and three anonymized case studies from real project experience.

2. The CCPP Electrical Topology at a Glance

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

1. Introduction: Why CCPP Electrical Design Lives or Dies at the Interconnection

30% — 60% — 90% — IFC Milestone Scope, Deliverables & Studies

An IEEE/ANSI-Based Design Framework for Gas Turbine + Steam Turbine Generation

Notice & Disclaimer

3. Governing Codes & Standards Matrix (U.S. / IEEE-ANSI Basis)

This document is published by Keentel Engineering as original technical content for general informational purposes. It reflects Keentel Engineering's independent design methodology, professional experience, and interpretation of publicly available industry codes and standards.

IEEE, ANSI, NFPA, NEMA, UL, ICEA, NERC, and FERC, and all referenced standard designations, are trademarks or registered marks of their respective organizations. Keentel Engineering is not affiliated with, sponsored by, or endorsed by any of these organizations. References to codes and standards are provided for identification of governing design criteria only; the reader is responsible for obtaining and applying the current published editions.

Case studies contained herein are anonymized and generalized composites drawn from professional experience. Client names, project names, locations, and identifying commercial details have been removed or altered. Nothing in this document constitutes engineering services, a professional recommendation, or a substitute for project-specific engineering performed under the responsible charge of a licensed Professional Engineer.

Before mapping scope to milestones, it helps to fix the reference single-line architecture that the design phases progressively mature:

Gas turbine generator and steam turbine generator — typically 11–24 kV machine voltage (18 kV class is common for large frames), hydrogen- or air-cooled, each with its own excitation system and generator protection package.
Generator step-up (GSU) transformers — designed per IEEE C57.12.00, connecting each generator (via isolated-phase bus on large frames) to the switchyard at transmission voltage.
Switchyard / point of interconnection (POI) — high-voltage breakers rated per IEEE C37.06, disconnect switches, buswork, revenue metering, and the utility protection interface, laid out per NESC clearances.
Unit auxiliary transformers (UAT) and station service transformers (SST) — tapping the generator bus or switchyard to feed the 13.8 kV medium-voltage auxiliary system.
13.8 kV MV switchgear — metal-clad, arc-resistant switchgear feeding boiler feed water pumps, circulating water pumps, condensate pumps, cooling tower fans on larger frames, and GT/ST auxiliary skids.
480 V LV switchgear and MCCs — fed from 13.8 kV–480 V station service transformers, powering motors under roughly 200 HP, valve actuators, package skids (water treatment, lube oil, fuel gas conditioning), HVAC, and lighting/small power via 480–208/120 V transformers.
125 VDC and UPS systems — station batteries sized per IEEE 485 to carry protection, switchgear control, turbine controls, and emergency lube oil pumps through a station blackout.

Five energy flows tie the plant together: natural gas fuel to the gas turbine; hot exhaust to the HRSG; the steam/condensate/feedwater loop through the steam turbine, condenser, condensate pumps, and boiler feed pumps; cooling water between the condenser and cooling tower; and electrical output from both generators to the grid. Every one of those mechanical flows has an electrical consequence — most visibly in the large motor loads that dominate the auxiliary load list.

A Note on IEC Equipment

Gas turbine generator packages are frequently manufactured overseas and arrive with IEC-designed components (IEC 60034 rotating machines, IEC 62271 switchgear). On a U.S. project these must be harmonized with the ANSI/IEEE basis of design — ratings translation, protection interface, and testing acceptance are recurring scope items that belong in the 30% design basis document, not in a construction-phase RFI.

The table below is the standards backbone Keentel applies to a U.S. CCPP electrical design. It is organized by system rather than by document number, because that is how the standards are actually invoked during design.

The 30% (preliminary/schematic) phase exists to make the decisions that are expensive to reverse. For a CCPP electrical scope, the deliverable set centers on the design basis, the overall one-line, the load list, and — critically — the interconnection inputs. Voltage class selection happens here: this reference design fixes 13.8 kV as the MV auxiliary bus and 480 V as the LV utilization voltage, with 4.16 kV noted as the common alternative for mid-size motor blocks where motor economics favor it. That decision drives switchgear ratings, cable systems, and transformer counts for the rest of the project.

30% Scope of Work

Electrical design basis document: voltage classes (13.8 kV / 480 V / 125 VDC), grounding philosophy (HRG on MV auxiliaries per IEEE 142), short circuit design levels, codes and standards matrix, environmental and seismic criteria.
Overall key one-line diagram: both generators, GSUs, switchyard arrangement, UAT/SST scheme, MV and LV bus structure, sources of station service and backup power.
Preliminary auxiliary load list built from the mechanical equipment list — boiler feed water pumps, condensate pumps, circulating water pumps, cooling tower fans, GT/ST auxiliaries, HRSG loads — with duty cycles and diversity assumptions.
Preliminary UAT/SST and GSU MVA sizing; generator terminal voltage and iso-phase bus concept for large frames.
Switchyard concept plan: bus arrangement (ring, breaker-and-a-half, or single bus per POI requirements), preliminary equipment ratings per IEEE C37.06, NESC clearance envelope.
Interconnection engineering inputs: FERC LGIP study data, preliminary power flow and short circuit models, reactive capability at the POI, preliminary NERC applicability review (MOD, PRC, FAC families).
Preliminary equipment layout and electrical room sizing; hazardous area classification concept around fuel gas systems.
Class 4/5 electrical quantities and cost input to the project estimate.

Why 30% Is the Interconnection Milestone

Every input the interconnection studies need — generator reactive capability, GSU impedance, switchyard topology, auxiliary load at the POI — is set at 30%. If those numbers change at 60% or 90%, the interconnection studies get re-run, the queue position may be at risk, and the schedule absorbs the hit. Keentel's practice is to treat the LGIP data submission as a controlled 30% deliverable with the same rigor as the one-line itself.

Step 5: Design Enhancement — Where the Money Is

The 90% Trap

The most common CCPP failure mode at 90% is a study suite still running on assumed data. If the coordination study and arc flash analysis are built on estimated transformer impedances instead of factory test reports, the settings and labels issued at IFC are fiction — and the plant inherits the liability. Keentel gates the 90% studies on receipt of certified vendor data, and flags the gap on the milestone certificate if vendors are late.

At 60% (detailed design development), the architecture stops being conceptual. Equipment is sized and specified for procurement, the study suite produces its first real results, and the MV/LV systems take their final shape. This is the phase where long-lead procurement packages — GSUs, MV switchgear, generators' auxiliary interfaces — must be technically complete, because transformer and switchgear lead times drive the overall project schedule.

60% Scope of Work

Finalized one-line diagrams for HV, MV, and LV systems, including bus ratings, breaker frame sizes, and transformer impedances.
MV system detail design: 13.8 kV metal-clad, arc-resistant switchgear specification per IEEE C37.20.2 and C37.20.7 (compartmentalized construction, vacuum breakers, insulated bus, plenum exhaust routing); MV-105 shielded cable system per UL 1072/ICEA S-93-639 at the 133% insulation level required for the HRG grounding scheme.
LV system detail design: 480 V metal-enclosed switchgear per IEEE C37.13/UL 1558 with draw-out air circuit breakers and electronic trip units; MCC lineups per UL 845/NEMA ICS 18 with MCCB/MCP protection; feeder and branch design to IEEE 141 voltage drop limits (3% branch, 5% total).
Preliminary study suite per IEEE 399: load flow across operating scenarios (startup, base load, one-unit trip), first-pass short circuit per IEEE 551 to verify equipment interrupting duties, and motor starting studies confirming ≥85% terminal voltage on the largest MV motors (boiler feed pumps are usually the binding case).
Procurement specifications: GSU and UAT/SST per IEEE C57.12.00, MV/LV switchgear, MCCs, DC systems, UPS, iso-phase bus, NEMA MG-1 large motors coordination with mechanical.
Grounding grid preliminary design per IEEE 80 using measured soil resistivity; lightning shielding layout per IEEE 998 for the switchyard and outdoor structures.
DC system sizing per IEEE 485: battery duty cycle built from the actual blackout load profile — protection, breaker control, turbine controls, emergency lube oil pumps — with IEEE 946 architecture (chargers, distribution, monitoring).
Cable and raceway design development: tray routing, duct banks, segregation of power/control/instrumentation, preliminary cable schedule.
Protection philosophy document: generator protection per IEEE 242 (21 distance backup, 59N stator ground, 87G differential, plus 40, 46, 24, 81 elements), transformer differential zones, MV/LV coordination strategy.

At 90%, the design must be construction-complete in substance; only vendor-data backfill and final review comments separate it from IFC. The study suite is finalized against actual purchased-equipment data (real transformer test impedances, real motor characteristics, real relay models), protection settings are calculated, and the arc flash analysis produces the labels that will physically go on the gear.

90% Scope of Work

Final power system studies per IEEE 399/551 using as-purchased equipment data: load flow, short circuit to the lowest 480 V panelboard with AIC margins verified, motor starting re-run with certified motor curves.
Protection coordination study per IEEE 242/3004.5: complete time-current characteristic (TCC) sets demonstrating that a 480 V fault clears at its MCC breaker before the upstream 13.8 kV feeder responds, generator and transformer relay settings per the protection philosophy, coordination with the utility's POI relaying.
Arc flash hazard analysis per IEEE 1584 and NFPA 70E: incident energy at every bus, label schedule, and mitigation design where categories are unacceptable (maintenance-mode switching, arc-resistant credit, remote racking).
NERC compliance engineering: PRC-019 coordination of generator voltage regulation controls and protection, PRC-024 (transitioning to PRC-029 for applicable resources) frequency/voltage ride-through verification, MOD-025/026/027 test plans, FAC-008 facility ratings methodology.
Final grounding grid design per IEEE 80 with step/touch potentials demonstrated below tolerable limits; final IEEE 998 shielding verification.
Complete cable schedule, conduit/tray fill calculations, cable pulling calculations, and raceway drawings; segregation and separation verified.
Full drawing set at 90%: one-lines, three-lines, schematics, wiring diagrams, panel schedules, grounding plans, lighting, cathodic protection interfaces, hazardous area drawings.
Relay settings files drafted for the microprocessor protection platform; SCADA/DCS I/O lists and network architecture coordinated with controls.

6. The 90% Design Milestone — Studies Closed, Settings Issued, Field-Ready

4. The 30% Design Milestone Fixing the Basis Before It Fixes You

7. IFC — Issued for Construction and Beyond

8. The Study Suite: The Analytical Spine of All Four Milestones

IFC is not a drafting exercise; it is the point at which the engineer of record stakes a license on the package. Every review comment is closed, every hold is lifted, every calculation is checked and signed, and the drawings are stamped by the responsible Professional Engineer.

IFC Scope of Work

Stamped, signed IFC drawing and specification packages for all electrical systems, incorporating final vendor data and constructability review comments.
Final relay settings issued under configuration control, with setting sheets traceable to the coordination study revision.
Final calculations package: studies, grounding, DC sizing, voltage drop, cable ampacity and derating, lighting — indexed and archived for the owner and for NERC evidence.
Commissioning support documents: energization sequence, protection functional test plans, MOD-025/026/027 field test procedures, breaker and transformer acceptance test criteria.
Construction-phase engineering: RFI responses, field change management, factory and site acceptance test witness, red-line capture.
As-built close-out: conformed drawings, final settings as-left records, updated models delivered to the owner for ongoing NERC MOD compliance.

Studies are not a milestone deliverable so much as a thread that runs through all four. Each study matures in lockstep with data quality:

How Keentel Delivers This Scope

Keentel Engineering provides CCPP electrical design as a full owner's engineer or detailed-design scope — one-lines through IFC, the complete IEEE 399 study suite, protection and settings, grounding, DC systems, and the interconnection and NERC compliance layer that most design firms treat as someone else's problem. Because our practice is built around grid interconnection, the POI requirements are engineered into the plant from the 30% design basis, not discovered at commissioning.

Case Study

13. About Keentel Engineering

Case Study 1 — 2×1 CCPP: Interconnection Requirements Rescued at 30%

Profile: A nominal 700 MW-class 2×1 combined cycle facility (two gas turbines, two HRSGs, one steam turbine) interconnecting to a 230 kV transmission system in a U.S. RTO footprint. Keentel's role: electrical design lead and interconnection engineer. All identifying details anonymized.

The Challenge

The developer had inherited a conceptual design from an earlier feasibility effort in which the GSU impedances and switchyard arrangement had been chosen for cost, with the interconnection application treated as a parallel administrative track. Early cluster study results indicated the POI short circuit duty was near the interrupting capability of the assumed 230 kV breakers, and the reactive range the interconnection agreement would require at the POI could not be delivered through the assumed GSU impedance across the full generator capability curve.

The Keentel Approach

Rebuilt the 30% design basis with the interconnection requirements as first-order inputs: POI reactive envelope, ride-through obligations, and short circuit headroom were written into the basis document alongside voltage classes and grounding philosophy.
Re-optimized GSU impedance jointly against short circuit contribution and reactive transfer, iterating the interconnection power flow cases with the plant one-line rather than treating them as separate models.
Moved the switchyard from a single-bus concept to a breaker-and-a-half arrangement with breaker ratings selected per IEEE C37.06 against the study-derived duty plus system growth margin.
Locked the LGIP data submission as a controlled 30% deliverable with a formal design freeze on interconnection-coupled parameters.

The Outcome

The revised parameters passed the restudy without a queue position change, and no interconnection-driven equipment change occurred after 30%. The GSUs were released for procurement at 60% against final, study-consistent impedances. The developer's independent engineer specifically credited the integrated interconnection/plant-design model with removing what it had flagged as the project's largest single schedule risk.

Case Study 2 — MV/LV Auxiliary System: HRG and Arc Flash Engineered Together

Profile: A 500 MW-class 1×1 combined cycle plant with a 13.8 kV auxiliary backbone and 480 V utilization system. Keentel's role: MV/LV detail design, protection, and safety studies from 60% through IFC. All identifying details anonymized.

The Challenge

The plant's availability model could not tolerate an auxiliary bus trip on a single line-to-ground fault — a forced GT runback traced to a grounded cable had burned the owner on a prior asset. Simultaneously, preliminary arc flash screening showed several 480 V switchgear buses trending toward incident energies that would have made routine racking operations impractical under NFPA 70E.

The Keentel Approach

Implemented high-resistance grounding on the 13.8 kV auxiliary system per IEEE 142, with pulsing ground-fault location and alarm-and-locate operating procedures, so a first ground fault alarms rather than trips.
Carried the HRG decision through its consequences: MV-105 shielded cable at 133% insulation level per UL 1072/ICEA S-93-639, and charging-current verification to keep the fault current within the resistor's design window.
Specified arc-resistant 13.8 kV metal-clad switchgear per IEEE C37.20.7 with engineered plenum exhaust routing coordinated with the building design — a detail frequently discovered as an interference at construction.
On the 480 V system, combined IEEE C37.13/UL 1558 switchgear with electronic trip units and a maintenance-mode (arc energy reduction) switching scheme, then re-ran IEEE 1584 with maintenance mode credited to bring worst-case buses into workable categories.
Delivered the full TCC coordination set per IEEE 242/3004.5 demonstrating an LV fault clears at its MCC breaker before the 13.8 kV feeder relay responds — selective clearing preserved despite the fast maintenance-mode settings.

The Outcome

The plant entered commercial operation with a single-ground-fault ride-through capability on its MV auxiliaries, arc flash labels at every bus reflecting the mitigated energies, and a coordination study that survived the utility's POI relaying review without a settings change. In its first operating year, one MV cable ground fault was located and repaired during a planned reduction rather than forcing a unit trip — the exact scenario the design was built for.

Case Study 3 — Fast-Track Brownfield Addition: Motor Starting and the Study Suite Under Schedule Pressure

Profile: A gas turbine and HRSG addition converting an existing simple-cycle site to combined cycle, with new boiler feed water pumps and cooling infrastructure added to an aging auxiliary system. Keentel's role: power system studies , auxiliary system upgrade design, and IFC protection settings on a compressed schedule. All identifying details anonymized.

The Challenge

The conversion added multi-thousand-horsepower boiler feed pump motors to a station service system originally sized for simple-cycle auxiliaries. Screening showed motor terminal voltage during BFP acceleration falling well below the 85% threshold under the weakest credible source configuration, and the legacy switchgear's interrupting ratings had never been re-verified against decades of transmission system growth. IFC was contractually fixed; the studies could not slip.

The Keentel Approach

Executed the IEEE 399 suite as a single integrated model — load flow, IEEE 551 short circuit to the lowest 480 V panelboard, and dynamic motor starting — so every design change propagated through all three analyses in one pass instead of three sequential revisions.
Resolved the BFP starting problem with variable frequency drives, which eliminated the voltage-dip constraint, cut pump energy at part load, and removed a planned UAT upsizing from the project scope.
Identified two legacy MV breakers with insufficient interrupting margin against present-day duty; specified replacements within the existing lineup footprint to avoid switchgear building modifications.
Gated the 90% studies on certified factory test data for the new transformers and motors, and issued the IEEE 1584 arc flash labels and final relay settings from that verified model under configuration control.

The Outcome

The addition energized on the contractual date. First BFP starts matched the dynamic study predictions within measurement tolerance, the replaced breakers cleared their commissioning duties with verified margin, and the owner received a single indexed calculation archive that has since served as the plant's NERC evidence baseline. The VFD decision, made analytically at 60% rather than as a field fix, was later estimated by the owner to have paid for the entire study scope several times over.

Frequently Asked Questions

Keentel Engineering is a power systems consulting firm specializing in grid interconnection, generation and substation design, transmission engineering, and NERC compliance. Our practice spans EMT and PSS/E modeling, POI interconnection engineering, utility-scale generation and storage design, owner's engineer services, and the milestone-based detailed design delivery described in this document.

What differentiates our CCPP work is the integration of interconnection engineering with plant electrical design under one roof — the same team that builds your one-lines builds your LGIP models, coordinates your POI protection, and assembles your NERC evidence. Interconnection is a first-order design input in our process, and it shows in schedules that hold.

Engage Keentel

For CCPP electrical design, owner's engineer support, interconnection studies, or NERC compliance engineering, contact Keentel Engineering — Tampa, FL • Austin, TX. Deliverables are executed under the responsible charge of a licensed Professional Engineer.

ERCOT Batch Zero Guide: Large Load & Ride-Through Rules

Sun, 05 Jul 2026 13:55:09 GMT

Jul 05, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Texas/ERCOT Rewrote the Rules for Big Power Users: A Plain-English Guide to ERCOT's Batch Zero, Ride-Through Requirements, and What Your Project Must Do Now

Every large load project seeking to connect lands in one of three categories:

Base Load

Projects that were already far along energized, or with completed and valid studies, an executed interconnection agreement, and hard evidence of commitment such as purchased high-voltage transformers and breakers and construction underway. Base Load projects keep the capacity their prior work supports. They're modeled in the study but not re-evaluated, and they can keep moving toward energization while the batch study runs.

Studied Load

Projects with real momentum site control, posted financial security, notarized attestations but without the full Base Load paper trail. These projects go into the system-wide study, and ERCOT determines how many megawatts each can draw, year by year, based on what the transmission grid can reliably support once everyone is assessed together.

Excluded

Projects that missed deadlines or eligibility criteria. Excluded projects cannot receive approval to energize until a future batch study — there is no individual-study side door anymore.

One detail that surprises people: a project can be excluded because its utility missed a submission deadline, not just because the developer did. Smart developers actively project-manage their transmission or distribution service provider's paperwork rather than assuming it's handled.

The Timeline

Two numbers tell the story. By mid-2026, large power users had asked to connect more than 438,000 MW of new demand to the Texas grid — roughly ninety percent of it data centers. ERCOT's all-time record for the entire state's peak demand is about 85,500 MW. In other words, the request queue was more than five times the size of everything Texas has ever consumed at once.

The old process studied each project individually. With thousands of megawatts arriving every month, that meant endless restudies, multi-utility coordination headaches, and a growing backlog. Nobody not developers, not utilities, not ERCOT could get a straight answer about what the grid could actually support.

There was a second, quieter problem. Since October 2022, ERCOT has documented repeated incidents where large electronic facilities tripped offline during routine voltage dips — the kind of momentary disturbance the grid experiences whenever lightning strikes a line and protection equipment clears the fault correctly. A data center's uninterruptible power supply (UPS) is designed to protect servers by instantly switching to batteries or generators at the first sign of trouble. That's great for the servers. But when several gigawatts of load vanish from the grid in a fraction of a second, the sudden imbalance can destabilize the entire system. ERCOT identified clusters of large loads capable of dropping more than 3,200 MW simultaneously during a severe disturbance — a genuine reliability threat.

So ERCOT built a package of five interlocking rules to fix both problems at once.

The Five Rules, and What Each One Does

Batch Zero: The Three Buckets

Ride-Through: Your Facility's New Job During Grid Disturbances

Two Numbers Every Studied Load Gets: LPC and MPC

Batch Zero was designed to separate real projects from speculative queue positions, and it does that with dollars:

$100,000 per MW of financial security at the intermediate-agreement stage (cash, an investment-grade parent guaranty, or a letter of credit from a highly rated bank). On a 500 MW project, that's $50 million posted before you have a study result.
Another $100,000 per MW as a non-refundable interconnection fee when you execute the interconnection agreement paid to the transmission utility and credited against its rate base, so it ultimately benefits ratepayers.
Full cash payment (CIAC) for all direct interconnection costs the radial lines and substation work built specifically for you with no allowance and no rate recovery, trued up to actual costs.
Clawbacks with teeth. Withdraw your request, or miss a phased-energization milestone by six months, and the utility draws your security against its sunk costs, refunds only 20% of what remains, and sends the other 80% to the transmission rate base. CIAC is never refundable. ERCOT reallocates your freed-up capacity to someone else.
Even a fully successful project gets its security back slowly: 20% of the balance at energization, further amounts as milestones are hit, and the remainder only after five years of sustained operation at contracted peak demand.

Add the required disclosures site control via deed or five-year lease, sworn officer attestations on permits and site studies, and mandatory disclosure of any "substantially similar" interconnection request you're pursuing elsewhere (the anti-queue-shopping provision) and the message is clear: reserving grid capacity in Texas is now a binding, expensive act.

Here is the concept in one sentence: when the grid stumbles, your facility must stay on the grid and keep consuming, within defined limits, instead of instantly fleeing to batteries and generators.

Frequency

The grid normally runs at 60 Hz. Covered facilities must operate continuously anywhere between 58.8 and 61.2 Hz, and must hang on for about five minutes (299 seconds) in the bands just outside that. Only beyond roughly 57.0 Hz or 61.8 Hz may the facility choose to trip. Protection systems may not trip on an instantaneous frequency reading they must use filtered measurements or time delays, because momentary sensor blips shouldn't take a gigawatt offline.

Voltage

The requirements form a ladder. Between 90% and 110% of nominal voltage, operate continuously. For deeper sags, the facility must remain connected for a defined time two seconds down to 80% of nominal, half a second down to 50%, a quarter second down to 20%, and 0.15 seconds even below that. Two performance duties ride along:

The one-second recovery rule. If the facility reduced consumption during a deep sag, it must return to at least 90% of its pre-disturbance consumption within one second of voltage recovering. The grid needs the load back promptly, or the recovery itself becomes a disturbance.
The 125% current cap. During any disturbance outside the normal band, the facility may not draw more than 125% of its maximum normal operating current.

Three provisions catch facilities off guard. First, sag-counting trip schemes are banned your controls can't trip because they saw, say, five sags in an hour if each individual sag was within the ride-through envelope. That directly targets common industrial power-quality relay logic. Second, protection must be set to the equipment's maximum capability, not just the minimums in the tables. Third, there's a vintage split: projects qualifying after January 1, 2028 face a stricter duty to keep consuming during moderate sags (down to 50% voltage), with no "if capable" escape hatch so anyone specifying power electronics today for a 2028+ energization should buy to the stricter standard.

Enforcement

Enforcement is real. After a ride-through failure, the owner must investigate root cause within 90 days of ERCOT's request, submit a corrective plan within 90 days after that, and implement it within 180 days of approval. And if ERCOT finds that continued operation poses an imminent reliability risk, it can order the facility disconnected until compliance is demonstrated. For a revenue-generating data center, that provision deserves board-level attention.

Who's Exempt?

Facilities that received approval to energize by November 14, 2025, or that had completed studies and utility confirmation by that date, are grandfathered. Everyone else in the queue is covered.

What Your Team Should Be Doing Right Now

ERCOT enforces all of the above through simulation, which means your model is your compliance case. The bar is now high:

Whole-facility scope. The model must represent everything the IT load, the UPS and power conversion, the cooling plant, the protection and control systems in formats compatible with ERCOT's study platforms (PSS/E, PSCAD, TSAT).
Real control loops, not approximations. Generic textbook representations are unacceptable. The model must capture the actual inner control behavior of your power electronics.
Hardware-validated converter models. For electronic loads, the PSCAD model must be benchmarked against actual hardware testing including voltage ride-through and subsynchronous response. A model assembled from standard PSCAD library blocks fails by definition, because a generic block has never been tested against your vendor's hardware. The good news: validation is a hardware-type test, so results for a given converter product are reusable across every facility that uses it.
Format migration. Facilities that previously submitted the older composite load model (CMLD) format must transition to EPRI's PERC1 format.
Three checkpoints. Models are reviewed before the stability study begins (no model, no study), before each quarterly stability assessment, and for electronic loads one final time before energization, when you must submit as-built models with a documented comparison against the previously studied data and a sworn attestation that the model matches actual field settings. ERCOT's review takes 10 business days, extendable by 20 put it on your critical path.
A living obligation. Change your technology, controls, or relay settings in a way that affects ride-through including converting a crypto mining site to an AI data center — and you've triggered a new interconnection study, even if your megawatts don't change.

How Keentel Engineering Can Help

The Money: What Commitment Actually Costs

This framework merges three disciplines that rarely live in one team: interconnection process management, power-electronics dynamic modeling, and protection engineering. That intersection is where we work. Keentel Engineering supports large load developers with applicability and grandfathering assessments, PERC1 model development and conversion, hardware-benchmarked PSCAD converter validation, ride-through and stability studies, protection coordination that meets the new measurement-window and filtering rules without sacrificing equipment protection, DWG survey preparation, and as-built model packages for the pre-energization gate.

If your project is in Batch Zero or aiming for the next batch the time to engineer compliance is before ERCOT's checkpoint, not after a deficiency notice. [ Contact our team] for a readiness review.

Why Did ERCOT Do This?

ERCOT compliance rules evolve quickly. This article reflects the framework approved by the Public Utility Commission of Texas on June 18, 2026. Always verify current requirements against official ERCOT postings before making filing or design decisions or contact our team and we'll do it with you.

The Short Version

Frequently Asked Questions

If you are building a data center, crypto mining facility, hydrogen plant, or any other project in Texas that will draw 75 MW or more from the grid, the rules of the game changed in mid-2026. ERCOT the operator of the Texas power grid replaced its one-project-at-a-time interconnection study process with a system-wide "batch" process. The first batch is called Batch Zero, and the door to get into it is nearly closed: key submissions were due July 10, 2026, with the utility-side package due July 24, 2026, and a final deficiency-cure deadline of August 31, 2026.

At the same time, ERCOT imposed brand-new technical performance rules on large electronic loads requirements to "ride through" grid disturbances instead of tripping offline and dramatically raised the bar for the computer models developers must submit to prove their facilities will behave.

This guide explains the whole framework in plain English: why it happened, what the five interlocking rule changes actually say, what it costs, and what your project team needs to do. At the end, we answer the 20 questions we hear most often.

The study produces two figures for each project. The Low Power Consumption (LPC) is the firm amount megawatts you can draw with no curtailment obligation. The Maximum Power Consumption (MPC) is your full desired peak. A project that makes no special election is capped at its LPC. To access the capacity between LPC and MPC, you choose one of two pathways:

PCLR the flexibility path (Form W)

A Provisional Controllable Load Resource can consume above its LPC, all the way to its MPC, in exchange for following ERCOT's real-time dispatch instructions including instructions to reduce consumption when the grid is constrained. Think of it as: you get the power sooner, but ERCOT holds a dimmer switch on the portion above your firm allocation. PCLRs also can't sell ancillary services while in provisional status.

WLPUN the bring-your-own-generation path (Form X)

A Withdrawal-Limited Private Use Network pairs the load with co-located on-site generation. The generation carries part of the load, and ERCOT allocates an annual megawatt amount for the incremental portion served from the grid. To qualify, the generation project needed its own ERCOT approval by the July 10 deadline.

Both elections were due to the utility by July 10, 2026, on notarized forms.

The Digital Twin: Why Dynamic Models Became the Hardest Deliverable

Confirm your classification. Did your submissions land by July 10? Did your utility's package go in by July 24? Watch for ERCOT's August 7 notice and be ready to cure any deficiency before August 31 — that window is short and unforgiving.

Audit your ride-through posture. Review UPS transfer settings, protection relay logic, sag-response behavior, and the one-second recovery capability against the envelopes above — before ERCOT's simulations do it for you.

Get your model house in order. Inventory what was submitted, verify PERC1 conversion status, chase vendors for hardware-validated converter models, and confirm your DWG survey still matches the current design.

Plan the as-built gate. Build the pre-energization model submission and ERCOT's 10–30 business day review into your commissioning schedule.

Treat design changes as regulatory events. Any protection setting change or technology conversion now needs a compliance check first.

If you're a PCLR elector, calendar March 1, 2027 for Form W Part B and the interconnection agreement — missing it means withdrawal.

Think of the package as a cascade. Each rule answers one question:

1. NPRR1308 Who is covered?

It defines a Large Load as one or more facilities at a single site with combined peak demand of 75 MW or more. It then defines a subset originally called a Large Electronic Load (LEL), later renamed Large Computational Load (LCL) as a Large Load where at least half the demand is power-electronic, computational equipment. That's the data center and crypto mining profile. The extra technical rules attach to this subset, because it's the power-electronics interface, not the megawatt-hours, that creates the trip risk.

2. NOGRR282 What performance is required?

This rule creates formal frequency and voltage ride-through requirements. Covered facilities must stay connected and keep consuming power through defined grid disturbances instead of instantly jumping to backup power. We break down the actual numbers below.

3. PGRR144 How do you prove your facility can do it?

ERCOT can't physically test every facility, so it requires a high-fidelity dynamic model essentially a digital twin of your facility's electrical behavior validated against real hardware tests, checked at three checkpoints: before your stability study, before your quarterly assessment, and (for electronic loads) before you're allowed to energize.

4. PGRR145 When and under what commitment do you connect?

This is Batch Zero itself: the eligibility criteria, the system-wide study, the classifications, the deadlines, and the financial commitments.

5. NPRR1325 The market-rules companion.

It carries the definitions into the ERCOT Protocols and, in its final form, the two optional operating pathways (PCLR and WLPUN, explained below) that let projects access more capacity sooner in exchange for flexibility.

PRC-028-1 & NOGRR255 Compliance Guide | SEL-2240 Axion

Sun, 05 Jul 2026 09:48:29 GMT

Jul 05, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Achieving PRC-028-1 &

NOGRR255 Compliance with the

SEL-2240 Axion & SEL RTAC

4.1 Why This Platform

A traditional standalone DFR is a single centralized chassis with every CT and PT circuit hauled back to one panel. That model works, but at an IBR collector substation it usually means long secondary cable runs, a large panel footprint, and the real killer on brownfield sites dependence on spare CT cores that often do not exist. The SEL Axion inverts the model: compact, modular acquisition nodes are placed close to the signals, and a deterministic EtherCAT network carries time-aligned samples back to a central SEL RTAC that performs all recording, triggering, retention, and file generation. Per SEL's published specifications, the platform scales to 192 analog channels or 1,728 digital channels, records triggered oscillography at up to 24 kHz with 24-bit resolution, continuously records at 3,000 samples per second for more than 20 days, streams synchrophasors at up to 120 messages per second, and stores as many as 500,000 SOE records with 1 ms accuracy with native IEEE C37.111 COMTRADE output and IEEE C37.232 COMNAME file naming.

Critically, every AC analog input module in an Axion system samples at precisely the same instant, with sub-microsecond alignment across nodes and when connected to IRIG top-of- second synchronization across geographically dispersed locations. For disturbance analysis, that means every voltage and current in the plant shares one time reference without post-processing.

4.2 The Hardware Building Blocks

4.3 Storage Sizing Do the Math Before You Order

Continuous Recording Groups are storage-hungry: each voltage or current channel configured for continuous recording consumes roughly 600–700 MB per day at 3 kHz. A typical single-transformer collector substation lands around 30 analog channels — call it 20 GB per day — so the 20-day regulatory minimum needs about 400 GB, and a comfortable 45-day design target lands near 900 GB. A 2 TB SSD covers that with margin; SEL offers drives up to 8 TB for the SEL-3555, SEL-3560E, and SEL-3350 where multi-transformer stations push channel counts higher.

A point owners frequently miss: because the NOGRR255 retention and provision obligations are already live, a facility with any existing recording capability (relay event reports, RTAC SER logs) must already be retaining that data and be able to deliver it to ERCOT on request. An interim-compliance configuration of the existing relays is therefore a standard first step in every Keentel deployment, protecting the owner during the construction window.

3. What the Standards Actually Require: SER, FR, and DDR

4. The Solution Architecture: SEL RTAC + SEL-2240 Axion DFR

7. Beyond the Checkbox: What Owners Actually Gain

5. Installing an Axion Node per Transformer: Step-by-Step

Treating this purely as a compliance cost misses the operational value. With a commissioned RTAC + Axion DFR, an Apparent Performance Failure investigation under ERCOT's ride-through rules with its 90-day plan and 180-day implementation clocks starts from complete, time-aligned data instead of guesswork. Impedance-based fault location on the interconnection line shortens patrols. Continuous PMU data feeds model validation and oscillation analysis. And because the entire configuration lives in one templated, no-code SEL-5033 project, the second site in a fleet deploys in a fraction of the engineering hours of the first. Keentel Engineering provides the full lifecycle gap analysis, design, procurement, installation, commissioning, and the audit-ready evidence book as a turnkey substation service .

Ready to start your own compliance program?

6. Compliance Mapping: Requirement to Platform Capability

Keentel Engineering's substation services team delivers gap analysis, DME design, procurement, installation, commissioning, and audit-ready documentation for PRC-028-1 and NOGRR255 for a single site or an entire fleet. Visit www.keentelengineering.com to schedule a consultation.

Because the PRC-028-1 dates vary by facility category, here is the full matrix:

An in-depth technical guide from the Keentel Engineering substation services team on designing, installing, and commissioning a compliant Digital Fault Recorder (DFR) system for inverter-based resources using the SEL-2240 Axion distributed I/O platform and the SEL Real-Time Automation Controller (RTAC).

1. Why These Standards Exist: The IBR Visibility Gap

Case Study 1 Brownfield 100 MW Solar Facility: DFR Retrofit Without Rewiring a Single Feeder CT

Background

A 100 MW-class utility-scale solar facility in the ERCOT region interconnected at 138 kV through a single collector substation: one three-winding main power transformer, a 34.5 kV collector bus, and four medium-voltage feeder circuits. The site had been built several years earlier with a strong protection and automation backbone SEL line, transformer, bus, and feeder relays, an SEL-3555 RTAC serving as SCADA gateway, a GPS clock with IRIG-B distribution, and a fiber LAN but no dedicated disturbance recording device of any kind. A third-party compliance review had flagged the disturbance recording methodology as undefined and the largest outstanding gap against PRC-028-1 and NOGRR255, with time-synchronization commissioning evidence also missing.

Challenge

No spare CT cores existed at any of the four feeder positions — a conventional centralized DFR would have required breaking in-service secondary circuits, extended outages, and significant rewiring cost.
The existing SEL-3555 was mission-critical for SCADA and plant control interfaces; any DFR duty added to it could not jeopardize those functions.
NOGRR255 retention and data-provision obligations were already legally in effect while the project was still in design — the owner had exposure during the construction window.
The owner needed audit-grade evidence, not just hardware: settings files, commissioning records, and procedures were all absent from the site records.

Solution

Keentel designed a two-node SEL-2240 Axion DFR reporting to the existing SEL-3555. Node 1, in the control house, carried SEL-2245-42 AC Protection Modules wired conventionally to the 138 kV line PTs and transformer high-side CTs (where spare secondaries existed) plus an SEL-2244-2 digital input module for breaker positions. Node 2, at the medium-voltage switchgear lineup, used an SEL-2245-43 12-channel module paired with non-intrusive split-core CTs clamped around all four feeder circuits and the bus PTs — zero secondary rewiring, with field calibration compensating split-core gain and phase error. The nodes daisy-chained on LC fiber back to an SEL-3390E4 EtherCAT card added to the RTAC, which was upgraded to R153 firmware with FileIO, DDR, and Continuous Recorder licenses and a 2 TB SSD. Fault recording was set at 24 kHz — above both the 64 samples-per-cycle NERC floor and the 128 s/c NOGRR255 requirement for equipment installed after January 1, 2024 — with a 2.5-second pre-trigger and 6-second records for margin over the ERCOT 5-second total-record floor. Continuous Recording Groups ran on all station channels with a 45-day retention target, comfortably above the rolling 30-calendar-day ERCOT requirement. Trigger settings followed the verified floors: voltage more sensitive than the 1.1/0.9 p.u. thresholds, residual overcurrent, over/underfrequency at the transformer and feeder assets per PRC-028-1 R3, and relay-trip cross-triggering from all protection groups. Inverter fault codes, mode changes, and ride-through flags already flowing through the RTAC from the plant controller were mapped as DFR custom channels — embedding the SER content requirements directly into the COMTRADE records. As an interim measure delivered in week two, Keentel configured and documented the existing relays' event recording and retention so the site met the already-effective NOGRR255 obligations during construction.

Implementation Highlights

Total on-site outage exposure: one four-hour window at the transformer position; the four feeder circuits required no outage at all thanks to the split-core approach.
DFR configuration through the SEL-5033 menu-driven asset model — line asset with impedance fault location, transformer generic assets, four feeder line assets with voltage (105%/95%), sequence-component, and frequency/ROCOF triggers — compiled at 0 errors / 0 warnings.
Commissioning proved the full chain: secondary-injection triggers to COMTRADE, on-demand CRG web retrieval, timestamp verification of every IED against UTC, and a live demonstration of the ERCOT data provision workflow against the 7-calendar-day NOGRR255 clock, including SER export in the Attachment 1 CSV field order. The closeout package also delivered the DME Equipment Reporting Database and a maintenance & testing plan built around the PRC-028-1 R8 process (restore within 90 days or file a Corrective Action Plan) with the 30-day data-availability verification implemented through automated DFR-alarm-to-SCADA notification.

Results

Lessons Learned

The CT question decides the retrofit economics resolve conventional-versus-split-core per position during the gap analysis, before the BOM is priced. And treat the already-effective NOGRR255 obligations as day-one scope: the interim relay-recording configuration cost a few engineering hours and removed months of compliance exposure.

Lessons Learned

Treat the first site as a product, not a project: every hour invested in templating the asset model and test plans pays back multiplied across the fleet. Confirm legacy RTAC capability early discovering mid-project that an installed controller cannot host the DFR extension wrecks both budget and milestone sequencing. And read the final standard, not the draft: the unit-level requirement changed materially between the circulated drafts and the adopted PRC-028-1, and re-baselining to the final language cut meaningful hardware cost from the program.

Case Study 2 Multi-Site Wind Fleet: Legacy RTAC Replacement and a Template That Scale

Background

A renewable owner-operator held a fleet of wind facilities in the 100–200 MW class across ERCOT and neighboring regions, most built in the early-to-mid 2010s. The sites shared a familiar vintage architecture: SEL feeder and transformer relays, capacitor banks with their own relays, and older-generation RTAC hardware — including SEL-3505 units at several sites — serving as SCADA gateways. Facing the PRC-028-1 fleet milestones (50% by January 1, 2029 and 100% by January 1, 2030 for existing BES IBRs) and NOGRR255's faster equipment clock at the Texas sites, the owner needed a repeatable program, not a series of one-off projects.

Challenge

The legacy SEL-3505 RTACs could not perform the fault recording or dynamic disturbance recording functions the standards require — a hardware replacement was unavoidable at those sites, and existing SCADA logic had to migrate without disrupting operations.
Wind turbine generators were distributed across kilometers of collector feeders, and unit-level obligations were a moving target: the owner had scoped its program against draft-era language requiring 128 s/c oscillography at units on the last 10% of each feeder, but the final adopted PRC-028-1 contains no unit-level fault recording — the final unit-level obligation is SER-type data (fault codes, alarms, and ride-through mode status), and for units in commercial operation before the effective date it applies only if the units are capable.
Fleet-level milestone accounting: the owner had to sequence which sites counted toward the 50% tranche and prove progress to its compliance organization.
Each site had different feeder counts, CT ratios, and cap-bank arrangements — but the owner's budget assumed template reuse, not bespoke engineering per site.

Solution

Keentel executed a pilot-then-template program. At the pilot site, the legacy RTAC was replaced with an SEL-3555 running R153+ firmware with the SEL-3390E4 EtherCAT card and DFR licensing; existing SCADA logic and protocol maps were converted into the new project so the gateway function carried over intact. A control-house Axion node (SEL-2245-42 modules) covered the main power transformer and 34.5 kV bus; a second node covered the feeder lineup and the capacitor bank — the cap bank defined as a Generic asset exactly as SEL's asset model intends for reactive devices. For the unit level, the program pivoted to the final standard's actual requirement: Keentel obtained OEM capability documentation for every turbine converter model in the fleet — manufacturer letters and equipment specifications that, under the “if capable” provision for existing units, are themselves acceptable compliance evidence — and integrated the available fault codes, alarms, and ride-through mode status tags from the turbine SCADA into the RTAC SER and DFR custom channels with no new field wiring. At two sites where the owner elected unit-level oscillography for its own ride-through analytics, compact remote Axion nodes at selected pad-mount locations were linked over existing single-mode fiber — comfortably within EtherCAT's roughly 15 km single-mode reach — recording at 24 kHz, with deterministic EtherCAT and IRIG top-of-second synchronization keeping the remote measurements time-aligned with the substation channels. Every design artifact — asset model, trigger templates (including ROCOF triggers tuned for the weak-grid corner of the system), channel schedule format, panel drawings, test plans — was built as a fleet template.

Implementation Highlights

Pilot site: 26 weeks including the RTAC replacement and SCADA migration; template sites thereafter averaged 16–18 weeks.
Engineering hours per site fell by roughly half from the pilot to the third deployment — the direct payoff of the SEL-5033 no-code asset model and template reuse.
OEM capability letters closed the unit-level obligation for the majority of the fleet at documentation cost only; the elective remote turbine-pad nodes recorded at 24 kHz for the owner's own analytics. Because R8 took effect for BES IBRs on January 1, 2026, every site's DFR alarm logic (recording stopped, storage low, EtherCAT abnormal) was mapped to SCADA — the automated failure detection that both the 90-day restore-or-CAP process and NOGRR255's 30-day verification obligation assume.
A fleet milestone tracker mapped each commissioned site against the NOGRR255 2-year/4-year and PRC-028-1 2029/2030 dates, giving the owner's compliance team defensible progress evidence.

Results

Frequently Asked Questions

Lessons Learned

On a greenfield, the cheapest disturbance monitoring program is the one specified before steel is ordered: dedicated CT cores cost almost nothing at procurement and a fortune after energization. Engage the inverter OEMs on unit-level SER capability during factory coordination new units get no “if capable” relief and let the DFR channel math, not habit, size the RTAC hardware and storage.

2. Who Is In Scope, and When

The design pattern Keentel applies most often is one Axion node per main power transformer position (covering the high-voltage interconnection and the transformer itself) plus one node per collector bus (covering the medium-voltage bus and its feeders), all reporting to a central SEL-3555 RTAC in the control house. At a single-transformer collector substation that is two nodes; at a two-transformer station, three to four. Here is the full workflow, from paper to energized recorder.

Step 1 Define the monitored areas from the single-line diagram

Start with the SLD and mark every location the standards require: the high-voltage line/POI, the main power transformer high and low sides, the collector bus, each feeder breaker, any capacitor banks or dynamic reactive devices, and the IBR units on the last 10% of each feeder. Each marked location becomes a set of CT/PT signals and digital points, and every CT/PT group maps to an Axion analog module. Best practice: over-mark now and leave chassis slots empty a nine-slot SEL-2242 node rarely needs to ship full, and empty slots are free expansion for the future feeder positions most collector stations reserve.

Step 2 Verify the RTAC and order the right parts

Four hard gates decide the bill of materials check them before anything is purchased:

Firmware: the DFR extension requires RTAC firmware R153-V0 or later (Continuous Recording Groups arrived at R152). Older firmware means a planned upgrade, coordinated around the RTAC's existing SCADA and gateway duties.
EtherCAT interface: an SEL-3555 needs a free PCI slot for the SEL-3390E4 Network Adapter Card (or SEL-3390T Time and Ethernet Adapter Card) — this card is the dedicated EtherCAT port to the Axion network.
Licenses: confirm FileIO, DDR, and Continuous Recorder licenses; add ACSELERATOR RTAC SEL-5033 v1.37.153.8000+ and SEL-5601-2 Synchrowave Event on the engineering workstation.
Storage: size the SSD from the channel count (600–700 MB/day/channel at 3 kHz continuous) against your retention target 30 days is the ERCOT floor (20 under NERC alone), 45 days is a comfortable design point.

Step 3 Plan the CT/PT connections (and decide conventional vs. split-core)

At the transformer node, the SEL-2245-42 (3CT/3PT) module takes conventional 1 A/5 A CT secondaries and up to 300 V PT inputs with two-wire connections the standard approach where spare CT cores or shared circuits with verified burden exist. Where they do not most commonly at feeder positions on brownfield sites specify the SEL-2245-43 with non-intrusive split-core CTs: the clamp-on CTs install around existing conductors without breaking a single secondary circuit, and field calibration compensates the gain and phase error the split-core introduces. This one design decision routinely eliminates the largest outage and rewiring cost in a retrofit. Perform a CT burden check on every shared circuit, and land all new wiring through test switches with shorting provisions for safe future maintenance.

Step 4 Mount the nodes and build the EtherCAT network

Each node is a SEL-2242 chassis (rack, panel, or surface mount, modules vertical with 0.5 in clearance) fitted with an SEL-2243 power coupler (dual couplers for redundant AC + DC station supplies), the analog modules, and a digital input module. The network rules are strict and worth memorizing: EtherCAT is non-routable the connection from the RTAC's EtherCAT port to the first node's power coupler Port 1 must be direct, with no Ethernet switches, and subsequent nodes daisy-chain coupler-to-coupler. Copper RJ45 links are limited to under 3 meters (panel-internal only); LC fiber reaches roughly 2 km multimode or 15 km single-mode, and backplanes can be linked as far as 5 km apart which is also what makes remote nodes at inverter skids feasible over site fiber for the unit-level 128 s/c requirement.

Step 5 Wire the digital inputs

Bring every breaker 52A contact, disconnect switch status, relay trip/operate output, and station alarm into the SEL-2244 module at the nearest node. These become 1 ms SER points and for breakers falling-edge fault-record triggers, so a breaker opening captures a full waveform record even if no analog threshold tripped.

Step 6 Configure the DFR in SEL-5033 (no code required)

The DFR Extension turns configuration into a menu-driven workflow SEL documents configuration time dropping from days to under an hour. The sequence:

Enable EtherCAT on one port of the SEL-3390E4 card via the RTAC web interface (the port becomes dedicated to EtherCAT).
Create an RTAC project (firmware R153+) and insert the Digital Fault Recorder extension from the Extensions menu.
Set General Settings: nominal frequency, phase rotation, station name, fault recording rate (24 kHz recommended), record length (e.g., 6 s with 2.5 s pre-trigger margin over the ERCOT 5-second total-record floor and the 2-cycle pre-trigger minimum), continuous-recording retention duration (e.g., 45 days), and enable the synchrophasor server.
Define the Axion nodes: chassis size and the module in each slot, then enter every CT and PT ratio as X:1 (a 1200:5 CT is entered as 240; a wye-connected 700:1 PT as 700).
Define substation assets and associate them with modules: the HV interconnection as a Transmission Line asset (voltage + current + impedance-based fault location), the transformer high and low sides as Generic assets (the asset type SEL lists for transformers, inverters, reactors, and capacitor banks), feeders as Transmission Line assets, and voltage-only points as Bus assets.
Set triggers per asset: voltage triggers (a common pattern is 105% overvoltage / 95% undervoltage with a 16 ms pickup), overcurrent, sequence-component triggers (V1 low, I2/I0 high catch unbalance and ground events), and over/underfrequency triggers (with positive and negative ROCOF) on the main power transformer and feeder assets frequency triggering at the MPT and feeders is explicitly required by PRC-028-1 R3. Enable relay-trip cross-triggering from all protection groups via the digital inputs.
Configure digital inputs (channel names appear in the event records name them properly), enable falling-edge triggers on breaker inputs, and map inverter fault codes, operating-mode changes, and HVRT/LVRT/HFRT/LFRT ride-through flags as custom channels any tag already in the RTAC project (for example from the plant controller or inverter SCADA) can be embedded directly into the DFR records.
Toggle Build DFR to True and save: the extension automatically generates every recording device, Continuous Recording Group, PMU, channel, calculation, and all IEC 61131-3 logic. The finished project must compile with 0 errors and 0 warnings.

Step 7 Commission and prove it end-to-end

A compliant recorder is one you can demonstrate, not just install. The Keentel commissioning sequence:

Time synchronization validation: GPS startup report, IRIG-B signal verification at every device, and timestamp comparison of all IEDs against UTC. The DFR's Synchronized status indication confirms high-quality IRIG-B or PTP at the RTAC and all connected analog modules.
Secondary injection trigger tests at each asset: prove every voltage, overcurrent, sequence, frequency, and digital trigger produces a correctly named COMTRADE record with intact pre-trigger data.
Retrieval demonstration: generate an on-demand Continuous Recording Group file from the RTAC web interface (select date, time, duration, and channels — 3 kHz, PMU, and digital) and download fault records from Event Collection; export SOE and fault-location CSVs from the file manager. This same workflow becomes the documented ERCOT 15-day data provision procedure.
Retention validation: verify storage headroom and the rolling-buffer behavior against the retention target; confirm the DFR Alarm logic (EtherCAT abnormal, CPU burden over 75%, storage under 10%/4 GB, RAM under 10%) is mapped to SCADA.
Documentation closeout: as-left RTAC project backup, settings files, channel schedule, trigger settings document, test reports, and the maintenance & testing plan — the audit evidence package.

Case Studies

Case Study 3 Greenfield Solar + Storage Hybrid: Compliance Designed In, Not Bolted On

Background

A developer building a solar-plus-battery hybrid facility in the 200 MW class solar PV and a co-located BESS sharing a collector substation with two main power transformers and a 34.5 kV bus with six feeder positions plus dedicated storage feeders engaged Keentel during detailed design, before the substation package was released for construction. As a new BES IBR with commercial operation after April 1, 2025, the facility fell under the accelerated new-resource schedule R1 through R7 compliant by July 1, 2026 or commercial operation, whichever is later, with R8 already in effect since January 1, 2026 so disturbance monitoring had to be operational at COD, and the interconnection agreement's ride-through obligations referenced IEEE 2800-aligned ERCOT requirements from synchronization.

Challenge

Two transformers and mixed solar/storage feeders pushed the channel count well beyond a small retrofit the DFR had to be sized, not assumed.
BESS inverters introduce bidirectional flows and fast mode transitions (charge/discharge/idle, grid-support functions); SER had to capture mode changes and ride-through flags with enough fidelity to defend performance during commissioning-era grid events.
The construction schedule offered no room for DFR rework: CT/PT allocations, panel space, fiber routes, and the EtherCAT topology had to be right on the IFC drawings the first time.

Solution

Because the facility was still on paper, compliance was engineered into the base design. Each main power transformer received its own Axion node the one-node-per-transformer pattern with SEL-2245-42 modules on dedicated DFR CT cores and PT windings specified into the equipment procurement (no split-core compromise needed on a greenfield). A third node covered the collector bus and all feeder positions, with slots reserved in each SEL-2242 chassis for the expansion feeders in the site's later phase. The central SEL-3555 was specified from day one with the SEL-3390T time-and-Ethernet card, DFR licensing, and storage sized by calculation: roughly 45 analog channels at 600–700 MB/day each set the SSD requirement, with the platform's 192-analog-channel ceiling and the above-96-channel SEL-3555 requirement both checked against the phase-two build-out. The EtherCAT fiber topology direct RTAC-to-node, no switches, node-to-node daisy chain was drawn onto the raceway and fiber schedules before construction. BESS and PV inverter fault codes, fault alarms, and HVRT/LVRT/HFRT/LFRT ride-through mode status were mapped as DFR custom channels from the plant controller integration — and because these were new units, the “if capable” relief available to existing fleets did not apply: unit-level SER capability per PRC-028-1 R1 was made a factory-acceptance requirement with both inverter OEMs during procurement rather than discovered after energization. Fault recording was configured at 24 kHz with 6-second records, and retention at 45 days against the rolling 30-day ERCOT floor. Redundant time sources (GPS primary with a second source under the platform's PTP/IRIG redundancy support) eliminated the single-point time-base failure.

Implementation Highlights

Zero DFR-driven change orders during construction the CT/PT, panel, and fiber provisions were on the IFC set from the start.
DFR commissioning ran inside the substation commissioning window; the end-to-end trigger-to-COMTRADE-to-retrieval demonstration including Attachment 1 SER CSV export and the 7-calendar-day ERCOT provision workflow was witnessed as part of energization testing, and the data-provision procedure, R8-ready maintenance & testing plan, and DME Equipment Reporting Database were in the operator's hands at COD.
During an area grid disturbance in the first months of operation, the recorder captured time-aligned 24 kHz records and inverter mode flags across both transformers — the facility demonstrated compliant ride-through from its own data, closing the inquiry without a corrective-action process.

Results

Over the past decade, several major grid disturbances including the Blue Cut Fire, Canyon 2 Fire, and the Odessa events in Texas revealed a systemic problem: when inverter-based resources (IBRs) such as solar, wind, and battery storage plants tripped or reduced output during grid faults, operators and engineers frequently had no high-resolution data to explain what happened. Protection relays at conventional plants had long provided oscillography, but IBR facilities were often built with minimal recording capability, no unit-level visibility, and inconsistent time synchronization. Event investigations stalled, models could not be validated, and the same failure modes repeated across the fleet.

Two regulatory instruments now close that gap. NERC PRC-028-1 Disturbance Monitoring and Reporting Requirements for Inverter-Based Resources became effective April 1, 2025 and establishes mandatory Sequence of Events Recording (SER), Fault Recording (FR), and Dynamic Disturbance Recording (DDR) for applicable IBRs across North America. In Texas, ERCOT NOGRR255 amended the Nodal Operating Guides effective August 1, 2024 with closely aligned disturbance monitoring equipment (DME) requirements, plus obligations for data retention, data provision to ERCOT, and ongoing maintenance and testing. Together with the ride-through requirements of NOGRR245 and the performance standards PRC-029 and PRC-030, these rules create a closed loop: high-resolution data captures what happened, ride-through criteria define what should have happened, and corrective-action processes fix the difference.

The practical question for asset owners is no longer whether to install disturbance monitoring, but how to do it economically at dozens of sites that were never designed for it. This guide presents the architecture Keentel Engineering deploys most often for brownfield and greenfield IBR facilities: the SEL RTAC + SEL-2240 Axion Digital Fault Recorder, a solution SEL documents as designed to exceed compliance with NERC PRC-002 and PRC-028.

Both frameworks organize disturbance monitoring into three data categories, each with specified locations, trigger conditions, and performance minimums.

3.1 Sequence of Events Recording (SER)

SER is the time-stamped digital log of discrete events. PRC-028-1 R1 requires circuit breaker positions for breakers associated with the main power transformer (s), collector bus(es), and shunt reactive devices, and at the IBR unit level all fault codes, all fault alarms, and high/low voltage and frequency ride-through mode status, recorded when triggered by ride-through operation or tripping. An important nuance in the final standard: for IBR units already in commercial operation before the effective date, the unit-level data applies if the units are capable and acceptable evidence includes equipment specifications or a letter from the manufacturer documenting capability or the lack of it. Timing accuracy under R6 is ±1 millisecond of UTC for recording devices and ±100 milliseconds for IBR unit device clocks; ERCOT's NOGRR255 language is tighter still (GPS-based sub-microsecond timing for fault recording), so in Texas the design should be built to the ERCOT numbers.

3.2 Fault Recording (FR)

FR is triggered oscillography raw AC voltage and current waveforms captured around a disturbance at the main power transformer high side, the collector feeder breakers, and any shunt dynamic reactive devices. PRC-028-1 R2 requires phase-to-neutral voltages, phase and residual currents, and real and reactive power on a three-phase basis for each of those elements. The NERC performance floor in R3 is 64 samples per cycle but NOGRR255 raises it to 128 samples per cycle for any fault recording equipment installed or replaced after January 1, 2024 (legacy equipment set as close to 128 as it allows). Records must include at least 2 cycles of pre-trigger data, with a total record length of at least 2 seconds under PRC-028-1 and at least 5 seconds under NOGRR255 design to the ERCOT number. Mandatory triggers: neutral (residual) overcurrent (NOGRR255 sets the pickup at 0.2 p.u. of rated CT secondary or less), AC phase undervoltage (below 0.9 p.u. for two cycles or longer), AC phase overvoltage (above 1.1 p.u. for two cycles or longer), phase overcurrent at 1.5 p.u. or less or protective relay tripping for all protection groups (cross-triggering from relay operations), and under PRC-028-1 R3 over- and underfrequency at the main power transformer and collector feeders. A design note on unit-level oscillography: early PRC-028 drafts contained a 128 s/c fault-recording requirement at IBR units on the last 10% of each collector feeder, and that language circulated widely but the final adopted standard does not include unit-level fault recording; the final unit-level obligation is the SER-type data described in 3.1. Verify the corresponding ERCOT unit-level language against the final approved Operating Guide before scoping inverter-skid recording hardware.

3.3 Dynamic Disturbance Recording (DDR)

DDR captures the slower electromechanical and control-system dynamics continuously: per PRC-028-1 R4, for each main power transformer — one voltage, the corresponding current, real and reactive power on a three-phase basis, and frequency. R5 sets the performance: input sampling of at least 960 samples per second and an output recording rate of at least 60 times per second. Continuous recording with a rolling retention buffer is the cleanest way to satisfy it.

3.4 Retention, Format, Delivery, and the Obligations People Miss

PRC-028-1 R7 requires data retrievable for 20 calendar days and provided within 15 calendar days of a request; NOGRR255 goes further a rolling 30-calendar-day retention and delivery within 7 calendar days of a request unless the requestor grants an extension. Design to the ERCOT numbers and both are satisfied. Formats are fixed: SER in ASCII CSV per the standard's Attachment 1 field order (Date, Time, Local Time Code, Plant Name, Device, State), FR/DDR in IEEE C37.111 COMTRADE (1999 revision or later) or CSV, files named per IEEE C37.232 COMNAME (2011 or later). Three obligations owners routinely miss: (1) any data actually provided to ERCOT, the Regional Entity, or NERC must then be stored for at least three years; (2) NOGRR255 requires a current DME equipment database location, type, make/model, operational status, monitored equipment, and the complete monitored-points list deliverable within 30 days of request; and (3) under PRC-028-1 R8, a discovered failure of recording capability must be restored within 90 calendar days or covered by a Corrective Action Plan submitted to the Regional Entity, while NOGRR255 adds a 30-day data-availability verification cycle (or an automated failure-notification system) and 30-day failure reporting. These are operational programs, not hardware which is why the deliverable set must include a data-provision procedure and a maintenance & testing plan, not just a commissioned recorder.

PJM EMT Model Development Guidelines for IBR Projects

Fri, 03 Jul 2026 15:33:58 GMT

Jul 03, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

PJM's New EMT Model Development Guidelines for Inverter-Based Resources

The guideline applies to all IBR solar PV, BESS, Type III (DFIG) and Type IV wind, and co-located or hybrid plants such as PV-plus-BESS. The submission trigger depends on where your project sits in the process:

Two traps hide in this timeline. First, the as-built stage is not a copy-paste of the as-designed submission it adds a Parameter Verification Report comparing installed PPC and inverter settings against the model, and any divergence between studied and installed settings is treated as a qualified change under FAC-002-4, reviewed before commercial operation. Second, hybrid conversions are resubmission events. Adding a BESS to an operating solar plant re-opens the entire deliverable set.

When PJM Will Actually Run EMT Studies

Not every project will face a full EMT interconnection study. In the planning horizon, PJM performs EMT analysis for three scenario classes:

Low system strength projects electrically located in weak areas where PSPD results are unreliable by construction.
Poor stability performance in PSPD studies when the positive-sequence model already shows marginal damping or questionable recovery, EMT is used to look closer.
Topology or conditions with stability risk configurations such as high local IBR concentration, series compensation exposure, or unusual outage conditions.

The critical point for developers: PJM decides whether an interactions study is required based on project size and electrical location but the model must be submitted regardless, so that the option is always on the table without restarting the clock.

Who Must Submit, and When: The Compliance Timeline

The 18 Model Quality Tests, Decoded

What the PSCAD Model Must Contain

The appendix checklist is the technical heart of the guideline. Every item must be reviewed and marked by the model supplier, with a written explanation for any non-conformance, and the checklist is signed by an authorized representative. The items group into four families:

Documentation and configuration control

The model must carry an identification mechanism control revision codes, settings files, or equivalent so that during commissioning PJM can verify the field settings match what was studied. Vendor name and model version must be visible in the .pscx case, and a site-specific sample test case (single-machine infinite bus at a representative SCR) must be included.

Accuracy features

This is where generic library models die. PJM requires the full detailed inner control loops of the power electronics the approximation used in transient-stability modeling is expressly insufficient. The strongly recommended path is embedding actual hardware code into a PSCAD component (a “real code” model). Models assembled from master-library blocks or translated from block diagrams require validation against hardware performance, and against field measurements at the as-built stage. Additional accuracy items: a full IGBT or firing-pulse-based switching representation (a three-phase sinusoidal source is not acceptable), plant-level control with real communication delays and ride-through mode transitions, SSO mitigation with enable/disable capability, transformer magnetizing curves, dynamic reactive devices, DFIG machine slip representation, and detailed balanced/unbalanced protection systems implemented in actual hardware code.

Usability features

Study engineers must be able to work with the model: user-accessible control and hardware options, diagnostic flags that identify why a model tripped, protection functions that can be disabled, scalable active power capacity, dispatch capability below nameplate, external reference variables changeable mid-simulation, self-initialization to ordered output without hand-holding, and tolerance of the 5–20 µs time-step range (10 µs recommended).

Efficiency features

Compatibility items that make PJM's study workflow scale: no dependence on a specific Fortran or PSCAD version (5.0.1+; Intel OneAPI HPC compiler support recommended), Visual Studio 2015+ compilation, multiple-instance support, PSCAD “timed snapshot” and “multiple run” support, no global variables, no multiple layers, replication via copy/copy-transfer, and named modeling-support contacts.

In our experience, checklist items 4 (real-code inner controls), 7 (plant-level control fidelity), 13 (protection systems), and 19 (self-initialization) account for the majority of model rejections and resubmission cycles across ISOs and utilities.

Section 4 of the guideline defines a standardized test battery. Every test reports total P and Q, per-phase RMS voltage (positive sequence when benchmarking against PSS/E), and frequency at both the POI and the Point of Connection, at two zoom levels full simulation and a 1–2 second window around the disturbance. For BESS projects, all tests run at both full charge and full discharge; hybrid plants must cover five operation modes (generation alone; BESS alone charging; BESS alone discharging; generation plus BESS charging; generation plus BESS discharging), with written justification for any mode that does not apply.

PSCAD vs. PSS/E: The Benchmark Requirement

Ride-Through Performance Tests (Tests 7–18): Does the Plant Survive?

How Keentel Engineering Can Help

The 35-Item Requirements Checklist: Where Model Quality Is Really Defined

Keentel Engineering provides end-to-end EMT deliverable support for PJM and other ISO/RTO interconnection processes: PSCAD model development and OEM real-code integration, execution of full model quality test batteries, PSCAD-PSS/E benchmarking, Model Requirements Checklist certification, weak-grid EMT interconnection studies, as-built parameter verification, and NERC MOD-026/MOD-027 model tuning. Our team holds P.E. licensure and IEEE Senior Member standing, with offices in Tampa, FL and Austin, TX.

PJM's stated rationale mirrors what grid operators worldwide have learned the hard way: inverter-based resources (IBRs) connect to the grid asynchronously through power electronics, and their control systems depend on a stable terminal voltage magnitude and angle to operate. In weak grids — areas with low short-circuit strength — that terminal voltage becomes highly sensitive to the IBR's own current injections. The result can be adverse controller reactions, sub-synchronous oscillations (SSO), subcycle overvoltages, and destabilizing interactions between neighboring IBR plants.

Conventional positive-sequence phasor-domain (PSPD) tools such as PSS/E are built for the classical stability band of roughly 0.1 Hz to 10 Hz. They are the right tool for rotor-angle and voltage stability, but they simply cannot represent switching transients, torsional interactions, harmonics, fast inner-loop control dynamics, or phase-locked-loop (PLL) behavior during severe voltage depressions. Those phenomena live in the electromagnetic transient domain — microsecond time steps, three-phase instantaneous waveforms, and full converter control representations.

The regulatory scaffolding has been converging on this conclusion for years. IEEE 2800-2022 defines ride-through, reactive current support, and fast-response capabilities for transmission-connected IBRs that only EMT simulation can meaningfully verify. NERC's reliability guidance on EMT modeling for BPS-connected IBRs, the May 2025 Level 3 Alert on IBR performance, and the MOD-026-2 requirement for Transmission Planners and Planning Coordinators to jointly define dynamic model requirements for both PSPD and EMT models all push in the same direction. PJM's guideline is the region-specific implementation of that national trajectory.

The Interconnection Rulebook Just Changed Again

In March 2026, PJM's System Planning Division released Revision 0 of its EMT Model Development Guidelines for Inverter-Based Resources a document that formally makes detailed electromagnetic transient (EMT) modeling a standing requirement of the PJM interconnection process. For developers of utility-scale solar, battery energy storage (BESS), Type III and IV wind, and hybrid plants, this is not an incremental paperwork update. It changes what must be delivered, when it must be delivered, and how rigorously it will be tested before your project can advance through the queue.

The headline: beginning with Transition Cycle 2, PJM will screen projects connecting in low-system-strength areas and require a PSCAD model at Decision Point II. From Cycle 1 onward, every BPS-connected project must submit a PSCAD model at Decision Point II accompanied by a 35-item Model Requirements Checklist, a Model Quality Test report spanning 18 defined test cases, and a PSCAD-to-PSS/E benchmark demonstrating that your EMT and positive-sequence models tell the same story.

If your project touches the PJM queue in 2026 or beyond, an EMT model is no longer a contingency deliverable you might be asked for. It is part of the definition of a complete submission.

This article walks through the full guideline in depth: why PJM is doing this, who must comply and on what timeline, what the PSCAD model must contain, how the 18 quality tests work, what the cross-platform benchmark demands, and where submissions most often fail. A companion FAQ follows at the end.

The Deliverable Package and Submission Logistics

Case Study 1 — 200 MW Solar PV in a Weak-Grid Pocket: Passing the Model Quality Battery at SCR 3

Background

A utility-scale solar developer held a queue position for a nominal 200 MW (AC) PV facility interconnecting to a 230 kV transmission system in a region with limited nearby synchronous generation. Grid screening identified the point of interconnection as a low-system-strength location the composite short-circuit contribution at the POI yielded an effective SCR near 2.8 at full plant output, below the guideline's default test value of 3. A PSCAD model, the 35-item Model Requirements Checklist, and a complete Model Quality Test and Benchmark Report were required at the project's next decision point.

The Challenge

The inverter OEM's public PSS/E library model existed, but the corresponding real-code PSCAD model had to be obtained under NDA, configured for the site, and validated — with roughly ten weeks of runway.
The plant used two inverter designs (a legacy unit for the first phase block and a newer high-voltage unit for the balance), obligating two distinct aggregated units rather than one lumped equivalent.
Initial flat-start runs at the default SCR = 3 / X/R = 5 setup failed the most basic test in the battery: the plant-level controller hunted around its voltage setpoint with sustained ±0.8 MVAR oscillation — eight times the 0.1 MVAR steady-state tolerance.

Engineering Approach

The team first reconciled the plant electrical data — inverter nameplates, pad-mount and main power transformer impedances, and the equivalent collector branch derived from the as-designed cable schedule between the PSS/E case and the PSCAD model, eliminating the divergence that would otherwise contaminate every downstream benchmark. The two aggregated units were built with the OEM real-code components, each behind its own aggregated unit transformer and collector equivalent, joined at the collector bus ahead of the main step-up transformer and tie line to the POI.

The flat-start oscillation was traced to interaction between the power plant controller's voltage-regulation gain and the weak-grid voltage sensitivity: gains tuned by the OEM at a strong test bench (SCR ≈ 10) were unstable at SCR 3. Working within the user-accessible parameter set the checklist requires the model to expose, the PPC voltage-control response time and droop were retuned, and the coordination between the PPC and the inverters' local volt-VAR loops was corrected so the two control layers stopped fighting. The retuned settings were documented as the settings of record for both models an early payoff of the checklist's identification-mechanism requirement.

With the model stable, the full battery was executed: functional Tests 1–6, fault ride-through Tests 7–9, PRC-024-3 profile Tests 10–12 and 17–18, the system strength sequence of Test 13, phase-angle Tests 14–15, and the ROCOF sequence of Test 16 — each plotted at POI and point of connection, full-window and transient zoom.

Results

Lessons Learned

Vendor bench tuning is not grid tuning. Any model destined for a weak POI should be shaken down at SCR = 3 on day one — Test 1 is the cheapest possible place to discover a controller problem.
Two inverter types means two aggregate models. Budget the extra configuration and testing time up front rather than discovering the requirement in checklist review.
Retunes must propagate to both platforms. A PSCAD-only fix guarantees a benchmark failure later.

Two design details deserve emphasis. First, momentary cessation is treated as a defect, not a feature: for the profile tests it must not be observed at all, and where equipment limitations make it unavoidable, the developer must disclose it to PJM in the test report. Second, the damping ratio floor of 0.3 is a quantitative bar oscillatory recoveries that would have skated through a qualitative review now fail on the numbers, and the required transient-zoom plots make marginal damping impossible to hide.

Lessons Learned

Hybrids multiply everything. Five modes × two power extremes × 18 tests is unmanageable without automation build the scripting before the testing, not during it.
Charge-mode ride-through is a distinct behavior. A BESS that rides through faults while discharging can still fail while charging; both directions must be verified explicitly.
Generic PPC models rarely survive contact with a multi-OEM hybrid. Plan for real-code plant control from the outset.
Diagnostic flags are not bureaucratic garnish they turned a mystery trip into a two-week fix instead of a two-month vendor escalation.

Case Study 2 — Solar-Plus-Storage Hybrid: Conquering the Operation-Mode Matrix and a Momentary Cessation Finding

Background

A hybrid facility pairing approximately 150 MW of solar PV with a co-located 4-hour BESS behind a shared POI required a complete EMT deliverable package. As a hybrid, the guideline's full operation-mode matrix applied: every quality test executed for generation alone, BESS alone charging, BESS alone discharging, generation plus BESS charging, and generation plus BESS discharging with BESS cases at both full charge and full discharge power.

The Challenge

The raw test matrix exceeded 90 simulation cases before benchmark reruns — a volume problem as much as an engineering one.
The PV inverters and BESS inverters came from different OEMs with different real-code models, coordinated by a third-party power plant controller whose vendor initially proposed a generic PPC block.
During LVRT profile testing (Test 10), the BESS inverters entered momentary cessation at the deepest step of the PRC-024-3 low-voltage profile while in charging mode — a direct violation of the test's acceptance criterion that momentary cessation not be observed.

Engineering Approach

The generic PPC proposal was rejected early on checklist grounds: generic representations are acceptable only when the final controller exactly matches the generic model, and this multi-OEM hybrid demonstrably did not. The PPC vendor's actual control firmware was embedded as a real-code component, capturing the measurement filtering, communication latency between the PPC and the two inverter fleets, and the hybrid dispatch logic that arbitrates solar and storage output at the shared POI limit.

The momentary cessation finding was dissected at the inverter level using the diagnostic flags the checklist requires models to expose. The trip signature showed the BESS units' DC-side protection interacting with the charging power flow: at the profile's zero-voltage step, the units blocked current while charging although they rode through correctly while discharging. Working with the OEM, the ride-through parameter set was corrected — the low-voltage current-blocking threshold and the charge-mode ride-through behavior were aligned with the unit's actual certified capability, which supported continuous current injection. Retesting confirmed reactive current injection through the depressed-voltage window with no cessation in any operation mode.

To manage case volume, the team scripted the matrix: automated case generation, batch execution using PSCAD's multiple-run capability (itself a checklist item), and automated extraction of P, Q, RMS voltage, and frequency at both measurement points into the standardized two-zoom plot format. Human review time went into interpreting marginal results instead of producing figures.

Results

Frequently Asked Questions

Lessons Learned

Parameter drift is the rule, not the exception. Firmware updates and commissioning adjustments accumulate silently between Decision Point II and COD — schedule the verification early enough to absorb what it finds.
Quantitative reruns beat qualitative arguments in a qualified-change review. A targeted retest package converts a schedule threat into a documented non-issue.
Treat the one-year MOD-026/MOD-027 tuned-model obligation as part of the same campaign; the marginal cost of capturing the right commissioning data is small, and the cost of recreating it later is not.

Why EMT? The Physics Behind the Policy

PJM requires the project to be modeled up to the Point of Interconnection (POI) using an aggregate or partially equivalent plant representation. Where a single physical plant is split into multiple queue positions, PJM explicitly prefers that the whole-plant model accompany each individual submission. If a plant uses more than one inverter type or design, each type gets its own aggregated unit; lumping dissimilar inverters into one equivalent machine is not acceptable.

The aggregate model must include, at minimum:

Aggregated generation or energy storage model
Aggregated generator (unit) transformer
Equivalent collector branch (line or cable equivalent)
Main step-up transformer(s)
Generator tie line to the POI
Supplemental devices where present STATCOMs, synchronous condensers, capacitor and reactor banks, harmonic filter banks
Station service load

The external grid is represented by an AC source behind a Thevenin impedance, with a default short-circuit ratio (SCR) of 3 and X/R of 5 unless a specific test dictates otherwise. Models must be delivered in PSCAD™/EMTDC™-compatible formats (*.pswx, *.pscx, *.pslx).

Functional Tests (Tests 1–6): Does the Model Behave?

Because PJM's planning workflow still runs on PSS/E, the guideline requires proof that the EMT and PSPD models agree. For as-designed submissions (cycle process, Necessary Study), PSCAD and PSS/E signals must be overlaid and aligned on every plotted signal for Tests 1–6 (functional), Test 7 (3LG), Test 10 (LVRT), and Test 12 (HVRT). For as-built submissions, the benchmark expands to the full quality-test set minus four cases that positive-sequence tools cannot legitimately represent: SLG (Test 8), line-to-line (Test 9), and both ±25° phase-angle tests (Tests 14–15). The as-built package must also include the PSS/E UDM and library model if not previously submitted.

The intent is stated plainly: the two models must align well enough that PJM can pivot to the PSCAD model to chase issues surfaced in PSS/E studies without questioning whether it represents the same plant. In practice this benchmark is where mismatched droop settings, inconsistent transformer data, and divergent protection thresholds between your PSS/E and PSCAD models get exposed.

Case Studies

Everything is zipped and uploaded through the Data Submission Platform under Generator Information > Stability Models for Queue Point projects or Project Capability Tab > Dynamic Files for NextGen with a 250 MB upload limit (larger packages go through Secure Share via your Project Manager). Note the phrase “all cases and files”: PJM reserves the right to rerun any test, so the submission must be reproducible, not just reported.

Where Submissions Fail and How to Get Ahead of It

Having built, tested, and benchmarked PSCAD and PSS/E models across ISO/RTO processes, we see the same failure modes recur:

Vendor model lead time. Real-code PSCAD models come from the inverter OEM, often under NDA, sometimes weeks after request. Projects that wait until Phase II to open that conversation compress their own testing window to nothing.
Flat-start failures at SCR 3. Models tuned in vendor test benches at high SCR frequently hunt or oscillate at PJM's default weak-grid setup. Test 1 is deceptively simple and fails more often than any ride-through case.
PSS/E-PSCAD divergence. The two models are usually built by different teams from different data snapshots. Reconciling MVA bases, transformer impedances, PPC droops, and protection settings before running the benchmark saves an entire resubmission cycle.
Hybrid mode coverage gaps. Five operation modes, at charge and discharge extremes, across 18 tests multiplies quickly. Missing modes without documented justification is an automatic deficiency.
As-built parameter drift. Commissioning settings that deviate from the studied model become FAC-002-4 qualified changes on the eve of commercial operation the most expensive possible moment to discover them.

Keentel Engineering supports developers, EPCs, and owners across exactly this workflow: PSCAD model development and vendor-model integration, execution of the full 18-test quality battery, PSCAD-PSS/E benchmarking, checklist certification support, EMT interconnection studies for weak-grid POIs, and as-built parameter verification. If PJM's new guideline intersects your queue position, the time to scope the modeling effort is before Decision Point II is on the calendar not after the deficiency letter arrives.

Case Study 3 Standalone BESS at As-Built Verification: Parameter Drift, Benchmark Expansion, and a Qualified-Change Save

Background

A standalone transmission-connected BESS in the 100–150 MW class approached commissioning. The as-built stage carries the guideline's heaviest evidentiary burden: a PSCAD model reflecting installed site parameters, the PSS/E UDM and library model, an expanded Model Quality Test and Benchmark Report covering more tests than the as-designed stage, and a Parameter Verification Report demonstrating that PSCAD model parameters both PPC and inverter match the equipment actually configured in the field.

The Challenge

Commissioning records revealed drift between studied and installed settings: the field PPC carried a different voltage-droop characteristic and a slower active-power recovery ramp than the as-designed model, and the inverters had received an OEM firmware update that revised two protection thresholds.
The as-built benchmark scope expands beyond the as-designed set (functional tests plus 3LG, LVRT, HVRT) toward the full battery, excluding only the unbalanced faults and phase-angle jumps that positive-sequence tools cannot represent — meaning ROCOF, the frequency ride-through profiles, and the system strength test all now required PSCAD/PSS/E overlays.
Any unresolved difference between studied and installed settings would constitute a qualified change under FAC-002-4, requiring review before commercial operation — with the COD date already fixed in offtake commitments.

Engineering Approach

A line-by-line Parameter Verification Report was built first, not last: every user-settable PPC parameter and every inverter setting exposed by the real-code model was tabulated against the commissioning configuration exports, with three disposition categories — matched, model updated to field value, and field value flagged for engineering review. The droop and ramp-rate changes fell in the second category: the PSCAD and PSS/E models were updated to the installed values, and the delta was assessed for study impact.

The firmware-revised protection thresholds required more care. One threshold change was benign; the other narrowed an undervoltage protection band in a way that could plausibly interact with ride-through performance. Rather than argue the point qualitatively, the team reran the affected ride-through tests with the installed thresholds. Results showed the plant still met every acceptance criterion — damping ratios above 0.3, no cessation, recovery inside limits — and the rerun results were packaged as the technical basis for the qualified-change review, allowing the reviewing parties to confirm existing studies remained valid without new full-scope studies.

The expanded benchmark then proceeded on the updated, verified model pair: charge and discharge runs for the functional suite, 3LG, both PRC-024-3 voltage profiles, both frequency profiles, ROCOF, and the system strength sequence, with PSCAD and PSS/E traces overlaid per signal. The final package — model files, checklist, test and benchmark reports, parameter verification report, and every runnable case — was assembled under the 250 MB platform limit by externalizing raw output data to a Secure Share transfer coordinated with the project manager.

Results

Power Transformer Testing and Commissioning Guide 2026

Thu, 02 Jul 2026 12:46:13 GMT

Jul 02, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Power Transformer Testing &Commissioning

Recording nameplate data is not a test, but it is mandatory — and it forces the test crew to become intimately familiar with the unit. The main tank, the under-load tap changer, bushing CTs, bushings, fuses, fan and pump motors, surge arresters, and disconnect switches all carry individual nameplates, and every pertinent field on the test data sheet should be completed.

Terminal marking follows ANSI convention. Two-winding transformers use H for the higher-voltage winding and X for the lower-voltage winding; viewed from the high-voltage side, the H1 bushing is on the right. Transformers with three or more windings add Y and Z designations in order of decreasing voltage rating, with H assigned to the highest-voltage (or, for equal voltages, highest-kVA) winding.

Because transformer BIL is relatively low, a surge arrester belongs on every transformer bushing connected to a line or high-voltage bus. Each arrester and each section of stacked arresters should be meggered at 2,500 VDC. Verify the nameplate kV rating matches the application (mismatches cause failures and outages), confirm stacking order against manufacturer recommendations, and record all nameplate data. Hardware matters: grading rings for arresters 115 kV and above, proper standoff insulators where required, and ground connections routed so discharges actually pass through the surge counter arresters have been installed with counters inadvertently shorted past. Newer metal-oxide voltage limiters often omit counters entirely, since they clip surges faster than counters can respond.

4. Auxiliary Wiring Checks and Hand Meggering

The Keentel Perspective

Commissioning test data is not just a gate to energization — it is the fingerprint of the transformer in its healthy, as-installed condition. Every ratio, impedance, winding-resistance, and insulation reading taken on day one becomes the diagnostic baseline against which every future maintenance test, post-fault assessment, and end-of-life decision will be compared. Sloppy baseline data devalues the asset for its entire service life.

2. The Recommended Test Sequence

Transformer testing is most efficient and safest when it follows a deliberate sequence. A proven field sequence looks like this:

Inspect the transformer and all shipped components for shipping damage and moisture ingress.
Verify nameplate data and drawings for correct voltages and the external phasing connection to the line or bus.
Check calibration of all thermal gauges, the hot-spot heater circuit, and RTDs, and verify alarm contact settings (typical staging: one cooling stage running continuously, second stage near 80 °C, third stage near 90 °C, hot-spot alarm near 100 °C with trip near 110 °C where applicable).
Point-to-point wire check and megger all transformer wiring: fans, pumps, alarms, heaters, tap changers, and interconnecting cables.
Vacuum-dry large banks (a common threshold is units above roughly 150 MVA). Never apply test voltage to windings during vacuum drying, and keep terminals shorted and grounded during oil circulation because of static charge buildup on the winding.
After oil filling, confirm an oil sample has gone to the chemistry lab and record oil level and temperature at completion.
Power-operate pumps and fans to verify rotation, and exercise the under-load (UL) tap changer, heaters, and alarms.
Perform winding tests: ratio and polarity (TTR preferred for units above 1 MVA), impedance, DC winding resistance, and megger/power-factor of windings, bushings, and arresters waiting 24 hours after oil filling before power-factor testing.
Load CT circuits overall and flash CTs for polarity.
Trip-check all bank protection schemes and confirm the gas-collection relay is free of gas before energization.
Energize while monitoring bank currents, voltages, and UL tap-changer operation; verify phasing to the system before picking up load. Where possible, let large units soak energized for eight hours before carrying load.
Make in-service checks on meters and relays, release to operations, and turn in complete test reports including all test data, moisture and oil data, problems encountered, and in-service readings.

3. Nameplate Data and Terminal Markings

6. Insulation Power Factor Testing

Every wire on the transformer should be verified before energization control panels, terminal cabinets, cables routed to the unit, and every screw, nut, and bolt terminal torqued for tightness, including CT leads where they originate at the bushing connection boxes. A thorough wire check protects a unit that may be difficult, expensive, or impossible to replace; it also cross-checks the drawings against physical reality and familiarizes the crew with the equipment.

All low-voltage wiring should be meggered at 250 or 500 VDC. The emphasis is deliberate: large transformers carry numerous small terminal boxes interconnected by conduit that can accumulate moisture or leak, and wire pulled through metal conduit is occasionally scraped down to bare copper. Boxes mounted on vertical surfaces should have small drain holes; horizontal-surface boxes need sound weather seals, and any questionable gasket should be replaced. Completing wire checks and low-voltage meggering early allows control and alarm power to be applied safely which in turn makes tap-changer and cooling-system testing far faster than hand-cranking mechanisms.

5.1 Ratio Verification

Bushing CTs are best tested by the Current Ratio method — ideally before they are mounted, or with high-capacity current leads pulled through the CT windows before bushings are inserted. Where primary current injection is impossible, a Tap Voltage Ratio test (energizing the full winding and measuring the drop across each tap) can verify tap ratios, since voltage ratios are directly proportional to turns ratios. But the voltage method is a last resort: it cannot establish the true installed polarity or verify the primary-to-secondary ratio, and testing at rated current always provides more assurance that the CT will perform in service.

5.2 Flashing CTs for Polarity

CT polarity is commonly verified by “flashing” — applying 6–12 VDC to the transformer bushings and making/breaking the circuit with a hot stick. Battery positive connects to the polarity (high-voltage) terminal; battery negative is touched to the non-polarity end or station ground through the hot-stick lead. An analog voltmeter across the CT secondary (polarity side to X1) should deflect upscale on make and downscale on break if polarity is correct. Series current-limiting resistance is advisable with automotive batteries, and shorting the windings on the side opposite the flash limits arc energy.

Safety Warnings — DC Flash Testing

A winding carrying DC current generates a large voltage kick when the circuit is opened. Always make and break the connection with an insulated hot stick never by hand.

Never make or break connections directly on a lead-acid battery terminal; charging batteries evolve hydrogen and have exploded from connection sparks.

Never place your body in series with the test circuit by holding a battery clip in one hand while touching the terminal with the other.

Flashing does leave residual magnetism in the CT core which, in theory, could contribute to CT saturation and improper relay operation on initial energization. After a high-current DC flash, demagnetize the core: apply AC excitation gradually, drive the CT just into saturation with the secondary open, then slowly reduce the AC quantity to zero.

Bushings first

Every bushing should be power-factored before insertion into the transformer. If no power-factor set is available during assembly, at minimum measure C1 and C2 capacitance with a bridge — a capacitance shift can flag a serious internal bushing problem before it is buried in the tank — and megger the bushing and its tap at 2,500 V (checking first that small bushing taps may be limited to 500–1,000 V).

Then the transformer

Power-factor the complete transformer soon after drying and oil filling, and re-test all bushings, since readings shift slightly after assembly. A complete data set includes winding-to-winding, winding-to-ground, and bushing tests, plus a winding excitation test where the test set has sufficient capacity. Critically, wait 24 hours after completion of oil filling before power-factor testing — oil and insulation need time to reach equilibrium for meaningful readings.

7.1 Subtractive vs. Additive Polarity

Winding polarity describes the relative direction of instantaneous voltage at the terminals, determined by winding direction and lead routing. If adjacent HV/LV terminals share the same instantaneous polarity, the transformer is subtractive; if diagonally opposite terminals match, it is additive. Polarity matters whenever windings or transformers are paralleled, and when establishing three-phase connections. As a rule, transformers above 500 kVA are built subtractive.

7.2 DC Flashing Method

With a small dry-cell battery intermittently connected to the HV winding and a DC voltmeter on the LV terminals, an upscale kick on make and downscale kick on break indicates subtractive polarity; the reverse indicates additive. The inductive kick on break is much larger than on make. Crisp “make/break” call-outs between the battery operator and meter observer, a healthy battery, and low-resistance bushing connections make deflections easy to read.

7.3 AC Method

Energize the HV winding, jumper the two adjacent HV/LV bushings together, and read a polarity voltmeter across the open pair: VPOL = VH − VX indicates subtractive polarity; VPOL = VH + VX indicates additive. The AC polarity check is conveniently performed during the voltage-ratio test.

7.4 Voltmeter Ratio Method

The voltmeter method uses a fused 115 VAC source, a variable autotransformer, an ammeter, and precision voltmeters (0.5%-class or 4½-digit DVMs for 1% ratio accuracy). Test voltage is typically 1/1000 of winding rating or less, which makes evaluation intuitive: if the nameplate tap says 112,750 V : 13,800 V, energize the HV winding at 112.75 V and expect 13.80 V on the low side. Read every tap-changer position, pre-calculate expected values on the data sheet, connect voltmeter leads directly to bushing terminals (never through the source leads), and monitor input current to catch shorted turns or connection problems immediately.

7.5 TTR — The Preferred Method

The Transformer Turns Ratio test set compares the unit under test against an adjustable-ratio reference transformer using a null detector, determining polarity by comparison in the same operation. TTR is the preferred ratio method, with accuracy on the order of 0.1%, and should be performed on every new high-voltage power transformer at installation and on any unit that has been overhauled or relocated. Its repeatability makes it the diagnostic base for future maintenance comparisons — and the single most useful data set when a shorted turn is suspected. The voltmeter method still earns its place: it verifies proper tap-changer make/break behavior and converts easily into the impedance test setup.

High-Ratio Hazard

Extreme caution is required when test-energizing high-ratio transformers. Even low test voltages applied to a low-voltage winding are stepped up at the HV terminals — for example, 120 V applied to the wye side of a 230 kV delta / 13.8 kV wye bank can generate roughly 3.5 kV between H-side bushings. Apply test voltage to the highest-voltage winding whenever the hands-on nature of the test allows, fence off exposed areas, and warn all personnel.

7. Ratio and Polarity Testing

Impedance measurements — compared against nameplate and retained for future diagnostics — detect shipping damage, loose or high-resistance connections, and in some cases shifting of the core and coils from mechanical or electrical damage. A short-circuit impedance test should be performed on:

Installation as a new unit
Any overhaul
Any internal connection change (other than tap-changer operation)
Reinstallation at another location
Any unit subjected to a severe fault where winding or tap-changer damage is possible

Because power transformers are highly linear, impedance can be measured at rated frequency using currents and voltages that are a small fraction of rating. Practical rules: use a sinusoidal (preferably regulated) source above 7 V, RMS meters of 0.5% class or 4½ digits, measure voltage directly at the transformer terminals, and record ratio and impedance for every tap position. Short-circuit jumpers must be heavy (2/0 or 4/0), short, and clamped tight — a high-resistance short reads as high impedance. In practice the resistance of heavy shorting leads introduces well under 0.5% error even on high-ratio windings; contact resistance is the more likely culprit.

Acceptance guidance:

measured impedance should compare with nameplate within about 5% (nameplate comparison is only valid on the nominal tap); if it deviates more, re-measure from the alternate direction or use phase-to-phase methods. Comparisons between phases of the same unit should agree within about 2%. On a 9.0% impedance transformer, expect field values between roughly 8.8% and 9.2%.

8.1 Connection-Specific Nuances

Percent-impedance formulas depend on winding connection, and this is where mistakes happen. A few field-critical points:

Delta-delta units read double. A single-phase impedance measurement on a delta-delta transformer yields roughly twice the nameplate %Z, because the nameplate lists a three-phase wye-equivalent through-impedance. Matching nameplate requires a three-phase short on the opposite winding, phase-to-phase test quantities, and averaging the three measurements.
You cannot short one winding of a delta. Any short applied to one delta winding also short-circuits the other two; test current flows in all windings. The same limitation applies to wye windings without an accessible neutral.
Zero-sequence impedance. For grounded-wye windings, short all three phase bushings together, apply single-phase voltage from the shorted bushings to neutral, and compute %Z0 with the appropriate 3× KVA factor. Zero-sequence quantities matter because delta-wye grounded banks are the principal sources of zero-sequence current during ground faults, and correct ground-relay operation depends on them. Zero-sequence impedance is infinite looking into an ungrounded wye or a delta winding.
Verify before tear-down. Always calculate %Z from measured quantities and confirm agreement with the nameplate before dismantling the test setup, so anomalies can be re-checked on the spot.

8. Short-Circuit Impedance Testing

9. DC Winding Resistance Measurement

5. Bushing CT Testing: Ratio, Polarity, and the Flash Test

All large transformers carry temperature indication top-oil, hot-spot, or both — feeding alarms, cooling controls, recorders, and SCADA. Every indicator, recorder, and control must be functionally checked and calibration-verified, classically by immersing all sensor bulbs in a slowly heated oil bath against a reference thermometer while recording all devices simultaneously. Alarm/control contacts are set and verified for pickup on rising temperature and dropout on falling temperature (typical dropout 5–10 °C below pickup). Handle capillary tubes with extreme care — a kink is unrepairable — and know which wells are true oil wells before removing anything from a pressurized tank.

Hot-spot indication is a thermal image

Hot-spot indication is a thermal image: a replica temperature, not a direct winding measurement. A dedicated low-iron bushing CT drives well heaters whose thermal characteristics are factory-designed from heat-run data, so the well tracks the hottest point in the winding. Field verification injects heater current in steps through a resistive load box (the CT wiring need not be lifted its high reflected impedance forces test current through the heaters), monitoring the RTD with a resistance bridge until temperature stabilizes at each step, and comparing against the manufacturer's time-current/temperature curves. The heater tap is factory-set and should not be adjusted the goal is functional verification, not recalibration. Afterward, confirm the CT is actually terminated to the heater circuit: a DC ohmmeter across the injection terminals should read near zero (the CT looks like a short to DC the opposite of its AC behavior). Overheating is the enemy of insulation life: oil temperatures above 100 °C are presumed to accelerate paper deterioration.

Winding resistance verifies that no loose connections, discontinuities, or internal lead problems exist and that shipping has not loosened connections inside the tank. Field data is compared with factory values to validate the as-received condition, used to determine average winding temperature, and retained as a diagnostic benchmark for the transformer's later life.

9.1 Method Selection and Technique

Two families of methods are used: bridge measurement and the volt-amp (V/I) method. The Wheatstone bridge suits windings above 1 Ω (accuracy around 0.05%, but test-lead resistance must be measured immediately after the test and subtracted). The Kelvin bridge and four-terminal volt-amp methods suit low-resistance windings below 1 Ω, because four-terminal measurement excludes lead resistance — connect current leads outside the potential leads, closest to the unknown resistance. Universal rules: connect directly to winding/bushing terminals (never through bus), keep test current below 15% of winding rating to avoid heating, allow 30 seconds to several minutes for current to stabilize before reading, measure winding temperature before and after, bridge every tap position where time allows (at minimum the maximum-winding tap and nominal tap), and operate tap changers several times beforehand so contact wiping establishes low resistance.

9.2 Temperature Correction

Resistance readings are only comparable when normalized. For copper windings, RB = RX × (234.5 + TB) / (234.5 + TX), with standard base temperatures of 20 °C, 75 °C, or 85 °C. For aluminum windings the constant becomes 228.1, and for 97% hard-drawn copper, 241.5. The same relationship, inverted, converts a measured resistance into average winding temperature — the basis of heat-run temperature determination.

The Stored-Energy Hazard

A transformer winding is a very large inductor. Unlike a capacitor, which discharges through a short circuit , an inductor releases its stored energy when the current path is opened — and the voltage generated by a collapsing field can reach kilovolts, arc through open switches, destroy test instruments, and deliver a potentially fatal shock.

Always provide a discharge path (short circuit, discharge resistor, or a temporarily shorted second winding) before de-energizing the DC source. Do not open the short until a series ammeter reads zero — discharge can take 30 seconds to 2 minutes, and prudent practice is to wait at least five minutes before touching connections. Sizing the discharge resistor at R = 100/Ix limits peak discharge voltage to 100 V.

Operate shorting switches with an insulated stick, never watch the discharge arc, never circulate test current through fused switches without jumpering the fuses, and never position yourself in series with a test lead.

10. Tap Changers: NLTC and ULTC

Tap changers are either no-load (NL) — requiring de-energization to move — or under-load (UL), which permit voltage regulation while carrying power. During ratio, impedance, resistance, and TTR testing, every available tap position should be exercised and recorded, especially on new units.

NL tap changers create an open circuit during transitions. While testing, excite the tap winding with a small current and watch the output meter to confirm contacts drop out and pick up at the correct points, that contacts seat firmly, and that spring pressure feels consistent position-to-position. Anything unusual demands investigation. UL tap changers must never open-circuit the winding — monitor excitation current continuously through the full range. Verify local and remote (SCADA) operation across the entire range, confirm the mechanism cannot drive past its stops, check position agreement between local and remote indication, and test the supervisory cutoff switch in both positions.

Three single-phase installations with UL tap changers also carry an out-of-step alarm scheme built on odd/even cam contacts and a time-delay relay. Test it by deliberately placing one phase a single tap off — and understand its blind spot: the scheme only alarms for odd-numbered position differences. Phases sitting at taps 3, 5, and 7 are all two steps apart and generate no alarm, yet the resulting circulating currents cause imbalance, heating, and possible overload. The companion interlock cam set should block remote operation until all phases are back on identical taps.

11. Power Meggering, Polarization Index, and Dielectric Absorption

Winding insulation is evaluated with a 2,500 VDC power megger: winding-to-winding for all combinations, winding-to-ground for all windings, and windings/bushings-to-CTs, with CT secondaries and other low-level circuits tested at 500 V. High-voltage cables tying the transformer to the system are meggered at 2,500 V as well.

A flat one-minute reading tells you very little about a transformer: the winding is a large RC/RL structure, and readings may keep climbing for 30 minutes or more. Ten minutes is the industry-standard interval. A liberal screening rule for the 10-minute reading is 1 megaohm per kV of winding rating a transformer meggering below that line is definitely not safe to energize. Because insulation resistance falls as temperature rises, normalize all readings to 20 °C using published correction tables or a curve-fit correction.

Polarization Index

(PI) is the 10-minute reading divided by the 1-minute reading. Above 2 indicates good insulation; below 1 indicates poor condition; between 1 and 2 is marginal. PI is most meaningful when the 10-minute reading is low a low PI with a high absolute megger value can simply reflect a unit with a naturally high 1-minute reading, and should be ignored.

Dielectric Absorption Ratio

(DAR) is the 60-second reading divided by the 30-second reading; below 1.25 warrants investigation. Plotting the first ten minutes on log-log paper produces a dielectric absorption curve good insulation plots as a rising straight line, while wet or deteriorated insulation flattens and finishes low. These DC tests complement AC power factor testing and become essential when power factor testing is unavailable, on new installations, or when moisture or damage is suspected. Note the physics: DC megger voltage stresses primarily the paper (about 75%), while AC power-factor voltage stresses primarily the oil.

12. Surge Arresters

Transformers that depend on forced air/oil cooling need reliable auxiliary power usually two independent sources through an automatic transfer switch, since many large units are rated for an hour or less of operation without cooling. Commissioning must include phase-rotation checks on every source feeding the transfer switch (fans and pumps provide no cooling running in reverse), deliberate source-kill/restore sequences to prove transfer logic, and verification of transfer-switch alarms. On the cooling system itself, verify all manual and automatic control functions, observe every fan and pump for correct rotation by watching oil and air flow, and prove remote cooling-status indication and loss-of-cooling alarms. Work the bugs out during commissioning once the bank is released to operations, getting coolers out of service for troubleshooting is difficult.

High-voltage power transformers are among the most important — and most expensive — pieces of equipment in any substation. Whether a unit arrives new from the factory or is relocated from another site, it cannot simply be bolted down, filled with oil, and energized. Before the first energization, the commissioning team must verify that the transformer is dry, that no damage occurred during shipping, that internal connections have not loosened in transit, that the measured ratio, polarity, and impedance agree with the nameplate, that the major insulation structure is intact, and that no wiring insulation has been bridged or scraped.

The scope of that verification scales with the machine. Physical size, voltage class, and kVA rating dictate how much preparation, how many auxiliary devices, and how much testing is required before a transformer can be certified ready for energization. A distribution-class unit may need only a handful of checks; a large EHV autotransformer bank demands a coordinated campaign of insulation, ratio, impedance, winding-resistance, CT, cooling-system, and protection testing — and a single responsible test engineer who confirms every item is satisfactorily complete before the final go/no-go decision.

How disciplined pre-energization testing protects the most expensive asset in your substation from nameplate verification and TTR to winding resistance, insulation evaluation, and final trip checks.

1. Why Transformer Commissioning Testing Matters

13. Temperature Devices and the Thermal Image

15. Mechanical Protection: Pressure Relief, Buchholz, and Sudden-Pressure Relays

14. Auxiliary Power, Transfer Switches, and Cooling

Keentel Engineering provides substation engineering, owner's engineer, and commissioning advisory services for utility, IPP, data center, and industrial clients including transformer test plan development, factory and site acceptance test review, energization plans, protection trip-check matrices, and baseline data management for fleet diagnostics. If you are installing, relocating, or repowering large power transformers, our licensed professional engineers can help you get the baseline right the first time.

Case Studies

Every main tank and UL tap-changer tank has its own pressure relief, usually with a lever that can be manually tripped and reset to prove alarm contacts (pressure relief alarms it does not trip breakers). Oil-level indicators on every tank and bushing must be verified before energization; magnetically coupled gauges can be exercised with an external magnet, and some can only be proven by jumpering the alarm wiring.

Buchholz relays

Buchholz relays provide two protective actions: a fast pressure-gate response to internal faults, wired to trip the lockout relay (which trips and blocks all associated breakers until manually reset after inspection), and a slower gas-accumulation function that alarms or trips depending on accumulation rate gas typically originating from corona breaking down oil into combustible gases. A sampling tube and shutoff valve allow gas collection for laboratory analysis. Most models test via levers or buttons; some require forcing dry air or nitrogen through per the manufacturer's book.

Sudden-pressure relays

Sudden-pressure relays serve units without Buchholz protection. Early designs were notorious for false trips from mechanical shock or external close-in faults; modern practice pairs improved relays with time-delay or overcurrent supervision. A field functional test simulates differential pressure by valving off the relay, bleeding a small amount of oil, closing the bleeder, and reopening the main valve the oil inrush operates the relay. Never open the bleeder with the main valve open (a geyser of oil results), and never bleed an in-service relay without first disabling the tripping circuits at the lockout relay.

16. Overall CT Loading, Trip Checks, and Energization

Before energization, CT circuits are loaded overall. On an assembled transformer, primary injection is impractical, so simulated current is injected at each CT secondary termination in the transformer terminal cabinet, with a return path from the control-house relay insertion points so magnitude and polarity can be verified end-to-end typically a two-person job using spare cable conductors.

Trip checks are the final gate. Transformer protection packages can include sudden-pressure, differential, phase/neutral/tertiary overcurrent, delta-winding ground detection, distance elements, and transfer trip, marshalled through lockout relays that multiply contacts, isolate control circuits, and provide interrupting capacity that protective relay contacts lack. Every contact of every protective device must be proven to its tripping function before the equipment enters service — though once breaker trip paths from the lockout are verified, further scheme testing can trip only the LOR to spare unnecessary breaker operations. Even slow-acting protections such as loss-of-cooling deserve rigor: a transformer that never trips for sustained overheating quietly loses service life.

Then, energization: confirm the gas relay is clear, energize while monitoring currents, voltages, and tap-changer behavior, verify phasing against the system before picking up load, and where practical let large transformers soak energized for eight hours before loading. Complete the record test data, oil and moisture results, problems encountered, in-service readings, and energization times because that record is the asset's baseline for the next several decades.

17. How Keentel Engineering Supports Transformer Commissioning

How Keentel Engineering Can Help

Keentel Engineering supports utilities, IPPs, data center developers, and industrial owners with transformer commissioning test plans, acceptance criteria, energization plans, protection trip-check matrices, owner's engineer oversight, and baseline data management. If your project involves installing, relocating, or re-commissioning power transformers, contact our licensed professional engineers before the outage clock starts.

Case Study 1: Impedance Testing Uncovers Transport Damage on a Relocated Transformer

Background

A transmission asset owner relocated a medium-sized, three-phase, delta-wye power transformer in the 12–50 MVA class from a decommissioned position to a new substation bay. The unit had a clean service history, and the construction schedule assumed a straightforward re-commissioning: assembly, oil processing, standard electrical tests, and energization within the outage window.

Challenge

Relocated units carry a specific risk profile: transport shock and handling can shift core and coil assemblies or loosen internal connections in ways that leave no external evidence. Industry practice is explicit that a short-circuit impedance test is required not only for new installations, but for any unit reinstalled at another location — precisely because impedance is one of the few field measurements sensitive to internal geometry.

Approach

The commissioning test plan treated impedance and ratio as fingerprint tests, executed per-phase with single-phase methods so results could be compared between phases:

Voltage ratio was verified at every tap position with precision meters connected directly to the bushing terminals, followed by TTR confirmation at 0.1%-class accuracy, since TTR data forms the diagnostic base for future maintenance.
Short-circuit impedance was measured phase-by-phase with heavy 4/0 shorting jumpers clamped tight to the bushing terminals, a regulated sinusoidal source above 7 V, and 4½-digit RMS instrumentation, with exciting voltage read directly at the transformer terminals to eliminate lead drop.
Percent impedance was calculated from measured quantities and compared to nameplate before the test setup was dismantled, so any anomaly could be re-verified on the spot.

Findings

Two phases measured within the expected band — comfortably inside the 2% phase-to-phase agreement criterion and within 5% of the nameplate value at the nominal tap. The third phase deviated beyond the 5% nameplate threshold and disagreed with its sister phases by well over 2%. Ratio and polarity on that phase were correct, and input current during the ratio test showed no evidence of a shorted turn — pointing away from a winding electrical fault and toward mechanical displacement. A re-measurement from the alternate winding direction and a phase-to-phase cross-check reproduced the deviation, ruling out test-lead and contact-resistance error (a high-resistance short reads as high impedance, so jumper quality was re-verified first). The pattern was consistent with core-and-coil shift sustained in transport.

Outcome

The unit was held out of energization and referred for internal inspection rather than being placed in service on schedule and discovered later through gassing or failure. Because it is unlikely that all phases of a transformer would sustain identical damage, the per-phase, single-phase test strategy was decisive: a three-phase average alone might have blurred the deviation. The owner avoided energizing a mechanically compromised unit, and the documented test data including both measurement directions gave the repair facility a precise starting point.

Key Takeaways

Impedance testing is mandatory for relocated units, not just new ones — transport damage is exactly what it detects.

Compare phases against each other (2% criterion) as well as against nameplate (5% at nominal tap); per-phase comparison is the sensitive indicator.

Calculate %Z from measured values before tearing down the test setup, and re-verify anomalies from the alternate direction before condemning a unit.

Case Study 2: CT Flash Testing Prevents a Differential Misoperation on First Energization

Background

A utility-scale substation project included a new large power transformer protected by a differential scheme fed from bushing CTs. The bushing CTs had passed factory tests, and secondary wiring had been installed by the construction contractor against issued drawings. The project was in the final pre-energization phase: CT circuits were to be loaded overall, flashed for polarity, and the protection scheme trip-checked before the bank was released.

Challenge

Differential protection depends entirely on CT polarity and ratio integrity: a single rolled CT connection converts a healthy transformer's inrush and load current into a false differential signal, tripping the bank on first energization — or worse, desensitizing the scheme to a real internal fault. Because CTs were already mounted in the transformer, primary current-injection ratio testing was no longer practical, and the verification burden shifted to secondary injection, overall loading, and DC flash polarity testing.

Approach

Each CT circuit was loaded overall by injecting simulated current at the secondary termination points in the transformer terminal cabinet, with a return path established from the relay insertion points in the control house over spare cable conductors — a two-person procedure verifying magnitude and polarity end-to-end through every terminal block.
CT polarity was then flash-tested: battery positive to the polarity (high-voltage) terminal through a current-limiting load box, battery negative made and broken with a hot-stick lead, and an analog meter across each CT secondary referenced to X1. Correct polarity shows upscale deflection on make and downscale on break, with the break kick notably larger.
Windings opposite the flashed side were short-circuited to limit arc energy, and all safety rules for DC flash work were enforced — insulated make/break only, no connections made or broken at battery terminals, and no one in series with the test circuit.

Findings

On one bushing CT circuit, the meter deflected downscale on make and upscale on break — the signature of reversed polarity. Tracing the circuit found the secondary leads rolled at the bushing connection box, an error invisible to point-to-point continuity checks because the conductors themselves matched the drawing designations. The overall-loading records for the remaining circuits confirmed correct magnitude and polarity through to the relays.

Outcome

The rolled connection was corrected and re-flashed, and because DC flashing leaves residual magnetism that could contribute to CT saturation and relay misoperation on energization each flashed CT core was demagnetized by applying AC excitation, driving the core just into saturation with the secondary open, and slowly reducing the excitation to zero. Trip checks then proved every protective contact through the lockout relay, and the bank energized cleanly, soaking energized before load pickup while currents, voltages, and relays were monitored in service. A first-energization differential trip with its outage extension, re-testing, and credibility cost was avoided entirely.

Key Takeaways

Flash every bushing CT for polarity before energization; wiring that passes continuity checks can still be rolled.

Overall loading from the CT secondaries to the relay panels verifies the complete measuring chain, not just the CT.

Always demagnetize CT cores after DC flash testing so residual flux cannot bias relay performance on first energization.

Case Study 3: Polarization Index Testing Catches Moisture Before Energization

Background

A new large power transformer arrived at a greenfield substation after extended storage and a long shipping sequence. Assembly proceeded normally: auxiliary wiring was checked and meggered at 500 VDC, bushings were capacitance-checked and power-factored before insertion, the tank was oil-filled, and an oil sample was dispatched for laboratory analysis. The electrical test program then moved to insulation evaluation of the main winding structure.

Challenge

Moisture is the silent schedule-killer in transformer commissioning. It does not announce itself in ratio or impedance data, and a single one-minute megger reading — the minimum many test forms require — provides almost no insight, because a transformer's large insulation structure can continue charging for 30 minutes or more. The team needed a defensible, quantitative basis for the energize/dry-out decision.

Approach

A 2,500 VDC power megger was applied winding-to-winding for all combinations and winding-to-ground for all windings, with readings recorded at 30 seconds, at each minute through ten minutes, and again at 15 and 20 minutes on a dedicated data sheet.
All readings were normalized to 20 °C using standard temperature-correction practice for oil-filled transformers, since insulation resistance falls as temperature rises.
Polarization Index (10-minute / 1-minute) and Dielectric Absorption Ratio (60-second / 30-second) were computed, and the ten-minute data was plotted on log-log paper as a dielectric absorption curve.
Results were screened against the basic energization rule of 1 megaohm per kV of winding rating on the 10-minute reading — a deliberately liberal threshold below which energization is definitely unsafe.

Findings

One winding-to-ground measurement told the story three different ways. The absolute 10-minute reading sat uncomfortably close to the megaohm-per-kV screening line; the PI computed in the marginal-to-poor band (well below 2, trending toward 1); the DAR fell below the 1.25 investigation threshold; and the dielectric absorption curve, instead of rising as a straight line on log-log axes, flattened early and finished low — the classic signature of moisture contamination rather than a localized defect. Because the 10-minute value itself was low, the low PI was meaningful (a low PI paired with a high absolute reading can be ignored, but this was not that case). The oil laboratory results, entered into the bank test record per standard practice, corroborated elevated moisture.

Outcome

Energization was deferred and the unit was processed for drying — with test discipline maintained throughout: no test voltages applied to the windings during vacuum processing, terminals shorted and grounded during hot-oil circulation to control static charge, and no personnel walking the tank under full vacuum. After processing, the insulation program was repeated in full. The 10-minute readings rose decisively clear of the screening rule, PI recovered above 2, the absorption curve straightened, and — 24 hours after oil filling was completed — power-factor tests confirmed the insulation system. The transformer energized on a revised but controlled schedule, and its post-dry-out insulation data set became the healthy baseline for the fleet diagnostics program.

Key Takeaways

One-minute megger readings are nearly meaningless on large transformers — record a timed series to ten minutes or beyond and compute PI and DAR.

Interpret PI in context: it matters most when the 10-minute absolute reading is low; normalize everything to 20 °C before comparing.

A flattened dielectric absorption curve plus low PI and low DAR is a moisture signature — dry the unit, then re-baseline before energization.

Frequently Asked Questions

Work With Keentel

Contact Keentel Engineering to discuss transformer commissioning support, substation design , protection & control, and grid interconnection engineering for your next project.

Fast Real-Time EMT Simulation for Low-Inertia Power Systems

Thu, 02 Jul 2026 11:18:11 GMT

Jul 02, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Fast & Real-Time EMT Simulation for Low-Inertia Power Systems

A useful way to organize the modern EMT toolchain is as a spectrum of study modes, all built on the same parallel real-time simulator technology:

Offline EMT simulation with generic control models the classical mode, appropriate for typical EMT studies and plant-level equipment stress evaluation.

Accelerated / parallel EMT simulation (SIL) with real-code controller emulation used for DER integration studies, interaction studies, and OEM controller model validation. Hundreds of contingencies can be screened in batch on in-house servers or the cloud.

Real-time simulation (controller HIL) with physical control system replicas used for protection and control design and testing and pre-commissioning tests, where the actual hardware must close the loop against the simulated grid.

Quasi-real-time or faster-than-real-time simulation the digital-twin mode for operations: transient security assessment and contingency analysis connected to the system state estimator, re-initialized from a fresh operating state every 5 to 10 minutes.

When the number of IBRs in a study footprint becomes very large, deploying physical control replicas for every power-electronic system becomes impractical or prohibitively expensive. The practical compromise is hybrid: the critical systems — for example, the new HVDC link or IBR plant being interconnected — run in HIL mode with physical replicas, while the remainder of the system is represented with previously validated generic controller models or black-box controller emulators. In other campaigns, parameter optimization and performance testing are executed entirely in SIL with black-box emulators, validated generic models, or a mix of both. In every mode, one operational requirement holds: simulation turnaround must remain fast enough for effective interaction with simulation specialists, who may need to modify test sequences, validate results, or investigate abnormal phenomena while a campaign is running.

How Real-Time EMT Speed Is Achieved

The Computational Wall: Why Large-Scale EMT Is Hard

If EMT is clearly the right tool, why has it not simply replaced RMS for wide-area studies? The answer is computation. Detailed EMT simulation of a large grid that integrates many HVDC systems and IBR plants is extremely intensive. A meaningful stability campaign typically requires simulating large-disturbance scenarios of 20 to 30 seconds each, across hundreds of contingencies, to properly assess grid stability, optimize controller settings, and identify worst-case conditions. Executed serially with conventional offline tools, such a campaign can take hours or days, depending on system size, contingency count, and solver efficiency.

The second obstacle is the models themselves. Plant controllers supplied by original equipment manufacturers (OEMs) are usually delivered as pre-compiled black-box binaries — dynamic-link libraries (DLLs) for Windows or shared objects for Linux built for specific simulation tools, with no universal interoperability standard or unified interface. Interfacing these black-box models with real-time simulators or with other simulation environments is complex and time consuming: absent a standard interface, adapting a single controller code with semi-automatic translation tooling can consume on the order of ten engineer-days. Some black-box implementations are also poorly optimized or embed unnecessarily detailed converter representations that demand very small time steps, dragging the entire simulation below real-time speed. In those cases, collaboration with the OEM to optimize the code is often the only path to real-time execution.

Finally, there is the economics of hardware testing. In real-time HIL testing with physical controller replicas, laboratory time is scarce and expensive. It is therefore essential to identify worst-case scenarios and optimized controller settings before committing them to the HIL bench which is exactly what fast, fully numerical SIL simulation provides.

The Study-Simulation Spectrum: From Offline EMT to Digital Twins

Proof at Scale: A 4,000-Bus, 150-Plant IBR Benchmark

Real-time simulation imposes a hard constraint: every simulation time step must be computed within the corresponding wall-clock interval. That constraint has driven the development of advanced fixed time-step solvers and computational techniques that maintain numerical stability and accuracy under strict timing budgets. Three techniques do most of the heavy lifting:

1. System decoupling and parallelization

Large networks are decoupled into smaller subsystems for parallel computation, exploiting the natural wave propagation delay of long transmission lines as a physically exact decoupling boundary. Decoupling shrinks the individual subsystem admittance matrices, which accelerates the solution dramatically compared to factorizing a single large matrix for the complete system. Mature real-time platforms automatically distribute the model across processors and manage inter-processor communication, so the engineering team is not hand-partitioning the network.

2. Compensation and interpolation for switching events

Power-electronic switching events rarely align with fixed time-step boundaries. Compensation and interpolation methods represent fast switching accurately within a fixed step, preserving accuracy at a relatively larger time step than an uncompensated detailed-switch model would require. For IBR converter representation, a switching-function converter model is the recommended compromise: it delivers close to the fidelity of detailed switch models at close to the real-time performance of average models.

3. Efficient parallel scaling on high-performance hardware

Combined with modern multi-core servers, these techniques scale remarkably well. A benchmark of a real utility interconnection of roughly 1,666 three-phase buses — including 111 electrical machines, 432 lines and cables, 338 three-phase transformers, multiple HVDC converters, static compensators, and wind plants — achieved hard real-time at a 40-microsecond time step on fewer than 60 processor cores. The parallel-scaling numbers are striking: a 15-second event that required 2,565 seconds on one CPU completed in 15 seconds on 56 CPUs, versus a theoretical linear-scaling expectation of 46 seconds. Actual efficiency exceeded 300%, a result of effective processor cache management and the move to more powerful CPUs on a Linux real-time environment.

Just as important for the broader industry: these EMT acceleration techniques, matured over decades in real-time simulators, are now being adopted by mainstream offline EMT tools as well — a rising tide that benefits every stakeholder performing power system analysis.

A typical utility-scale IBR plant model consists of the electrical circuit plus a hierarchy of controllers. Local and point-of-common-coupling measurements are sampled and fed to the controls; a plant-level controller determines power setpoints and provides protection and secondary control; a converter controller implements the primary control functions and DC-circuit protection; and duty cycles or gating pulses are fed back to the electrical circuit, depending on whether an average or detailed converter representation is selected. In practice, the plant and converter controllers arrive from the OEM as pre-compiled black-box code that must communicate with the grid solver at its own native time step, orchestrated by a co-simulation scheme.

Two automated import pathways now remove most of the historical integration pain:

Automatic import of tool-specific DLLs. Where the OEM controller is pre-compiled for a popular offline EMT tool, an automatic import function wraps the DLL in an interface layer so the resulting real-time controller block exposes exactly the same I/O and parameters as the original. An automatic open-loop validation against recorded signals is executed during import, confirming fidelity before the controller ever runs in closed loop.

Standards-based import. Where the OEM code follows emerging industry guidelines for real-code controller interchange, a dedicated interface integrates it directly. Controllers imported this way can execute in near-real-time on Windows and be distributed across parallel processors on the same simulator or a separate one.

Both pathways scale to hundreds of controller instances on a single standard simulator. For real-time HIL campaigns that mix black-box code with physical replicas, tooling is also emerging to execute Windows-compiled DLLs on top of a hard-real-time Linux environment — closing the last gap between OEM deliverables and the HIL bench.

How far do these techniques take you today? A synthetic continental-scale benchmark provides a concrete answer. The network comprises approximately 4,000 three-phase buses; 6,700 lines, loads, and switched shunt reactors; 2,000 transformers and synchronous machines; 150 IBR plants (solar and wind); 300 OEM pre-compiled controller DLLs; 70 FACTS and HVDC converters; and 100 protection relay models.

The same parallel EMT technology is moving to commercial cloud infrastructure, eliminating the need to purchase and maintain in-house server clusters and letting computation scale elastically with study demand across many users and departments. Cloud deployment raises an obvious question: can a grid simulated in a remote data center close the loop with physical control hardware in a local laboratory? For wide-area applications, the answer demonstrated to date is yes.

In a proof-of-concept configuration, a modified 118-bus transmission benchmark augmented with IBR generation is simulated on a cloud server. Four IBR plants are added two Type-3 wind turbine generation systems and two photovoltaic generation systems along with two switched capacitor banks located at buses electrically close to the IBRs to support voltage regulation at each plant's point of connection. Virtual phasor measurement units (PMUs) placed throughout the network stream time-stamped synchrophasor data over the standard synchrophasor communication protocol to a wide-area control (WAC) algorithm executing in real time on a local industrial controller — the same class of computer hardware used in actual control rooms. When the WAC detects an undervoltage event, it sends a command back to the cloud-hosted grid model to switch the appropriate capacitor bank via industrial TCP signaling.

Two disturbance scenarios illustrate closed-loop behavior. In the first, a three-phase-to-ground fault occurs with protection failing to pick up: voltage at a wind plant's point of connection collapses to roughly 0.62 per unit, the WAC issues its switch-on command 0.583 seconds after fault inception, the capacitor bank compensates the voltage to about 0.75 per unit, and after fault removal the voltage recovers to nominal and the bank is disconnected. In the second, an identical fault is cleared within 0.3 seconds by line disconnection; voltage never sags significantly, and the WAC correctly refrains from switching. Discrimination between the two cases acting decisively when needed, staying quiet when not is precisely what wide-area special protection scheme validation must demonstrate.

Computational performance on the cloud held comfortably within the real-time budget: across the seven cores used, average per-step execution times ranged from 14.7 to 39.2 microseconds against a 50-microsecond time step. Because the synchrophasor protocol carries time-stamped data referenced to the simulator clock, results remain valid for controller HIL even though generic Ethernet between cloud and laboratory is slower than the dedicated channels of production wide-area networks. The same architecture extends naturally to very large grids with hundreds of IBRs and distributed wide-area control and the WAC software itself can be hosted in the cloud for a fully digital SIL campaign before any hardware enters the loop.

What This Means for Asset Owners and Operators

Contingency screening that took days now fits in an hour worst cases are identified before expensive HIL bench time is committed.

Online EMT transient stability assessment on a 15-minute cadence is now technically feasible, opening the door to true operational digital twins initialized from the state estimator.

Wide-area monitoring, protection, and control schemes can be validated end-to-end first in fully digital SIL, then in HIL against control-room-grade hardware without owning a supercomputer.

Running on a 500-core Windows server cluster one high-performance 128-core machine plus 22 high-performance 4-GHz 18-core machines a 30-second EMT event completes in 90 seconds of wall-clock time. Roughly 100 cores carry the 4,000-bus electrical network at a 50-microsecond time step, while about 300 cores execute the manufacturer controller codes at their native 10-microsecond or 16.67-microsecond steps. Three-times-slower-than-real-time may sound modest, but its practical impact is enormous: a large contingency campaign that formerly consumed days can now be completed in about an hour, and a 90-second turnaround on a 30-second phenomenon is fast enough to support online EMT TSA delivering results on a 15-minute operational cadence.

Notably, real-time speed was within reach for the network solution itself; the limiting factor was a handful of black-box controllers whose implementations consume disproportionate compute. That finding carries a procurement lesson: OEM cooperation on controller code optimization should be negotiated up front, because it is frequently the binding constraint on real-time execution. Near-real-time results also accelerate a complementary workflow developing and validating reduced-order equivalents so that critical cases can be re-run in hard real-time with replicas and DLLs on smaller simulator facilities.

Two complementary acceleration paths have now been demonstrated at convincing scale. Proprietary clusters of high-performance standard computers can simulate a 4,000-bus continental-scale system with 150 IBR plants and 300 OEM controller binaries at one-third of real-time speed on 500 processors, and can run a real 1,666-bus utility grid HVDC and static compensation included in hard real-time on fewer than 60 processors. Commercial cloud servers, meanwhile, have proven capable of hosting wide-area control HIL tests against actual controller hardware, with elastic scaling for many simultaneous users and studies.

The benefits compound across the project lifecycle: rapid identification of critical scenarios for focused HIL analysis; faster EMT planning and IBR connection studies with more contingencies assessed in less time; operator-facing real-time EMT TSA tools and digital twins refreshed on operational timescales; and shorter turnaround when specialists must chase root causes of instability observed in the field or in simulation. As these acceleration techniques migrate into mainstream offline tools as well, the entire industry's analytical ceiling rises. For grids shedding inertia as quickly as they are adding inverters, fast and real-time EMT simulation is no longer a research luxury it is the foundation of reliable integration.

Conclusions

How Keentel Engineering Can Help

Interfacing OEM Black-Box Controllers at Scale

Keentel Engineering provides EMT and PSS/E modeling, point-of-interconnection engineering, IBR and BESS integration studies, protection and control design support, and NERC compliance services for utilities, developers, and large-load customers. If your project requires EMT model development, black-box controller integration, SIL/HIL test planning, or interconnection study support for low-inertia grid conditions, our licensed professional engineers can help you scope the right study campaign the first time. Contact Keentel Engineering to discuss your project.

For decades, transient stability assessment of large interconnections was performed almost exclusively with positive-sequence, phasor-domain (RMS) dynamic simulation. That approach was well matched to a grid dominated by synchronous machines, whose electromechanical dynamics evolve over hundreds of milliseconds to seconds. Power-electronic converters changed the equation. Many converter control and protection functions have time constants and reaction times far faster than a few milliseconds — entirely outside the bandwidth that positive-sequence RMS models can represent.

Three-phase RMS dynamic simulation overcomes some limitations of positive-sequence tools by capturing unbalanced conditions, but it still cannot represent the harmonic content and fast transients to which converter controls and protections actually respond. Industry experience has demonstrated cases in which an RMS model predicted successful ride-through of an HVDC link, while the corresponding EMT model correctly predicted sustained commutation failure and subsequent disconnection of the link. The two tools gave opposite answers to a question with major reliability consequences — and only the EMT result reflected reality.

The direction of the industry is unambiguous. Reliability organizations now recommend that EMT models be required for all newly connecting inverter-based resources, and a growing number of utilities have adopted EMT simulation for wide-area stability analysis, control design, and testing. Interconnection and interoperability standards for transmission-connected IBRs reinforce the same expectation: dynamic performance must be demonstrated with modeling detail that matches the physics of power-electronic equipment.

Keentel Engineering Insight

If your interconnection portfolio includes IBRs, HVDC, or FACTS, plan for EMT studies early. Retrofitting EMT analysis after an RMS-only study campaign frequently uncovers control interactions, ride-through failures, and protection misoperations that force late-stage redesign — the most expensive point in a project to discover them.

The modern bulk power system is undergoing the most fundamental change in its dynamic character since interconnection began. Inverter-based resources (IBRs) such as utility-scale wind and solar, flexible AC transmission system (FACTS) devices, and high-voltage direct current (HVDC) links are being added at record pace, while synchronous thermal and nuclear plants retire. The net effect is a steady, structural decline in system inertia. The fast power-electronic controllers that replace synchronous machines are expected to stabilize the grid, but they are also highly sensitive to fast transients, harmonics, and network imbalances phenomena that conventional planning tools were never designed to capture.

In this article, Keentel Engineering examines why detailed electromagnetic transient (EMT) simulation has become indispensable for large-scale, low-inertia grids; why simplified positive-sequence RMS studies fall short; how parallel computing on high-performance clusters and cloud infrastructure now makes real-time and even faster-than-real-time EMT simulation of multi-thousand-bus systems achievable; and how these capabilities enable hardware-in-the-loop (HIL) controller testing, software-in-the-loop (SIL) screening, wide-area control validation, and online transient stability assessment (TSA).

Why RMS Simulation Alone Is No Longer Enough

Case Studies

Case Study 1: Hard Real-Time EMT Simulation of a Complete Utility Transmission System with Embedded HVDC

Background

A large utility operates a transmission system of approximately 1,666 three-phase buses whose dynamic behavior is shaped by a demanding mix of conventional and power-electronic assets: 111 electrical machines with 86 governors, 81 excitation systems, and 54 stabilizers; 432 lines and cables; 338 three-phase transformers; 165 dynamic loads; six HVDC converters; ten static compensators; and six wind power plants. For a system of this composition, controller and protection interactions between HVDC links, static compensation, and wind generation cannot be adequately evaluated with phasor-domain tools the utility required a full EMT representation of the entire interconnected system, executable in hard real-time so that physical control and protection equipment could be tested in closed loop.

Challenge

EMT simulation of a complete utility grid not a reduced equivalent at a time step small enough to represent HVDC and FACTS control dynamics is a formidable computational problem. Executed on a single processor, a 15-second disturbance event required 2,565 seconds of computation: 171 times slower than real time, and far too slow for interactive study work, let alone hardware-in-the-loop (HIL) testing , which demands that every simulation time step complete within its wall-clock interval without exception. The engineering question was whether parallel decomposition could close a two-orders-of-magnitude speed gap while preserving numerical stability and waveform-level accuracy.

Approach

Physics-based network decoupling. The system was partitioned into subsystems along long transmission lines, whose natural wave propagation delay provides an exact decoupling boundary at the EMT time step. Decoupling shrinks each subsystem's admittance matrix, so many small matrices are solved in parallel instead of one very large matrix serially.
Automatic parallel task mapping. The simulation platform automatically distributed the model across processor cores and managed all inter-processor communication, eliminating manual partitioning and ensuring balanced core utilization.
Fixed-step solvers with switching compensation. Advanced fixed time-step solvers with compensation and interpolation techniques represented fast power-electronic switching accurately within the fixed step, allowing a practical time step without sacrificing converter fidelity.
Cache-efficient execution on modern hardware. The production benchmark ran on eight 8-core high-frequency server-class processor modules (3.5 GHz, large L3 cache) within a single high-performance shared-memor y server, on a Linux real-time operating environment.

Results

Case Study 2: Near-Real-Time EMT Screening of a 4,000-Bus Grid with 150 IBR Plants and 300 OEM Black-Box Controllers

Cloud-Based EMT and Wide-Area Control HIL

Case Study 3: Cloud-Hosted EMT Simulation with Hardware-in-the-Loop Wide-Area Control Validation

Background

A proof-of-concept program set out to answer a question with major cost implications for utilities and study consultants alike: can large-scale EMT simulation move to commercial cloud infrastructure — eliminating in-house simulator server ownership and scaling elastically across users and departments — while still supporting hardware-in-the-loop testing of wide-area monitoring, protection, and control (WAMPAC) equipment located in a physical laboratory? The test article was a modified 118-bus transmission benchmark (186 branches, 99 loads, 18 synchronous machines) augmented with four IBR plants: two Type-3 wind turbine generation systems and two photovoltaic generation systems. Two switched capacitor banks were added at buses electrically close to the IBR plants to support voltage regulation at each plant's point of connection.

Challenge

Cloud servers are not hard-real-time machines: the environment used was Windows-based and not optimized for deterministic microsecond scheduling, and the Ethernet path between a commercial cloud data center and a laboratory is slower and less deterministic than the dedicated communication channels of production wide-area networks. The scheme under test — a wide-area control (WAC) algorithm performing special-protection-scheme-class capacitor switching — had to demonstrate both dependability (acting correctly on genuine undervoltage events) and security (refraining from action on self-clearing disturbances) across that non-ideal communication fabric, with the physical controller executing on the same class of industrial computer hardware used in actual control rooms.

Approach

Grid model on the cloud. The complete EMT model, including the four IBR plants and both capacitor banks, was executed on a cloud-hosted parallel simulation service at a 50-microsecond time step across seven processor cores.
Virtual PMUs with time-stamped synchrophasor streaming. Virtual phasor measurement units placed throughout the network reported synchrophasor data to the laboratory over the standard synchrophasor protocol. Because the protocol carries data time-stamped against the simulator's 50-µs clock, measurement validity is preserved despite variable network latency the architectural insight that makes cloud HIL viable for wide-area applications.
Physical WAC controller in the loop. The wide-area control algorithm executed in real time on a local industrial controller computer. Upon detecting an undervoltage condition, it issued switching commands back to the cloud-hosted grid model via an industrial TCP protocol to energize the appropriate capacitor bank.
Dependability/security scenario pair. Scenario 1: a three-phase-to-ground fault with primary protection failing to pick up a sustained depressed-voltage condition requiring WAC action. Scenario 2: an identical fault at the end of a transmission line, cleared in 0.3 seconds by line disconnection a self-clearing event requiring no action.

Results

Background

A continental-scale synthetic transmission network of approximately 4,000 three-phase buses was adopted as the benchmark for a study campaign representative of a real interconnection with very high inverter-based resource (IBR) penetration. The model comprises roughly 6,700 lines, loads, and switched shunt reactors; 2,000 transformers and synchronous machines; 150 utility-scale IBR plants (solar and wind); 70 FACTS and HVDC converters; 100 protection relay models; and — critically — 300 original equipment manufacturer (OEM) controller models delivered as pre-compiled black-box dynamic-link libraries (DLLs). The objective was transient stability assessment and control performance analysis across large contingency sets, with sufficient speed to support both study-campaign screening and, ultimately, online assessment on operational timescales.

Challenge

Three obstacles compounded. First, raw scale: waveform-level EMT solution of a 4,000-bus network at a 50-microsecond time step, across disturbance events 20 to 30 seconds long and contingency sets numbering in the hundreds. Second, black-box integration: 300 vendor controller binaries, compiled for a specific offline simulation tool, with no unified interoperability standard — historical experience shows roughly ten engineer-days per controller for manual adaptation, an untenable 3,000 engineer-day exposure if attacked by hand. Third, heterogeneous timing: the vendor controller codes execute at native time steps of 10 or 16.67 microseconds, several times faster than the grid solution, requiring a co-simulation scheme to orchestrate data exchange without destabilizing either side.

Approach

Automated DLL import with open-loop validation. An automatic import function wrapped each vendor DLL in an interface layer exposing identical I/O and parameters to the original tool, and executed an automatic open-loop validation against recorded signals during import — confirming each controller's fidelity before closed-loop use. Controllers following industry real-code interchange guidelines were integrated through a dedicated standards-based interface.
Co-simulation across a compute cluster. The electrical circuits of the IBR plants and the bulk network ran on dedicated cores of the real-time simulator cluster while the OEM controller codes executed in parallel on companion high-performance machines, linked by fast communication channels. Automatic task mapping assigned processes for maximum throughput.
Switching-function converter models. IBR converters were represented with switching-function models — the recommended compromise delivering near-detailed-switch accuracy at near-average-model speed — keeping the grid time step at 50 µs.
Scalable hardware architecture. The campaign ran on a 500-core cluster: one high-performance 128-core server plus 2 2 high-performance 4-GHz 18-core machines. The architecture scales by simply adding processors as IBR plant count grows.

Results

Frequently Asked Questions: Fast and Real-Time EMT Simulation

Average per-step execution time on every cloud core remained below the 50-microsecond simulation time step, confirming comfortable computational margin. The closed-loop scenario pair demonstrated both required properties of a special protection scheme: decisive, correctly targeted action on the sustained undervoltage event, and correct restraint on the self-clearing fault. The results validate cloud-hosted EMT HIL for wide-area schemes communicating over TCP/IP-based protocols synchrophasor, Modbus, DNP3 and similar provided communication latency remains within pre-defined boundaries. The same architecture extends to very large grids with hundreds of IBRs and distributed wide-area control, and the WAC software itself can be cloud-hosted for a fully digital SIL campaign before hardware enters the loop. A Linux-based cloud environment is under implementation to further raise computational performance beyond the Windows-based proof of concept.

Key Takeaways

Time-stamped synchrophasor data is the enabling insight: simulator-clock time-stamping preserves measurement validity across non-deterministic cloud-to-lab Ethernet.

Cloud HIL is production-relevant today for WAMPAC-class schemes on TCP/IP protocols; microsecond-loop local controllers still require on-premises real-time simulators.

Both dependability and security of a wide-area special protection scheme were demonstrated in closed loop against control-room-grade hardware.

Elastic cloud scaling lets multiple specialists and departments run simultaneous studies without capital investment in simulator clusters.

Substation Protection Guide | Protective Relaying Fundamentals

Tue, 30 Jun 2026 15:02:52 GMT

Jun 30, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Fundamentals of Substation Protection

The lesson is immediate: Ground faults are by far the most common, which is why dedicated ground protection (the "N" and "G" functions you'll meet below) earns its place on nearly every feeder and transformer.

Faults are also unevenly distributed across equipment. Roughly half occur on overhead lines, with the remainder spread across switchgear, transformers, cables, instrument transformers, control equipment, and miscellaneous causes. Overhead lines are exposed to the weather and the public, so they fail most — and they are also the hardest to protect because they are long and electrically "fuzzy" at their far ends.

The following case studies are illustrative and generalized. No project names or locations are identified; each is a composite scenario drawn from common substation-protection situations, presented to show how the fundamentals are applied in practice.

The seven design tensions

The four building blocks

Transducers — The instrument transformers, namely current transformers (CTs) and voltage/potential transformers (PTs). The power system runs at thousands of amps and tens of thousands of volts; relays cannot touch those quantities directly. CTs and PTs scale them down to standard, safe levels — typically 5 A and 115 V — that relays are built to measure. They are the senses of the system.

Relays — The brain. A relay continuously watches the scaled-down quantities and decides whether conditions are normal or abnormal. If everything is normal, it does nothing. If a quantity crosses into abnormal territory, it issues a trip signal. That decision — normal or abnormal — is the entire job, and getting it right under every credible condition is the entire challenge.

Circuit breakers —The muscle. When the relay says "trip," the breaker physically interrupts the current and isolates the faulted section. The relay decides; the breaker acts.

Tripping and auxiliary supplies —usually a station battery. This is the detail beginners overlook: protection must work during a fault, which is precisely when the AC system may have collapsed. A reliable DC battery supply energizes the relay logic and the breaker trip coil independently of the faulted AC system, so the scheme can clear a fault even when the lights are out.

What protection is actually fighting

The abnormalities a substation faces fall into three broad families:

Overcurrent — overload, short circuit, or open circuit.
Ground potential — ungrounded equipment, dangerous touch and step potentials.
Surge voltages — lightning strokes, switching surges, and harmonics.

Of these, short-circuit faults dominate the protection engineer's attention, and not all faults are equally common. The well-known field statistics are worth memorizing because they shape design priorities:

Zones, overlap, and the no-blind-spot rule

There is no such thing as perfect protection, only a defensible balance among competing demands. A good scheme is judged against these criteria:

Reliability is the master quality, and it splits into two halves that pull against each other:

Dependability — certainty that the relay will operate when it should. A dependable scheme never misses a real fault.
Security — assurance that the relay will not operate when it shouldn't. A secure scheme never trips on load, transients, or external faults.

You can always make a scheme more dependable by making it more eager to trip, but every step in that direction costs security, and vice versa. Much of the craft of protection is finding the right point on that seesaw for a given application.

Sensitivity — the relay must respond to genuine abnormal conditions, even weak ones (like a high-resistance ground fault), while ignoring normal operating conditions.

Selectivity — when a fault occurs, only the breakers needed to isolate it should open, and no more. This is achieved through the concept of zones of protection, discussed below. Good selectivity means a fault on one feeder does not darken the whole substation.

Speed — faults must be cleared fast, because damage and instability both grow with time. Speed is classified as:

Instantaneous — no intentional delay.
High speed — operation in less than about three cycles.
Time-delay — an intentional delay, deliberately introduced for coordination.

Economics — maximum protection at minimum cost. The price of installing, operating, and maintaining the scheme is always weighed against the potential losses from equipment damage and outages.

Experience — history is a design input. Knowing the kinds of trouble a given system tends to suffer, how relays have actually performed, and how the system is operated and maintained all feed back into better schemes.

Industry standards — bodies such as IEEE and IEC, through ANSI and IEC standards, codify decades of collective experience so that every engineer is not forced to reinvent the basics.

The hardware that makes the trip decision has gone through three generations, and understanding the progression explains why modern substations look the way they do.

Electromechanical relays (1st generation) were the workhorses for most of the twentieth century. They use the induction-disc principle — the same physics as an old watthour meter. Current through a coil produces torque on a metal disc; the disc rotates against a spring until its contacts close and send a trip. They are robust and intuitive, but they have moving parts that wear, springs that fatigue with temperature, and they need periodic recalibration. They also impose a high burden on CTs and lose sensitivity at higher currents. Their saving grace, the draw-out construction, lets a technician pull the relay for testing without rewiring.

Static / solid-state relays (2nd generation) replaced the spinning disc with electronics. Timing characteristics come from RC circuits rather than a rotating mass, so there are no moving parts, reset is fast, and maintenance drops. They were widely used to retrofit electromechanical units. Their weaknesses live in the analog domain: RC timing drifts with temperature and has poor repeatability, and the AC-to-DC conversion at the front end is vulnerable to offset, harmonics, and noise. RC and LC filters could tame some of that, but at the cost of slower tripping.

Digital and numerical relays (3rd generation) are computers. A microprocessor or microcontroller samples the waveform, converts it to numbers, and applies digital signal processing to extract exactly the information it needs. The consequences are transformative:

Selectable characteristics and functions — one relay can be configured for many protection roles.
Metering and control built in.
Event and disturbance recording, enabling true post-mortem analysis of a fault.
Remote communication to laptops and control centers.
Self-monitoring — the relay watches its own health.
"All in one" — a single numerical relay can replace a panel full of discrete relays.

Numerical relays bring inherent immunity to DC offset and harmonics through tuned digital filtering, highly stable and repeatable performance, very few physical components, and the ability to act as a full bay-level controller rather than a mere protection element. The one nuance worth remembering from signal processing: a pure time-domain sample stream carries amplitude but, by itself, no phase or frequency information — which is exactly why the digital filtering and measurement algorithms inside the relay matter so much.

The single most important organizing idea in protection is the zone of protection a defined region of the power system that a given set of relays is responsible for. Two rules govern zones:

Every element of the power system must lie inside at least one zone. Nothing may be left unprotected.
Adjacent zones must overlap. If they merely touched, a fault exactly at the boundary could fall into a gap. Overlapping the zones typically arranged around the circuit breakers and their CTsv guarantees there are no "blind spots."

A typical substation is carved into zones such as bus protection, transformer protection, line (subtransmission) protection, feeder protection, and generator protection. The overlap is deliberately built at the breakers, so that whichever zone sees the fault, the boundary itself is always covered twice.

No component is perfect, so protection is layered.

Primary protection is the main scheme for a zone — the first to act, ideally fast and selective.

Backup protection stands ready in case the primary fails to clear the fault (a stuck breaker, a failed relay, a lost DC supply). Backup comes in two forms:

Local backup — alternate protection at the same substation, which trips additional local breakers if the primary device fails.
Remote backup — protection at neighboring substations that reaches into this zone and clears the fault from the far end if the local scheme fails entirely.

Remote backup is slower and trips more of the system, but it works even if the entire local substation's protection or DC supply is lost — a genuinely independent layer.

Primary and backup: assume the first line will sometimes fail

For protection to be selective, relays must distinguish a fault they should clear from one they should leave to someone else. They do it using one or more of these handles:

Current magnitude — a fault closer to the source draws more current; grading pickup levels lets relays sort near from far.
Time — deliberately delaying downstream-versus-upstream relays so the device closest to the fault acts first.
Current direction — knowing which way fault current is flowing, so a relay only responds to faults in front of it.
Distance (impedance) measurement — computing Z = V/I, which is proportional to the distance to the fault.
Current balance — comparing current in versus current out of a protected zone (the basis of differential protection).
Phase comparison — comparing the phase of currents at the two ends of a line.

Real schemes combine these. Time plus current magnitude gives the classic inverse-time overcurrent grading. Time plus distance gives the stepped, multi-zone reach of distance relays. Current balance gives differential protection its near-perfect selectivity.

How relays tell faults apart: methods of discrimination

The standard relay types you will meet

How relays evolved: from spinning discs to software

The situation

A distribution substation commissioned a power transformer rated well above the 5 MVA threshold, protected by a percentage differential relay (87T) with backup overcurrent (51) and backup ground (151G). During commissioning, every attempt to energize the transformer from the high side caused the differential relay to operate and lock out through the 86T lockout, even though no fault existed.

The investigation

The protection engineers walked the problem back through the fundamentals. A differential relay trips on the difference between current in and current out. With the transformer unloaded and healthy, that difference should be essentially zero yet the relay was seeing a large transient differential at the exact moment of energization. The signature pointed to magnetizing inrush:

when a transformer is first energized, it draws a brief, large, distorted current to magnetize its core, and that current enters the transformer without a corresponding current leaving the other side. To a naive differential element, inrush looks exactly like an internal fault.

The resolution

The fix followed directly from transformer-protection principles. The relay's settings were reviewed against two boundaries it must respect: the operating characteristic has to sit below the transformer damage curve but above the magnetizing inrush. The percentage-restraint characteristic and the relay's inrush-restraint (harmonic-based) feature were configured so the element would ride through the inrush transient while remaining fully sensitive to genuine internal faults. After the restraint was correctly set, energization proceeded cleanly, and a subsequent staged internal-fault test confirmed the relay still operated correctly for real faults.

The lesson

Differential protection is only as good as its ability to distinguish the kinds of unbalance that are normal (inrush, through-fault CT mismatch) from the kind that is not (an internal fault). The dependability security balance is not abstract here: a relay set purely for maximum dependability would trip on every energization, while one set for the restraint the transformer actually needs is both secure during inrush and dependable during fault.

Case Study 2 — The far-end fault that cleared slowly

The situation

A subtransmission line fed from both ends was protected by stepped distance relays (21) at each terminal, with directional overcurrent backup (67/67N) and auto-reclosing (79). Operators noticed a recurring pattern: for faults near one end of the line, the near breaker tripped instantly, but the far breaker took noticeably longer long enough that the disturbance was more severe than it should have been, and reclosing behavior was inconsistent.

The investigation

This is the classic limitation of a purely stepped distance scheme, and the engineers recognized it immediately. Zone 1 at each terminal is deliberately set to under-reach about 80–90% of the line so that it can trip with no intentional delay while being physically incapable of overreaching into the adjacent line. The consequence is a band near each end of the line (the last 10–20%) that the remote terminal sees in instantaneous Zone 1 but the local terminal sees only in time-delayed Zone 2, roughly 0.35 s later. So a fault in that band clears fast from one end and slow from the other precisely the symptom observed.

The resolution

The remedy was communication-aided protection. A permissive overreaching transfer trip (POTT) scheme was implemented over the existing teleprotection channel (85). With it, each terminal's overreaching zone, upon detecting a fault, sends a permissive signal to the other end; when both ends confirm the fault lies on the protected line, both terminals trip instantaneously, eliminating the delayed-end problem. The single-shot auto-recloser (about 300 ms) was retained for the now-symmetrical fast trips, with the standing rule preserved that backup-initiated trips block reclosing.

The lesson

Stepped distance protection is selective and fast for most of a line but inherently leaves an end-zone timing gap. Recognizing that the slow far-end clearance was a designed-in characteristic, not a relay defect, pointed straight to the correct fix. Communication does not replace the distance relays; it coordinates the two ends so the whole line is covered at high speed.

Case Study 3 — Choosing a busbar protection scheme for a security-critical bus

The situation

A substation upgrade required dedicated busbar protection for an important bus operating above the lower-voltage tier. The original design proposal reused an inexpensive low-impedance differential scheme based on time-overcurrent relays the same approach used elsewhere in the facility on lower-voltage buses. The protection review flagged a concern about its suitability for this particular bus.

The investigation

The reviewers returned to first principles of bus differential protection. A bus differential sums the currents of every circuit on the bus and operates on any imbalance; its great vulnerability is CT saturation during a heavy external (through) fault, when one CT may saturate and produce a false differential current that the scheme misreads as an internal bus fault. Low-impedance, overcurrent-based schemes are economical but are precisely the type most exposed to that failure mode acceptable at the lower-voltage tier (around 34.5 kV and below) where the review confirmed it was appropriately used, but a security risk on a higher-voltage, security-critical bus. The team also verified the wiring fundamental: every CT feeding the differential had to share the same ratio, or the scheme would see a standing false differential even under healthy load.

The resolution

For the critical bus, the design was changed to a high-impedance differential scheme using an overvoltage relay. By deliberately loading the CTs with a high impedance, the scheme forces differential current to manifest as a voltage and is far more tolerant of CT saturation, delivering the higher security the bus warranted consistent with standard practice of applying high-impedance schemes at 115 kV and above and on any lower-voltage bus that demands high security. CT ratios were standardized across all bus circuits, and the bus lockout (86B) was confirmed to trip and lock out every breaker on the bus. The existing low-impedance scheme was retained where it was genuinely appropriate, on the lower-voltage buses.

The lesson

Scheme selection is an exercise in matching the protection to the consequence of failure. The cheaper low-impedance scheme was not "wrong" it was wrong *for this bus*. Letting the security requirement and the CT-saturation risk drive the choice, rather than defaulting to the familiar low-cost option, is the economics-versus-security balance applied with judgment.

Protection engineering speaks in ANSI device numbers — a shorthand where each function has a number. A few you will see constantly: 50 instantaneous overcurrent, 51 AC time overcurrent, 52 the circuit breaker itself, 87 differential, 21 distance, 67 directional overcurrent, 79 auto-reclosing, 86 lockout, and the N or G suffix for the ground version of a function. Here are the workhorses behind the numbers.

Overcurrent relays (50/51) operate when current exceeds a preset value. They are simple and non-directional. On a radial feeder, you typically need three phase relays plus one ground relay — the ground relay (51N) measures the residual current, the sum of the three phase currents, which is zero in healthy balanced operation and nonzero the instant a ground fault appears.

Directional overcurrent relays (67) add a condition: they only operate for fault current flowing in a defined direction. This makes overcurrent protection usable on looped systems where current can flow either way, and it improves both selectivity and security. Their limit is that as source strengths change, a directional overcurrent relay can be fooled into seeing a fault it shouldn't — which is one reason distance relays are preferred on important lines.

Distance relays (21) measure the impedance of the line. Because line impedance is proportional to length, the relay effectively measures how far away the fault is, independent of source strength. They are the standard for double-ended transmission lines from roughly 69 kV up to 500 kV. Their characteristic shapes — impedance, reactance, mho, quadrilateral, composite — are plotted on the R-X plane, and the mho characteristic dominates in practice because it is inherently directional.

Differential relays (87) apply the simplest and most powerful idea in protection: what goes in must come out . They compare the current entering a protected element with the current leaving it. In healthy operation, or for a fault outside the protected zone, the two are equal and the relay sees nothing. For a fault inside the zone, the balance breaks and the relay trips. Differential protection is beautifully selective because it responds only to faults within its own boundaries — making it the protection of choice for transformers (87T), buses (87B), and generators.

Putting it together: standard substation schemes

The theory becomes concrete in the handful of schemes that protect almost everything in a distribution or subtransmission substation.

Radial line / feeder protection. A radial feeder is fed from one end, so simple overcurrent protection suffices. The standard package is instantaneous and time overcurrent for both phase and ground (50/51 and 50N/51N), a power circuit breaker (52), and an auto-recloser (79) with its recloser switch (43R). Reclosing matters because most overhead-line faults are transient — a branch brushing a conductor, a flashover that clears itself. The recloser trips, waits, and recloses; if the fault was temporary, service is restored automatically, and only a permanent fault results in a lockout.

Looped line protection. Lines that are fed from both ends cannot rely on plain overcurrent, because fault current can arrive from either direction. The standard package centers on a distance relay (21 for phase, 21G for ground), backed up by directional overcurrent (67/67N), with teleprotection (85) and auto-reclosing (79). The distance relay is applied as a stepped, time-graded scheme:

Zone 1 — Set to about 80–90% of the line impedance, no intentional delay. It deliberately under-reaches the far end so it can never mistakenly trip for a fault on the next line. Speed with guaranteed selectivity.

Zone 2 — Set to 100% of the line plus 20–50% of the shortest adjacent line, with a delay of around 0.35 s. It over-reaches to cover the far end of the line that Zone 1 left out, but its delay lets the adjacent Zone 1 act first for faults beyond the bus.

Zone 3 — Reaches still further (100% of the line + 100% of the longest adjacent line + a margin), with about a 1.0 s delay, serving as remote backup. Crucially, its reach must stay above the heaviest load impedance, or it will trip on load.

That stepped scheme leaves one awkward gap: a fault in the last 10–20% of the line is in Zone 1 from the remote end (instantaneous) but only Zone 2 from the local end (delayed) — so one end clears fast and the other lingers. The fix is communication-aided protection (for example, a permissive overreaching transfer trip, POTT): the two ends exchange a signal so that when both ends agree the fault is on the protected line, both trip instantaneously. Add single-shot auto-reclosing (around 300 ms, to give the breaker time to extinguish the arc) and the line is both fast and selective end to end. Note one safety rule: when a trip comes from backup protection rather than the main distance scheme, auto-reclosing is blocked — you don't reclose onto a fault your primary protection couldn't even see properly.

Transformer protection. Transformers are expensive, slow to replace, and vulnerable to both electrical and thermal stress. The threats and their defenses:

Internal faults → differential protection (87T), often with a percentage-restraint characteristic so it stays stable during through-faults and inrush while still catching genuine internal faults.

Overload / external fault backup → time overcurrent (51) and backup ground (51G/151G). Overcurrent relaying is standard for transformers of roughly 5 MVA and above; smaller distribution transformers often just use fuses. The overcurrent characteristic must sit below the transformer's damage curve but above its magnetizing inrush, so it protects the unit without nuisance-tripping every time it energizes.

Overheating → thermal relay. The rule of thumb is sobering: normal maximum working temperature is about 95 °C, and every 8–10 °C of sustained rise above rating roughly halves the transformer's insulation life.

Internal gas / incipient faults → Buchholz relay, which detects gas generated by an internal arc.

Sudden internal pressure → sudden-pressure relay / pressure relief.

A lockout relay (86T) ties it together: when 87T operates, the lockout trips and locks out both the high-side and low-side breakers, and they cannot be reclosed until someone investigates and manually resets.

Busbar protection. A bus fault is potentially the worst event in a substation because everything connects to the bus. Schemes follow the bus arrangement:

Single bus, double bus / double breaker, breaker-and-a-half, main-and-transfer with single breaker, and ring bus each have their own differential connection. When the bus differential (87B) operates, a lockout (86B) trips and locks out all breakers connected to that bus.
A vital wiring rule: all CTs feeding a bus differential must have the same ratio, or the scheme will see a false differential current under normal load.
The ring bus is the elegant exception — it needs no dedicated bus differential at all, because each bus section is already covered by the line or transformer protection on either side of it.

Two technologies implement bus differential:

Low-impedance schemes use time-overcurrent relays — inexpensive, but vulnerable to CT saturation, so they suit lower-voltage applications (around 34.5 kV and below).

High-impedance schemes use overvoltage relays and deliberately load the CTs with a high impedance, forcing the differential current to reveal itself as a voltage. They cost more but deliver much higher security, which is why they dominate at 115 kV and above, and on any lower-voltage bus where security is paramount.

The throughline

Strip away the device numbers and the scheme diagrams, and substation protection is one repeated idea applied at every level: define a zone, give it a fast and selective primary scheme, give it an independent backup, make adjacent zones overlap, and balance dependability against security for the realities of that particular system. The hardware has marched from spinning discs to self-monitoring computers, but the philosophy — clear the fault, protect the people, save the equipment, keep the lights on everywhere the fault isn't — is exactly what it was at the start.

Part 2 — Case Studies

Frequently Asked Questions

Every time you flip a switch and a light comes on, you're at the receiving end of a chain that began at a power plant, climbed to hundreds of kilovolts for long-distance transmission, was stepped down at one or more substations, and finally arrived at your premises at a usable voltage. That chain — generation, step-up, transmission, step-down, distribution, delivery — is the power system. It is also, at every point, exposed to faults: insulation that fails, trees that fall across lines, lightning that strikes, animals that bridge a gap, equipment that simply wears out.

When a fault happens, enormous energy concentrates at one spot in milliseconds. Left alone, it damages expensive equipment, starts fires, and can injure or kill people. Power system protection — also called protective relaying — is the discipline that detects these abnormal conditions and removes the affected part of the system quickly and accurately, before the damage spreads.

It is often described as a blend of science, art, and skill: science because it rests on circuit theory and fault analysis, art because two competent engineers can defend different settings, and skill because it is only as good as the person applying and commissioning it. The purpose statement is refreshingly simple and has not changed in a century:

Prevent injury to personnel.
Minimize damage to system components.
Limit the extent and duration of service interruption.

Everything else in this article is detail in service of those three goals.

Part 1 — Understanding Substation Protection from the Ground Up

Why protection exists at all

Case Study 1 — A transformer differential that kept tripping on energization

Utility-Scale Solar Engineering & Grid Interconnection

Tue, 30 Jun 2026 10:31:54 GMT

Jun 30, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Utility-Scale Solar & Storage: Engineering, Interconnection & Practice

Designing the array voltage window

Project snapshot

The single most consequential electrical calculation in PV design is the array voltage window — and it is driven by temperature, not by nameplate power. A string that looks correct at 25°C can over-voltage an inverter on the coldest morning of the year or fall out of the MPPT window on the hottest afternoon. Getting this wrong at utility scale multiplies the same error across millions of modules.

The cold-end limit: maximum open-circuit voltage

Module open-circuit voltage rises as temperature falls. To protect the inverter and stay within the system DC rating, the maximum string voltage must be evaluated at the coldest expected cell temperature — conventionally the ASHRAE extreme-minimum (2%) design temperature for the site. The corrected voltage uses the module’s VOC temperature coefficient (β):

VMAX = VOC × [ 1 + β × (TMIN − 25°C) ]

A worked example

A combined technical article, FAQ, and case-study collection from Keentel Engineering

Why utility-scale PV is its own engineering discipline

A rooftop array and a 150-megawatt solar power plant share the same underlying physics: silicon cells convert irradiance into direct current, modules are wired in series to build voltage, and an inverter converts that DC into grid-compliant AC. But the moment a project crosses from “behind-the-meter generation” into “a power plant that exports to the transmission system,” the engineering problem changes character. The questions stop being only about modules, conductors, and a service panel, and start being about how a multi-megawatt inverter-based resource behaves on the grid during faults, how it holds voltage at the point of interconnection, and whether the surrounding network can absorb its output.

At Keentel Engineering, our interconnection and power-systems work begins where the PV body of knowledge ends. This article walks the full arc from the module-level design fundamentals that every solar engineer must master, up through the collector system, the substation, and the studies that decide whether a plant can connect at all. The fundamentals never stop mattering; they simply become the foundation that the grid-facing engineering is built on.

The threshold that defines “utility-scale”

The National Electrical Code® recognizes large PV plants as a distinct class. NEC® Article 691 (Large-Scale Photovoltaic Electric Power Production Facility) may be applied to facilities of roughly 5 MW (5,000 kW) of generating capacity and larger that meet specific criteria the facility is not mounted on or in a building, public access is restricted, the plant is operated as a power-generation station, and the engineering is documented and stamped by a licensed professional engineer.

The practical takeaway: above this threshold, a PV plant is treated like a conventional generating station, not a building electrical system. That single distinction reshapes design, protection, commissioning, and above all interconnection.

The building blocks, scaled up

Every PV plant is assembled from the same hierarchy of components, but at utility scale each layer grows into infrastructure of its own.

From cell to source circuit

Modules are connected in series into strings (source circuits) to build DC voltage. Modern utility-scale plants are almost universally built around a 1500 V DC architecture rather than the 600 V or 1000 V systems common in distributed PV. The reason is purely economic: higher string voltage means lower current for the same power, which means fewer, longer strings, smaller DC conductors per watt delivered, and fewer combiner terminations across hundreds of acres.

From source circuit to inverter

How those source circuits reach the inverter divides utility-scale design into two dominant topologies:

Central-inverter (DC-combined) plants. Source circuits land in DC combiner boxes out in the array, the combined PV output circuits run to large central inverters (often 2–5 MW each), and the inverters feed a medium-voltage transformer. This topology minimizes the number of inverters and is well suited to large, uniform sites.
String-inverter (decentralized) plants. Many smaller three-phase string inverters are distributed throughout the field, each serving a handful of source circuits and feeding the medium-voltage collection system directly through skid-mounted transformers. This topology improves granular MPPT, simplifies replacement, and reduces single-point exposure, at the cost of more devices to monitor.

A protection nuance that matters at scale

DC arc-fault detection is straightforward on individual source circuits at a combiner, but it becomes technically impractical on large combined PV output circuits feeding a central inverter — the fault signature is swamped by the aggregate current. This is one of several reasons large-scale facilities lean on restricted access, qualified-person operation, and the alternative provisions of NEC® 691 rather than the prescriptive arc-fault rules written for accessible rooftop arrays.

From inverter to the point of interconnection

Above the inverter, the plant becomes a substation. Inverter AC output is stepped up by pad-mounted transformers to a medium-voltage

The layer the PV body of knowledge doesn’t cover: interconnection

The commissioning principles taught for any PV system scale directly into utility practice; what changes is the rigor, the documentation, and the fact that the tests are performed by qualified personnel under restricted access. The pre-energization electrical sequence still rests on a handful of fundamental tests:

Continuity of grounding and bonding. A low-impedance fault path is verified throughout the racking, equipment grounding, and grounding-electrode system. A useful field benchmark: the bonding resistance on a circuit should stay below 50 V divided by the circuit’s overcurrent rating (for example, under ~1.67 Ω on a 30 A circuit), keeping touch potentials below the hazardous threshold before a fault clears.
AC voltage and phasing. Inverter AC connections are verified for correct voltage and phase rotation before energization — a phase-rotation error at a multi-megawatt skid is not a tolerable surprise.
Insulation resistance. A megohmmeter verifies insulation integrity from ungrounded conductors to ground. At scale this is performed efficiently at the DC combiners — source circuits are isolated and tested, and combiner-to-combiner results are compared to flag the outlier that signals damaged wiring or a degrading module string.
Open-circuit voltage and polarity. Each source circuit’s V₀ᴄ and polarity are confirmed against the temperature-corrected design value the field check on the voltage-window calculation above.
Short-circuit current. Source-circuit Iₛᴄ is verified under controlled conditions to confirm string health and current balance across the array.

Beyond the pre-energization checks, I–V curve tracing characterizes representative strings against their expected curves to catch soiling, mismatch, shading, or connection problems early, and a performance-ratio test confirms the plant delivers its modeled energy. A first-order annual energy check peak-sun-hours × installed DC kW × system factor anchors expectations before detailed production modeling refines them.

Temperature-corrected per-module maximum: VMAX = 49.6 × [1 + (−0.0027)(−12 − 25)] = 49.6 × 1.100 ≈ 54.6 V. Dividing the 1500 V ceiling by 54.6 V yields 27.5, so the design is capped at 27 modules in series. That string reaches 27 × 54.6 ≈ 1473 V on the coldest morning safely under 1500 V.

The hot-end limit: minimum operating voltage

At the other extreme, maximum-power voltage (VMP) falls as cells heat. The string must stay above the inverter’s minimum MPPT voltage at the hottest expected operating temperature, or the plant sheds energy by tracking off its true maximum-power point. Checking the same 27-module string with a −0.35 %/°C VMP coefficient at a 70°C cell temperature gives roughly 27 × 35.2 ≈ 951 V — comfortably above the 875 V floor. The string design therefore survives both temperature extremes, which is the entire objective. We also subtract a small annual voltage-degradation allowance (about 0.5–1%/year) so the window remains valid across the 30-plus-year life of the plant.

Why this scales into real money

A one-module error in string length, applied across a 150 MW plant, can shift hundreds of thousands of strings. Choosing 27 versus 26 modules changes conductor counts, combiner loading, inverter utilization, and land use simultaneously. The voltage-window calculation is where module physics, inverter specifications, and site climate meet and it is the foundation every downstream system depends on.

PV is a current-limited source, and the Code sizes conductors with that in mind. Under NEC® 690.8, the maximum source-circuit current is the module short-circuit current multiplied by 1.25 (to account for irradiance above 1000 W/m²), and conductor ampacity is then sized with a further 1.25 continuous-duty factor before temperature and conduit-fill derating. Over hundreds of acres, these factors interact with two competing costs: copper/aluminum tonnage and lifetime energy lost to voltage drop.

At utility scale, voltage-drop budgeting stops being a code-minimum check and becomes a lifecycle optimization. A conductor sized to the bare ampacity minimum may pass inspection yet bleed energy for thirty years; an oversized run wastes capital up front. The right answer balances the present value of lost production against incremental conductor cost a calculation repeated across DC homeruns, the medium-voltage collector feeders, and the GSU connection. This is why collector-system design is treated as an engineering optimization, not a lookup.

Everything above produces a plant that works as a self-contained generator. Whether it is allowed to connect and how it must behave once connected is a separate engineering domain, and it is the core of Keentel’s practice. Three things distinguish a transmission-connected inverter-based resource from a building PV system:

1. It must ride through grid disturbances

A rooftop inverter is required to disconnect during a utility disturbance for safety. A utility-scale plant is required to do the opposite: ride through voltage and frequency excursions and keep supporting the grid, because a fleet of large resources tripping offline during a fault would itself become a reliability event. Ride-through and grid-support behavior are governed by standards such as IEEE 1547 for distribution-connected resources and IEEE 2800 for transmission-connected inverter-based resources, layered with NERC reliability requirements. Demonstrating compliance is an analysis problem, not a nameplate claim.

2. Its grid behavior must be modeled and proven

Interconnectio n studies require validated models of how the plant responds to the network. Keentel develops both positive-sequence dynamic models (PSS®/E) for system-wide stability and power-flow studies and electromagnetic-transient models (PSCAD™/EMTDC™) for the fast, fault-driven behavior of inverter controls that positive-sequence tools cannot capture. These models feed the feasibility, system-impact, and facilities studies that an ISO/RTO or utility runs before assigning an interconnection agreement and cost responsibility.

3. It must hold voltage and supply reactive power

Beyond real-power export, a utility-scale plant is expected to manage voltage at the POI and provide a specified reactive-power (VAR) capability across its operating range, coordinated by a plant controller. Reactive capability, voltage set-point control, and the protection scheme that isolates the plant for genuine faults while riding through transient events are all engineered, modeled, and field-verified.

Where the fundamentals and the grid meet

The string voltage window, conductor sizing, and combiner architecture decide what the plant can produce. The interconnection engineering ride-through, dynamic modeling, reactive capability, and protection decides whether and how that production is allowed onto the grid. A successful project needs both, and a defect in either layer can stall a project in the interconnection queue or in commissioning.

Commissioning a power plant

A utility-scale plant is a 30-plus-year asset, and its economics live or die on availability and performance retention. Module degradation (typically a fraction of a percent per year), inverter availability, soiling losses, vegetation and tracker maintenance, and connection integrity all compound over the asset life. A disciplined O&M program grounded in the same commissioning tests used as periodic diagnostics protects the production assumptions that the project was financed against. The insulation-resistance trend that looked like a commissioning formality becomes, years later, the early-warning system for a degrading array.

Operations, maintenance, and long-term performance

Keentel Engineering works across the full lifecycle of utility-scale solar and storage from interconnection feasibility through detailed design, modeling, and owner’s-engineer support:

Point-of-interconnection (POI) engineering feasibility, system-impact, and facilities study support, and interconnection-agreement technical review.

EMT and positive-sequence modeling PSCAD™/EMTDC™ and PSS®/E models for ride-through, stability, and protection studies.

Substation and transmission-line design collector systems, GSU and POI substations, and protection coordination.

Utility-scale renewables and BESS engineering PV, storage, and hybrid plant design and integration.

Owner’s engineer services and NERC O&P compliance independent review, commissioning oversight, and reliability-standard support.

Talk to Keentel

Planning a utility-scale PV or PV-plus-storage project or stuck on an interconnection study , a ride-through requirement, or a model-validation issue? contact@keentelengineering.com · 813-389-7871 · keentel.com · Tampa, FL & Austin, TX.

Where Keentel fits

Part II — Three Utility-Scale Engineering Case Studies

Conductors, current, and the economics of voltage drop

The challenge

Construction was complete and the owner needed confidence that the plant would meet its financed production target before final acceptance from the EPC. Early energization data showed two combiner zones underperforming their neighbors, and a handful of string inverters were reporting intermittent ground-related alarms. The owner engaged Keentel to run a disciplined commissioning and performance-verification program and to determine whether the issues were construction defects to be corrected before acceptance — or noise.

Keentel’s approach

Ran the pre-acceptance electrical test sequence. Continuity of grounding and bonding, AC voltage and phasing, insulation resistance, open-circuit voltage/polarity, and short-circuit current were verified and documented across the plant by qualified personnel under restricted access.
Isolated the underperformance with insulation-resistance comparison. Insulation resistance was measured at the DC combiners and compared zone-to-zone. The two outlier zones returned readings well below their peers — a clear, repeatable signature pointing to damaged conductor insulation rather than soiling or shading.
Confirmed the diagnosis with I–V curve tracing. Representative strings in the suspect zones were curve-traced against expected I–V characteristics, separating mechanical/wiring faults from module mismatch and confirming the specific strings needing rework.
Verified plant performance against the model. After corrections, a performance-ratio test confirmed the plant delivered its modeled energy, anchored by a first-order peak-sun-hours energy check and refined with detailed production modeling.

Engineering highlights

Comparison testing beats absolute thresholds. At scale, the fastest way to find the defect is to test identical sub-systems and flag the outlier — combiner-to-combiner insulation-resistance comparison turned a vague “two zones look low” into a precise rework list.

Acceptance backed by data. Documented test records gave the owner a defensible basis to require EPC corrections before final acceptance, protecting the production warranty and the financing assumptions.

Outcome

The underperforming zones were traced to a small number of insulation faults, corrected by the EPC before acceptance, and re-verified. The plant passed its performance-ratio test, and the owner accepted the asset with a complete commissioning record — turning an ambiguous early-data concern into a closed, documented item.

How Keentel supports utility-scale projects

Electrical design

Interconnection

Storage & hybrids

Commissioning & operations

Project snapshot

The challenge

The project had a viable site and a signed module supply agreement, but its interconnection request had stalled. The ISO’s system-impact study flagged two concerns: the plant’s low-voltage ride-through behavior during nearby transmission faults could not be confirmed from the manufacturer’s generic models, and a weak-grid stability question at the POI required electromagnetic-transient analysis that the developer’s original consultant had not provided. Without validated models, the pr oject risked falling to the back of the queue and absorbing avoidable network-upgrade cost allocations.

Keentel’s approach

Built and validated dual models. Keentel developed a positive-sequence dynamic model (PSS®/E) for power-flow and stability studies and an electromagnetic-transient model (PSCAD™/EMTDC™) tuned to the actual inverter control parameters, so the plant’s sub-cycle fault response could be demonstrated rather than assumed.
Demonstrated ride-through compliance. Fault scenarios at and near the POI were simulated to confirm the plant remained connected and provided grid support through the required voltage and frequency envelopes, consistent with IEEE 2800 and the applicable NERC requirements.
Resolved the weak-grid question. The EMT study characterized controller interaction with the local network at low short-circuit strength, and informed plant-controller tuning and reactive-capability settings that kept the plant stable across the studied operating range.
Reviewed the interconnection package as owner’s engineer. Keentel independently checked the collector-system design, GSU and POI protection coordination, and reactive-capability commitments before they were locked into the interconnection agreement.

Engineering highlights

Voltage-window discipline. The 1500 V string design was confirmed at the site’s ASHRAE extreme-minimum temperature for the cold-end limit and at peak cell temperature for the MPPT floor, with annual voltage degradation included — protecting both the inverter rating and lifetime energy capture.

Protection that rides through real events. The POI protection scheme was coordinated to isolate the plant for genuine faults while honoring the ride-through requirement for transient disturbances.

Outcome

With validated PSS®/E and PSCAD™ models and a defensible ride-through demonstration, the project cleared the study concerns and retained its queue position. The independent owner’s-engineer review caught two collector-protection coordination items before construction — far cheaper to fix on paper than in the field.

Case Study 2 — 100 MW PV + 50 MW / 200 MWh Storage: Engineering a Dispatchable Hybrid

Project snapshot

The challenge

The owner wanted to add four-hour storage to a PV project to firm output and capture energy-arbitrage value — without increasing the export limit at the POI. That created three coupled engineering problems: the plant now had to manage bidirectional power flow (charging from the array and the grid, discharging on dispatch); the interconnection models had to represent the resource’s behavior in both directions; and the export had to be held to 100 MW even when the PV and battery could momentarily deliver more, which is a controls and protection problem, not just a nameplate one.

Keentel’s approach

Selected and engineered the coupling architecture. An AC-coupled configuration at the collector bus was adopted to keep the PV and storage blocks independently serviceable and to simplify export-limit enforcement, with the trade-offs against a DC-coupled alternative documented for the owner.
Modeled the resource in both power-flow directions. PSS®/E and EMT models represented charging and discharging behavior, ride-through in both modes, and the plant controller’s response — because an interconnection study for a hybrid cannot assume a one-way generator.
Engineered the export-limit and reactive-support controls. The plant controller logic was designed to cap net export at the POI across all PV/BESS operating combinations while still meeting the reactive-capability and voltage set-point requirements at 230 kV.
Addressed storage-specific design. Bidirectional protection, state-of-charge-aware operation, and storage code provisions (e.g., NEC® Article 706 framework) were integrated into the overall plant design and reviewed independently as owner’s engineer.

Engineering highlights

One POI, two resources. The interconnection agreement was supported with models proving the combined plant respected the 100 MW export limit and ride-through requirements regardless of dispatch state.
Firming without a bigger interconnection. By holding the export limit and adding storage behind it, the project improved its capacity value and dispatchability without triggering the larger network upgrades a higher export request would have required.

Outcome

The hybrid was engineered to deliver firmer, more dispatchable output and arbitrage value while staying within its existing 100 MW interconnection envelope. Bidirectional modeling and controller logic that enforced the export cap let the project add four-hour storage without re-opening the interconnection capacity — the most expensive door to reopen on a utility-scale project.

Case Study 3 — 80 MW String-Inverter Plant: Commissioning and Performance Verification at Scaleph

Across these problem classes — interconnection, hybrid integration, and commissioning — Keentel Engineering works the full lifecycle of utility-scale solar and storage:

POI / interconnection engineering — feasibility, system-impact, and facilities study support and interconnection-agreement review.
EMT and positive-sequence modeling — PSCAD™/EMTDC™ and PSS®/E for ride-through, stability, and protection.
Substation & transmission-line design — collector systems, GSU and POI substations, protection coordination.
Utility-scale renewables & BESS engineering — PV, storage, and hybrid plant design and integration.
Owner’s engineer & NERC O&P compliance — independent review, commissioning oversight, and reliability support.

Start a conversation

Have a project facing one of these challenges? contact@keentelengineering.com · 813-389-7871 · keentel.com · Tampa, FL & Austin, TX.

Part III — Utility-Scale Solar & Interconnection FAQ

Plain-language answers to the questions asked most about large PV and storage projects

This FAQ collects the questions Keentel Engineering hears most often from developers, asset owners, EPC partners, and AHJs working on utility-scale solar and solar-plus-storage projects. Answers are general engineering information; project-specific, PE-stamped engineering and the governing code edition always control.

Defining the project

Part I — From PV Fundamentals to Utility-Scale Interconnection

How module-level physics scales into a transmission-connected power plant

Case Study 1 — 150 MW Single-Axis-Tracker PV Plant: Clearing the Interconnection Queue

Data Center Design Guide | Power, Cooling & Reliability

Mon, 29 Jun 2026 18:17:47 GMT

June 29, 2026 | White Paper

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

DATA CENTER DESIGN

Power, Cooling, and Reliability from the Grid to the Rack

Transmission Line Design: From 30% Design to IFC Guide

Mon, 29 Jun 2026 14:59:41 GMT

Jun 29, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Part One — Technical Blog

From 30% to IFC: How a Transmission Line Gets Designed

The 60% milestone intermediate design

Part One — FAQS

60% is where the line becomes structurally and electrically real. Loading is applied rigorously, structures are designed (or selected from the family) to those loads, and clearances are re-verified against the now-final wire conditions. The route is held; the engineering inside it is hardened.

What gets defined

Finalized structure spotting. Structures are optimized against the survey — balancing span length, structure count, height, and uplift — with angle and dead-end structures placed at PIs and line ends.
Structural loading and structure design . NESC district loads plus extreme wind and ice are applied per ASCE 74; each structure type is checked for strength against the controlling load combination, including longitudinal (broken-wire / unbalanced) cases on dead-ends.
Final sag-tension. Initial/final, everyday and maximum-condition tensions are locked, with the governing thermal case set from ampacity and the loading cases set from NESC.
Foundation preliminary design. Foundation type (drilled shaft, direct-embed, spread footing, pile) is selected per structure loading and the geotechnical data now in hand.
Insulation and hardware. Insulator strings, suspension/dead-end assemblies, dampers, and spacers are specified; phase spacing and swing are confirmed for switching and lightning overvoltage.
Grounding and shielding. Shield-wire/OPGW positioning is set for the target shielding angle; structure grounding is sized to footing-resistance targets.
Clearance re-verification. All ground, crossing, and edge-of-ROW clearances are re-checked against the final wire positions across the full load matrix.

The basis of design: what's true at every gate

Before the milestones make sense, a handful of inputs govern the entire design and are established up front in the Basis of Design (BOD). These do not change milestone to milestone unless the project scope itself changes — they are the constants the gates refine against.

Electrical basis

Nominal and maximum operating voltage, continuous and emergency (e.g., summer/winter, normal/contingency) ampacity ratings, and the resulting maximum conductor operating temperature. Ampacity drives conductor selection and the thermal case that governs clearance — a line rated for 100°C operation sags far more than one rated for 75°C, and that sag is what eats clearance.

Code and loading basis

The governing edition of the National Electrical Safety Code (NESC) sets minimum vertical and horizontal clearances and the NESC district loading (heavy/medium/light) plus extreme-wind and extreme-ice cases. ASCE Manual 74 informs structural loading methodology wind pressures, ice accretion, and the load factors applied to wire tensions and structure self-weight. State and utility amendments frequently add to these minimums.

Mechanical basis

Conductor and shield-wire/OPGW type, the tension limits (initial and final, everyday and maximum), and the ruling span assumptions feed sag-tension analysis. These set how tightly the wire is strung, which trades directly against structure height, span length, and fatigue life.

The 30% milestone preliminary design

Question answered: Can this line exist on this corridor, and roughly what does it cost?

At 30%, the goal is to prove feasibility and fix the route and the major design choices. The deliverable set is preliminary, but the decisions it captures are the most consequential of the whole project — a wrong call here propagates through every later gate.

What gets defined

Route selection and centerline. The corridor is chosen against constraints — wetlands, existing infrastructure, parcels, environmental and cultural resources — and a preliminary right-of-way (ROW) width is set from phase spacing and blowout.

Survey and terrain model. LiDAR or photogrammetric survey, or at minimum existing-data terrain, is loaded into the line-design model so the ground profile is real, not assumed.
Structure family. Steel pole vs. lattice, single vs. double circuit, tangent/angle/dead-end types, and a first cut at typical heights and a structure-type library.

Conductor and shield/OPGW selection. Size and type chosen to meet ampacity with margin; OPGW fiber count coordinated with the communications scope.

Preliminary sag-tension and structure spotting. A first automated spotting run places structures against the profile to test that spans, heights, and clearances are achievable.

Major clearance check. Ground clearance, crossings (roads, rail, other lines), and edge-of-ROW clearance to the worst-case (hot, blowout) condition.

Geotechnical and environmental kickoff. Preliminary subsurface assumptions and the permitting/environmental strategy are initiated — long-lead items that must start early.

Where 30% goes wrong

Spotting structures on an assumed flat profile instead of real survey every clearance is then fictional.

Choosing a conductor for ampacity alone and discovering at 60% that its sag forces taller, costlier structures.

Setting ROW width before blowout (wind swing) is checked, then renegotiating easements later.

30% deliverables

Preliminary plan-and-profile (P&P) sheets with spotted structures
Basis of Design document and design criteria
Preliminary structure list and material quantities (for early estimating)
Route map / alignment exhibit and ROW width basis
Class 4/5 cost estimate input and risk register

Issued for Construction (IFC)

Several engineering threads are not owned by any single milestone they tighten progressively at every gate. Understanding them as continuous, rather than one-time tasks, is what separates a coordinated design from a set of disconnected calculations.

Clearance

Checked preliminarily at 30%, rigorously at 60% against final wire positions, and documented at IFC. Clearance is governed by the worst-case sag — usually maximum operating temperature — not the everyday condition, which is why ampacity and sag-tension must be settled before clearance can be trusted.

Structural loading

From an NESC district assumption at 30% to a full load matrix (district, extreme wind, extreme ice, broken-wire) at 60%, to detailed foundation reactions at 90%. Loading is the spine that connects wire choices to structure choices to foundation choices.

Sag-tension

Preliminary at 30%, final at 60%, and expressed as stringing charts at 90%/IFC. Tension limits balance clearance (lower sag wants higher tension) against fatigue from aeolian vibration (high tension shortens conductor life), so they are an engineering compromise, not a maximum.

Right-of-way, environmental, and permitting

Initiated at 30% because it is long-lead, advanced through 60%/90%, and closed before IFC. Unresolved crossings or easements are among the most common causes of IFC slippage — the engineering can be finished while the line still cannot legally be built.

A transmission line is not drawn once and built. It is converged on deliberately, in design gates that each retire a defined band of risk before the next dollar is committed.

Every overhead transmission line is a negotiated settlement between physics, terrain, code, cost, and schedule. The conductor wants to sag; the code wants clearance; the landowner wants the structure somewhere else; the utility wants the lowest installed cost that still survives an ice storm in fifty years. No single calculation resolves all of that at once. So the industry designs in milestones — commonly 30%, 60%, 90%, and Issued for Construction (IFC) with each gate freezing a layer of decisions so the next layer has firm ground to stand on.

The milestone framework is not bureaucracy. It is risk sequencing. Route and electrical basis are settled first because everything downstream depends on them; foundation rebar is detailed last because it depends on everything upstream. Reordering those steps is how projects generate expensive rework. This guide walks the full progression: what each gate defines, what deliverables it produces, and where the costly mistakes hide.

The one-sentence version

30% proves the line can exist on this route; 60% proves it stands up to load and clearance; 90% proves it can actually be built; IFC is the buildable, stamped, coordinated package the contractor strings steel from.

Where 60% goes wrong

Designing structures to district load but skipping the longitudinal broken-wire case — then dead-ends are undersized.

Locking sag-tension before the final ampacity / max-temperature case is confirmed, so clearance is checked to the wrong wire position.

Foundations sized before geotech returns, forcing redesign when actual soil parameters arrive.

60% deliverables

Updated P&P with final spotting and clearance annotations
Structure loading trees / loading summary per structure type
Preliminary foundation design and reactions
Sag-tension report and stringing condition tables
Insulation, hardware, and grounding specifications
Updated structure list and refined material take-off

90% is the constructability and completeness gate. The design is essentially done; the work is closing gaps, finalizing details that depend on procurement and field input, and circulating the package for formal review. Few new decisions are made the focus is rigor, coordination, and catching the things that only show up when the whole set is assembled.

What gets defined

Final foundation design. Drilled-shaft reinforcement, anchor bolts, embedment, and rebar detailing completed against final reactions and final geotechnical recommendations.
Complete structural detailing. Final structure drawings, framing, attachment hardware, climbing/working provisions, and any guying.
Stringing and construction sequencing. Stringing charts, sag/tension setting tables, and sagging-condition references for the field; pulling/tensioning plan inputs.
Material take-off finalization. Full bill of materials reconciled against the structure list for procurement — conductor, OPGW, insulators, hardware, steel, foundations.
Constructability review. Access, lay-down, crane/equipment reach, outage/clearance coordination with adjacent energized facilities, and crossing methods.
Interdisciplinary and crossing coordination. Coordination with substation terminations, telecom (OPGW splicing/regen), road and utility crossings, and ROW/permitting closeout.
Formal review cycle. Internal QA/QC and client/utility review comments incorporated; open items tracked to closure.

Where 90% goes wrong

Treating 90% as ‘almost IFC’ and skipping a true constructability review — access and outage problems then surface in the field.

Bill of materials drifting out of sync with the structure list, causing procurement shortfalls or surplus.

Leaving crossing permits or substation interface details ‘to be resolved,’ which become IFC-blocking items.

90% deliverables

Near-complete drawing set: P&P, structure, foundation, grounding, hardware details
Final foundation design package and reactions
Stringing charts and sagging tables
Reconciled bill of materials / final take-off
Constructability review report and comment-resolution log

IFC is not a further design milestone so much as the formal release of a complete, internally consistent, professionally sealed package the contractor builds from. Every comment is resolved, every drawing cross-checks against every other, and a licensed professional engineer takes responsibility by stamp and signature. Changes after IFC flow through formal revision control they are tracked, costed, and re-issued, never quietly redlined.

What an IFC package contains

Complete, sealed drawing set plan-and-profile, structure drawings, foundation drawings, grounding, hardware and assembly details, and general notes.

Final bill of materials and procurement-ready specifications.

Stringing charts, sag-tension setting tables, and installation tolerances.

Construction notes, special conditions, crossing details, and access/outage requirements.

As-designed clearance documentation demonstrating NESC compliance across the load matrix.

Revision block and document control established for field changes (RFIs, IFC revisions).

The IFC discipline

After IFC, design intent is fixed and field changes are managed through revision control and RFIs. The value of the milestone gates is that, by IFC, almost nothing is left to discover the line on paper and the line in the field are the same line.

Threads that run the full length of the design

Designing against assumed terrain. Without survey-grade ground at 30%, every clearance and every span is a guess. The 30% gate exists to make terrain real before structures are committed.
Selecting conductor on ampacity alone. The cheapest conductor that carries the current may sag enough to force taller structures and longer foundations. Sag economics belong in the 30%/60% conductor decision.
Skipping load cases. District loading alone undersizes dead-ends and angle structures; the broken-wire and extreme-event cases at 60% are where those structures are actually sized.
Detailing foundations before geotech. Foundation rebar is the last thing to fix because it depends on the most upstream uncertainty — soil. The gate order protects against costly re-design.
Treating 90% as a formality. The constructability review is the cheapest place to find an access, outage, or crossing problem. Found in the field, the same problem is a change order.

The expensive mistakes and how the gates prevent them

Keentel Engineering designs transmission lines as a controlled convergence, not a single deliverable. Each gate is treated as a genuine decision point with a defined exit standard, so that risk is retired in the right order and downstream rework is minimized. Route and electrical basis are settled before structures; structures are designed to a full load matrix before foundations are detailed; and constructability is reviewed before the package is sealed.

Across milestones, Keentel maintains a single coordinated model survey, sag-tension, structure loading, and clearance live together rather than in disconnected spreadsheets so that a change in any one input propagates to every dependent check. The result at IFC is a package where the line on paper and the line in the field are the same line, and where the contractor builds from a set that has already answered the questions the field would otherwise raise.

Keentel transmission line design services

Route studies, structure spotting, and plan-and-profile development
Conductor and OPGW selection, sag-tension, and stringing charts
Structure and foundation design to NESC and ASCE 74
Clearance compliance, grounding, and insulation coordination
30% / 60% / 90% / IFC milestone packages and owner’s engineer review
contact@keentelengineering.com • 813-389-7871 • keentel.com

How Keentel approaches the progression

Standards & references

This article reflects general industry practice grounded in the following publicly available standards and references. Editions in force on a given project govern; the list is provided for orientation, not as project-specific design criteria.

National Electrical Safety Code (NESC), IEEE clearances and structure/wire loading (Rules 232–235, 250)

ASCE Manual of Practice 74 — Guidelines for Electrical Transmission Line Structural Loading

IEEE Std 738 — Calculating the Current-Temperature Relationship of Bare Overhead Conductors

IEEE Std 1313 — Insulation Coordination

IEEE Std 1283 / CIGRE — conductor thermal and aeolian vibration / fatigue guidance

Conductor and hardware manufacturer data; applicable ASTM material standards

Applicable state and utility-specific amendments to the above

PART TWO — Case Studies

The 90% milestone pre-final / review design

Plain answers to the questions clients ask about milestone-based line design.

The milestone framework

Inputs and standards

Conductors, sag, and clearance

Structures and foundations

Schedule, deliverables, and risk

Working with Keentel

Three anonymized milestone narratives — 30%, 60%, and the 90%→IFC boundary.

About these case studies

The following three case studies are illustrative and fully anonymized. They contain no client names, project names, utility names, or geographic locations. Voltages, configurations, and engineering specifics are representative of standard practice and are presented to show how milestone-based design resolves real problems. Any resemblance to a specific project is coincidental.

Case Study 1 — A 30% clearance finding that changed ructuramily

A single-circuit rebuild where the most important decision of the project was made at the first gate.

Engagement at a glance

Situation

An investor-owned utility client needed to increase the thermal capacity of an existing single-circuit line whose structures were near end of life. The initial direction was a like-for-like rebuild on the existing alignment, reusing the existing structure type with a larger conductor to meet the higher ampacity target. On paper, reusing the structure family promised the lowest cost and the simplest permitting, since the alignment and ROW were unchanged.

What the 30% gate revealed

When Keentel loaded survey-grade terrain into the line model and ran a preliminary sag-tension and spotting pass, the larger conductor — selected to meet the new ampacity — sagged enough at its maximum operating temperature to violate ground clearance across several spans where the original conductor had been compliant. The original structures had been sized for a smaller, lighter conductor strung at a different tension; the new thermal duty changed the governing sag case entirely.

Critically, this surfaced at 30% precisely because the check was run against real surveyed ground rather than the original design assumptions. Had the project advanced on the assumption that ‘same alignment, same structures’ was safe, the clearance shortfall would not have appeared until 60% — after structure procurement assumptions, estimates, and schedule had already been built around the wrong structure family.

The pivotal trade-off

Option A: keep the existing structure type and raise tension to reduce sag — rejected, because higher tension worsened aeolian vibration fatigue and still did not clear every span.

Option B: keep the structure type and shorten spans by adding structures — rejected on cost and ROW disturbance.

Option C: move to a taller structure family on the existing alignment — selected, because it recovered clearance across the full thermal matrix without adding structures or over-tensioning the conductor.

How it resolved through the gates

30%: Adopted a taller single-circuit structure family on the existing alignment; re-spotted to confirm clearance across the maximum-temperature case; updated the Basis of Design and preliminary estimate to reflect the structure change.

60%: Designed the new structure type to the full NESC district and extreme load matrix; finalized sag-tension at a tension that balanced clearance against fatigue; confirmed foundations were compatible with the existing ROW.

90%: Completed foundation detailing, stringing charts, and a constructability review covering outage coordination with the energized system during rebuild.

IFC: Issued a sealed package with documented clearance compliance across all load cases.

Outcome

The line met its new thermal target with full clearance compliance and no ROW expansion. Because the structure-family decision was made at 30% rather than discovered at 60%, the project avoided a mid-design re-estimate and the schedule slip that a late structure change would have forced. The episode is a clean illustration of why the 30% gate exists: the cheapest place to change a structure family is before anything has been built around it.

Lesson

Reusing an existing structure family is only safe if the new conductor’s governing sag case is checked against surveyed terrain. Ampacity changes the thermal sag, and thermal sag — not the everyday condition — governs clearance.

Case Study 2 — A 60% geotechnical and loading interaction that reshaped foundations

A greenfield double-circuit line where the load matrix and the soil report had to be reconciled before foundations could be trusted.

Engagement at a glance

Situation

A regional transmission provider was building a new double-circuit line to interconnect new generation. As a greenfield route, it carried the usual front-loaded ROW and environmental work, and the early estimate assumed a single common foundation type — a drilled shaft sized from generic soil parameters — across the line. This is a reasonable 30% assumption, but it is only an assumption until the subsurface investigation returns.

What the 60% gate revealed

Two things converged at 60%. First, the structural load matrix was applied rigorously: in addition to NESC district loading, the extreme-ice-with-wind case and the longitudinal broken-wire case on the angle and dead-end structures produced substantially higher foundation reactions than the everyday case had implied — particularly overturning moment at the larger line angles. Second, the geotechnical investigation returned variable subsurface conditions along the route, with weaker strata in part of the alignment than the generic 30% assumption.

The intersection of those two facts — higher-than-assumed reactions at angle structures, meeting weaker-than-assumed soil in some reaches — meant a single uniform foundation type would have been over-designed in good soil and under-designed in poor soil. Designing foundations before either input was firm would have produced exactly the kind of redesign the gate order is meant to prevent.

How it resolved through the gates

60%: Reconciled the full load matrix with the geotechnical report; segmented the line into foundation design reaches by soil condition; sized foundations per structure type and per reach rather than to a single uniform type.

90%: Completed final reinforcement detailing for each foundation reach; reconciled the bill of materials to the now-varied foundation quantities; constructability review confirmed access and drilling methods for the weaker-soil reaches.

IFC: Sealed package with foundation schedules keyed to structure and reach, and clear field documentation of which design applied where.

Outcome

Foundations were right-sized to both load and soil — neither uniformly over-built nor locally under-built. Because the loading matrix and the geotechnical data were deliberately reconciled at 60%, before any foundation reinforcement was detailed, the project avoided detailing and then re-detailing foundations when the soil report landed. The case shows why foundation design sits late in the gate order: it depends on the two inputs (final reactions and real soil) that are themselves only firm at 60%.

Lesson

Foundations are governed by the intersection of the load matrix and the subsurface investigation. A uniform foundation assumption is fine at 30% for estimating, but foundations should not be detailed until both the broken-wire/extreme cases and the geotechnical report are in hand.

Case Study 3 — An IFC constructability resolution under live-line and crossing constraints

An uprate where the engineering was essentially settled by 90%, but the package could not be issued until construction sequencing and a major crossing were resolved.

Engagement at a glance

Situation

A public power utility was uprating an existing 345 kV line, with a short relocation segment, to increase capacity. By 90% the line engineering was sound: structures were designed to the full load matrix, sag-tension and clearances were verified, and foundations were detailed. The remaining risk was not analytical — it was whether the line could actually be built within the available outage window while adjacent circuits stayed energized, and how to execute a crossing of a major existing facility along the route.

What the 90% constructability review revealed

The constructability review — the formal purpose of the 90% gate — surfaced three issues that, while not changing the design intent, would have blocked or endangered construction if they had reached the field unresolved:

Outage sequencing. The available outage window was too short to rebuild the full segment in one pass while keeping the parallel circuit energized; the stringing sequence had to be staged so that clearances to the energized circuit were maintained at every step, not just in the final condition.
Crossing execution. The line crossed a major existing facility, requiring a specific guard-structure and stringing method to protect the facility during conductor pulling — a means-and-methods detail that had to be reflected in the construction notes.
Access and lay-down. Part of the relocation segment had constrained equipment access, limiting crane reach and dictating which structures could be set by which method — affecting the construction sequence and schedule.

Why this belonged at the 90% → IFC boundary

None of these issues changed a calculated quantity. All three changed how and in what order the line gets built. The constructability review is the cheapest place to find them; the same issues found in the field during a live outage are safety events and change orders.

How it resolved into IFC

Staged the stringing and structure-setting sequence so that clearance to the energized parallel circuit was maintained at every intermediate construction state, and documented that sequence in the construction notes.

Specified the crossing guard structure and stringing method and added it to the IFC drawings and notes, with the temporary clearances called out explicitly.

Annotated access and equipment-method constraints by structure so the contractor’s plan aligned with the as-designed assumptions.

Closed the open items through the formal review-comment log before sealing, so the IFC set carried no ‘to be resolved’ notes.

Outcome

The IFC package was issued as a genuinely buildable set: the contractor received not only the line design but the sequence, crossing method, and access constraints that made it constructible within the outage window and safe alongside energized circuits. Because the constructability review was treated as a real gate rather than a formality, the issues were resolved on paper at the lowest possible cost — rather than discovered during a live-line outage, where they would have been the most expensive and the most dangerous.

Lesson

By 90%, analytical design can be complete while the project is still not buildable. The constructability review exists to convert a correct design into a build sequence — outage staging, crossings, access — so that IFC is not just ‘done’ but executable.

What the three cases share

Each case turned on a different gate 30%, 60%, and the 90%→IFC boundary — but all three illustrate the same underlying principle: the milestone framework retires risk in the order the design actually depends on it. The decisive issue in each project was caught at the gate designed to catch it, and therefore resolved on paper rather than in the field.

How Keentel applies this

Keentel treats every gate as a real decision point with a defined exit standard, working from a single coordinated model so survey, sag-tension, loading, and clearance stay consistent. The aim is simple: by IFC, the line on paper and the line in the field are the same line.

Transmission line design 30% through IFC, plus owner’s engineer review.

contact@keentelengineering.com • 813-389-7871 • keentel.com

From 30% to IFC — how a transmission line is designed, milestone by milestone.

Why a transmission line is designed in stages

Question answered: Does the line survive its loads and hold every clearance?

Question answered: nothing — all questions are closed. IFC is the buildable, coordinated, stamped package.

Notice

This document is published by Keentel Engineering for general professional information. It reflects standard industry engineering practice and Keentel's own methodology; it is not project-specific design guidance and does not establish an engineer-client relationship.

All case studies are illustrative and fully anonymized: no client names, project names, or locations are disclosed. Any resemblance to a specific project is coincidental.

Standards, codes, and organizations referenced (e.g., NESC, ASCE, IEEE, IEC, ASTM, OSHA) are the property of their respective owners. Keentel Engineering is independent and not affiliated with, endorsed by, or sponsored by any such organization.

AP1000 Supply Chain: Nuclear Grid Interconnection Guide

Mon, 29 Jun 2026 07:37:52 GMT

Jun 29, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Executive Summary

AP1000 Supply Chain: Grid Interconnection & Licensing Guide

The Lessons That Shape Every Boardroom Decision

An engineering and interconnection perspective on the federal supply-chain loan program for large light-water reactors, the bird-in-hand value of existing combined operating licenses, and what it takes to connect new nuclear units to the bulk power system.

Frequently Asked Questions

No serious discussion of new large reactors in the United States happens without reference to the two most recent attempts. The completed two-unit project in Georgia delivered working reactors, but only after roughly fifteen years of construction and a final cost in the mid-$30-billion range well above original estimates. A parallel two-unit project in South Carolina fared worse: it was abandoned in 2017 amid management failures across the utility, the vendor, and the contractors, leaving ratepayers with billions in stranded debt, pushing the vendor into bankruptcy, and resulting in fraud convictions and prison terms for several executives.

The South Carolina site has since attracted a new owner the private-equity firm that now controls the reactor vendor with a p lan to complete the two partially built units. A final investment decision is not expected before 2028, and because the original licenses were terminated, the project would require new licensing. Industry-wide, that completion effort is the bellwether : its cost and schedule outcome will heavily influence whether any utility commits to entirely new units, with or without federal incentives.

The Supply-Chain Loan Program: What Was Announced

The program issues a conditional loan commitment to finance the purchase of long-lead equipment needed to restart the domestic nuclear supply chain. As structured, it would back up to five projects, each pairing the reactor vendor with a utility or energy-company partner, and each delivering two large reactors at a single site. The combined facility size of the program is ten units. A separate, much smaller construction-loan facility sits alongside the equipment program.

The financing is a commitment, not a disbursement. Before funds flow, the parties must satisfy technical, legal, environmental, and financial conditions and execute definitive agreements. Each project also carries a substantial upfront equity requirement understood to be on the order of half a billion dollars from each of the two project owners before loan funds can be drawn which keeps real capital at risk and tests the seriousness of every prospective partner.

The vendor has reportedly signed letters of intent with several potential partners, each with an identified site, though none of the partners or sites have been named publicly. That secrecy is itself instructive: publicly traded utilities know they will face hard questions from boards and shareholders about whether a new large-reactor commitment meets the “prudent investor” standard, given the recent history of cost and schedule performance on this reactor class in the United States.

Why Long-Lead Equipment Is the Critical Path

Large reactors are gated by a handful of components that simply cannot be produced quickly. Reactor pressure vessels, steam-supply equipment, primary coolant pumps, and main turbine-generator sets each require specialized heavy-forging and fabrication capacity, and many of these items take three years or more to manufacture and deliver. A meaningful share of this capacity sits with heavy-industry firms in Japan and South Korea, because domestic capability atrophied during the long gap in new construction.

This is the strategic logic of financing equipment first. By aggregating bulk orders across multiple projects at a fixed price, the program aims to lower per-component cost, create supply-chain efficiencies, and critically compress schedule by getting forgings into the queue before sites are even finalized. For an engineering team, the implication is direct: the procurement clock and the interconnection clock should start together. A reactor vessel ordered today is of little value if the point of interconnection is still years from an executed agreement.

Putting the Numbers in Perspective

Recent international benchmarks for new large pressurized-water reactors point to an overnight cost in the range of $9,000 per kilowatt. Applied to a single large unit, that implies a capital cost approaching $10 billion; a two-unit project, with some economy of scale, lands in the high-teens of billions; and a full ten-unit build-out sits in the neighborhood of $90 billion in current dollars.

Against that backdrop, the headline incentive package on the order of $17 to $18 billion across the equipment and construction facilities represents roughly one-fifth of the total expected capital cost of building all ten units. That is a material de-risking of the long-lead procurement, but it is not a turnkey subsidy. The remaining four-fifths must still be financed by project owners who carry the construction, schedule, and market risk.

The figures at a glance

Benchmark overnight cost: ~$9,000/kW for a large PWR.
Single large unit: just under $10 billion.
Two-unit project: roughly $18–19 billion.
Ten-unit program: on the order of $90 billion in current dollars.
Federal incentive package: ~20% of total expected capital cost.
Per-project owner equity: substantial, committed upfront before loan draws.

Reading the Short List Site by Site

Duke the deepest bench

One utility carries six large-reactor licenses across three sites that were never built, including two units acquired through a corporate combination and two more that came off the active list when demand growth slowed and low natural-gas prices undercut the economics. None were pursued to construction which is precisely what makes them attractive now: the licensing and interconnection work exists, but the steel does not, leaving full flexibility on timing.

Turkey Point licensed but stalled on cost recovery

Two licensed units in Florida were suspended indefinitely after a state regulatory decision declined to let the utility recover construction costs during the build and found that the applicant had not demonstrated a realistic and practical intent to break ground in the near term. The license remains a real asset; the obstacle has been regulatory cost-recovery rather than engineering feasibility.

V.C. Summer a partially built site with terminated licenses

The two abandoned units retain enormous latent value as physical infrastructure, but because the licenses were terminated, completion would require fresh licensing. This is the project the rest of the industry is watching.

South Texas Project, Fermi 3, and North Anna 3 right sites, different reactors

These carry licenses for boiling-water or other designs rather than the program’s reactor type. The terminated boiling-water units in Texas faced strong ratepayer opposition from municipal participants two decades ago and are unlikely to revive in their original form; the other two were shelved chiefly over local demand. In each case the site and grid position remain valuable, but a reactor-type change would reopen licensing.

Sites With Withdrawn Applications

A second group of sites once pursued licenses but withdrew their applications, typically for other reactor designs. These represent thinner head starts the regulatory record is incomplete but the underlying sites, several with prior environmental and siting work, can still shorten a greenfield timeline.

Global Context: A Broader New-Build Wave

The federal government has signaled, for the first time in a generation, a willingness to underwrite the rebuilding of America’s large-reactor manufacturing base. A conditional commitment under a new supply-chain loan program would finance the long-lead equipment for ten large-scale light-water reactors organized as five projects of two units each with the explicit goal of shortening deployment timelines by up to three years. The reactor at the center of the program is the only large Generation III+ design currently licensed and operating in the United States.

For developers and utilities weighing participation, the financing is only half the equation. The other half is physical: where can two 1,100-megawatt units actually be built, licensed, and connected to the grid quickly enough to justify the equity? This paper argues that the answer is written largely in the existing license landscape and that the projects with the clearest path forward are those that can reuse a site, a switchyard, and a transmission corridor that the bulk power system was already studied to accommodate.

Key takeaways

Ten units at roughly 1.1 GW each represent on the order of 11 GW of new firm capacity seeking interconnection over the 2030s a scale that will reshape transmission planning in several regions.

Sites with an existing or recently held combined operating license (COL) for the chosen reactor type are a “bird in hand”: the licensing basis, point of interconnection, and transmission studies are partially or fully complete.

The interconnection scope switchyard, generator step-up transformers, line additions, dynamic and electromagnetic-transient modeling, protection coordination, and NERC compliance is a critical-path workstream that should begin in parallel with equipment procurement, not after it.

“Bird in Hand”: Why the Existing License Landscape Decides the Short List

The single most useful question a developer can ask is not “where would we like to build?” but “where is the licensing and interconnection groundwork already done?” A site that already holds or recently held a combined operating license for the chosen reactor type carries an enormous head start. The environmental review, the safety basis, the site characterization, and the point of interconnection have all been examined once already. A site licensed for a different reactor type, by contrast, would have to repeat much of that effort and expense to accommodate a new design.

The table below organizes the U.S. sites whose licensing history makes them logical candidates for new large light-water units. Sites already associated with the program’s reactor type are the strongest matches; sites licensed for other designs are listed because the underlying site, switchyard, and transmission position retain value even when the reactor type would change.

The Interconnection Reality: 11 GW Looking for a Home

Ten units at roughly 1.1 GW apiece is on the order of 11 gigawatts of new firm, dispatchable capacity enough to power millions of households entering the interconnection process over the 2030s. Whether a project succeeds or stalls often comes down to how cleanly that capacity can be absorbed by the surrounding transmission network. This is where the existing-license advantage becomes an engineering advantage, not merely a regulatory one.

A brownfield nuclear site typically comes with a high-voltage switchyard, established transmission corridors, and a transmission system that was once studied (and in many cases physically built) to export large blocks of generation. Reusing that interconnection position can save years of study time and hundreds of millions of dollars in network upgrades relative to a greenfield site that must be added to an interconnection cluster from scratch. The same logic increasingly applies to retiring coal sites and other large generators, where surplus interconnection service and generator-replacement provisions can let new capacity inherit an existing point of interconnection.

Why a brownfield interconnection position is worth so much

An existing point of interconnection and switchyard footprint reduces new right-of-way and permitting risk.

Prior system-impact and facilities studies narrow the unknowns in network-upgrade scope and cost.

Transmission that already moved large generation lowers the odds of deep, schedule-killing network upgrades.

Generator-replacement and surplus-interconnection pathways can preserve queue position and in-service dates.

Substation, Switchyard, and Transmission Scope

Connecting a two-unit station of this size is a major substation and transmission undertaking in its own right. Each unit needs generator step-up transformers sized for its full output, a high-voltage switchyard arranged for reliability (commonly a breaker-and-a-half or ring-bus configuration), and station-service and backup power that meet the stringent reliability requirements of a nuclear facility. Delivering two gigawatts-plus from a single site frequently requires new or reconductored extra-high-voltage lines, expanded terminal stations, and reactive-power resources to hold voltage within limits across a range of system conditions.

These are not afterthoughts to the reactor island; they are long-lead, permit-heavy scopes that belong on the critical path next to the reactor procurement. An owner’s engineer who can carry interconnection, substation, and transmission design alongside the nuclear scope keeps the grid connection from becoming the constraint that idles a finished plant.

Modeling, Studies, and NERC Compliance

Large synchronous reactors bring real strengths to the grid inertia, fault current, and voltage support but they must still earn their interconnection through rigorous study. That work spans steady-state power-flow and short-circuit analysis, dynamic stability simulation in tools such as PSS/E, and electromagnetic-transient studies in tools such as PSCAD for fast phenomena, weak-grid behavior, and transient recovery voltage on switchyard breakers. Protection coordination, relaying, and ride-through performance round out the package.

Layered on top is a thick stack of mandatory reliability obligations. Modeling and data standards, facility-rating and interconnection requirements, protection-system standards, and transmission-planning standards all apply, and each generator interconnection agreement carries its own testing and validation commitments. Getting the studies right the first time is one of the most effective schedule-protection measures available to a new-build project, because re-study and dispute are among the most common causes of interconnection delay.

What This Means for Developers and Owners

The domestic supply-chain push is one front in a wider movement. Several developments underscore how much momentum is building behind both large reactors and small modular reactors, and why supply-chain capacity is now a strategic concern worldwide:

A federal program is separately exploring whether nearly twenty metric tons of surplus plutonium could be converted into advanced reactor fuel, with several developers in active negotiation a potential bridge between waste liability and domestic fuel supply.

A U.S.-based supplier and a European state utility have jointly proposed up to four small modular reactors at a former coal site in the United Kingdom, building on a completed national design-assessment milestone, while the supplier’s first-of-a-kind twin units advance under regulatory review in the United States.

A major engineering firm has filed a notice of intent with the U.S. regulator to begin licensing a long-established heavy-water reactor platform, aiming at large industrial and data-center demand.

Canada has issued its first national nuclear strategy, targeting up to ten new large reactors domestically, expanded uranium output, and export of its reactor technology.

In Poland, a developer has agreed terms with a host city for a small modular reactor station, part of a multi-site program referencing a Canadian build already under construction.

In India, a large private conglomerate has announced plans to enter nuclear generation with an ambition of up to ten gigawatts of capacity by the mid-2030s, following the opening of the sector to private firms.

The common thread is demand data centers, artificial-intelligence compute, advanced manufacturing, and broad electrification colliding with a global shortage of heavy-reactor manufacturing capacity . Whoever can secure long-lead equipment and clean interconnection positions first will set the pace.

How Keentel Engineering Supports New Nuclear Interconnection

Treat the license landscape as a procurement input. The strongest candidate sites are those whose licensing, point of interconnection, and transmission studies are partially or fully complete. Screen the portfolio against this filter before committing equity.
Start interconnection in parallel with equipment orders. Reactor forgings and the generator interconnection agreement are both multi-year items. Sequencing them in series wastes the schedule advantage the financing is designed to create.
Quantify the network-upgrade exposure early. A feasibility-level interconnection screen — power flow, short-circuit, and a first cut at stability — can reveal whether a site is a clean connection or a deep-upgrade problem long before a formal study queue position is taken.
Plan the studies to NERC standards from day one. Modeling, facility ratings, protection, and transmission-planning compliance are not paperwork at the end; they shape the design and the in-service date.
Consider brownfield and generator-replacement pathways. Retiring thermal sites and previously licensed nuclear sites can offer an inherited interconnection position that materially de-risks schedule and cost.

Keentel Engineering is a power-systems and grid-interconnection consultancy built around exactly the workstreams that determine whether a large new generator reaches commercial operation on schedule. For nuclear and other large interconnecting projects, Keentel supports owners and developers across the full interconnection lifecycle:

Point-of-interconnection engineering and interconnection studies — feasibility, system-impact, and facilities-level analysis.

Substation and switchyard design, including generator step-up arrangements and high-voltage bus configurations.

Transmission-line design and network-upgrade scoping.

Dynamic and electromagnetic-transient modeling using PSS/E and PSCAD, including stability, short-circuit, and protection studies.

Owner’s engineer services that integrate the interconnection scope with the balance of plant.
NERC operations-and-planning compliance support across the applicable reliability standards.

Talk to Keentel

If your team is evaluating a site against the new-build financing program or weighing a brownfield interconnection position Keentel can run an early-stage interconnection screen and frame the substation and transmission scope before you commit capital.

Sources & References

This paper is grounded in publicly available program announcements and licensing records, interpreted through Keentel Engineering’s own engineering analysis. Primary references include:

U.S. Department of Energy program announcements describing the supply-chain and construction loan facilities and their structure.

U.S. Nuclear Regulatory Commission licensing records for combined operating licenses, including issued, terminated, and withdrawn applications (basis for Tables 1 and 2).

Publicly reported international cost benchmarks for large pressurized-water reactors.

Public statements and announcements regarding international new-build and small-modular-reactor programs referenced in the global-context section.

All interconnection, substation, transmission, modeling, and NERC-compliance analysis and recommendations represent the independent professional perspective of Keentel Engineering.

CenterPoint Energy Interconnection Requirements for Large Data Centers | Keentel

Sat, 27 Jun 2026 16:51:40 GMT

June 27, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

This specification defines the minimum requirements for primary service at 12,470Y/7,200 V (“12 kV primary service”) and 34,500Y/19,920 V (“35 kV primary service”), both three-phase, four-wire, grounded-neutral. The customer generally furnishes the equipment, and the customer’s first structure on which CNP conductors dead-end is the “Primary Service Pole.”

4.1 Service Voltage & Configuration

Service is provided at 12 kV or 35 kV. For a large load, the higher 35 kV voltage carries less current for the same power and can reduce the number of separate services required. Three-wire utilization equipment is permitted on 12 kV service (bonded to the CNP system neutral) but is not permitted on 35 kV service.

1. Executive Summary

CenterPoint Energy Interconnection

Requirements for Large Data-Center Loads

2. Purpose & Scope

CenterPoint Energy owns and operates the electric delivery system in its service territory and sets the technical conditions under which a customer may interconnect. CNP reserves the right to refuse to energize any service that fails to meet its specifications or its approved project drawings, and any deviation requires the written approval of CNP’s Manager of Power Quality Engineering or a designated representative. Where a specification and an approved drawing conflict, the drawing governs.

CenterPoint’s requirements sit on top of, and incorporate, broader codes and standards. Customer installations must comply with the National Electrical Code (NEC), the National Electrical Safety Code (NESC) / ANSI-IEEE C2, and local authority requirements. Harmonic control is governed by IEEE 519-2014 and Public Utility Commission of Texas (PUCT) Substantive Rule 25.51. Parallel and closed-transition generation is governed by PUCT Substantive Rules 25.211 and 25.212 and CNP’s customer-generation specification .

4.3 Main Protective Device

Every primary-service customer must provide, install, and maintain a main protective device — either a single group of three fuses or an electronically / electromechanically controlled circuit breaker (hydraulic breakers are not allowed). Fuses may be used only when demand or connected transformer capacity is at or below the thresholds in Section 4.2; above 4,320 kVA at 12 kV or 11,950 kVA at 35 kV, a three-phase circuit breaker (pole-mounted recloser, pad-mounted, or indoor) is required. Important rules for breaker installations include:

Automatic reclosing is not permitted.
The customer determines and sets the relay settings, performs functional testing, and provides the settings in an electronic data file plus certified test reports; CNP verifies coordination and loading before energization.
The breaker control must be accessible to CNP at all times and arranged for both CNP and customer locks.
A three-phase, gang-operated bypass switch is recommended; if omitted, maintenance outages must be scheduled with CNP under the Operational Agreement.
Bushing CTs must be on the source side with relay accuracy of C100 or better, and the breaker must have capacitor-trip or battery backup for trip power.

4.4 Disconnecting Device, Overvoltage Protection & BIL

A disconnecting device may be required in addition to the main protective device (for example, when the main device is a pad-mounted breaker). It must be rated for the maximum expected load and fault duty and meet the same BIL minimums as the service. Lightning arresters are recommended on the Primary Service Pole, must be tested to ANSI/IEEE C62.11, and must include a ground-lead isolator; where the main device is pole-mounted, arresters connect on its load side.

4.5 Transformer Connection Rules

Two winding-connection restrictions apply to the customer’s medium-voltage system and must be reflected in the one-line and transformer selections:

Grounded-wye primary to delta secondary transformers shall not be used on the customer’s 12 kV or 35 kV distribution system.
Transformers with ungrounded primary connections (delta or ungrounded wye) shall not be used on the customer’s 35 kV distribution system.

4.6 Capacitor Banks

A customer wishing to install a 12 kV or 35 kV capacitor bank must submit a description of the bank and its control to CNP, and CNP may require load- or power-factor-sensitive switching. Customers with harmonic-producing loads may be required to install harmonic filters in accordance with the harmonic specification . Because capacitors can create resonance with harmonic sources, capacitor application on a data-center system must be coordinated with the harmonic evaluation.

4.7 Overhead vs. Underground Primary Service

Overhead service centers on the customer’s Primary Service Pole, which must be designed for CNP’s adopted extreme-wind loading (110 mph north of US 59 / Hwy 90, 132 mph south of US 59 / Hwy 90) and located a minimum of 50 feet from the CNP metering pole with all-weather truck access.

Underground service is handled by CNP’s Major Underground Engineering group, may be served from outdoor (pad-mounted) or indoor facilities, and is delivered with a project-specific Terms & Conditions package; cable from the metering cabinet or vault must be in concrete-encased conduit (direct burial is not permitted). For a large campus, the design stage typically develops a primary overhead route into the site with major underground utilities on site connecting to the overhead system.

5. Harmonic Distortion Control (Spec 083, Rev. 3)

4.2 Service Capacity & Service Splitting

Two capacity thresholds are decisive for a large load. First, a service whose connected transformer capacity exceeds 4 MVA at 12 kV or 6 MVA at 35 kV may be required by CNP to be split into two or more separate services, and an Operational Agreement is required when a customer is served from multiple services. Second, the choice between fuses and a circuit breaker as the main protective device is set by capacity (see Section 4.3). Because these thresholds are far below a data-center-scale load, multiple services, a circuit-breaker main device, and one or more Operational Agreements should be anticipated.

8. The Article 9.0 Submittal Package

Values are percent of maximum demand load current Iₗ. Limits apply to odd harmonics; even harmonics are limited to 25% of the values shown. TDD = Total Demand Distortion.

5.3 Voltage Distortion Limits

IEEE 519 also defines voltage-distortion limits at the PCC, which filter and mitigation designs must respect:

10. What This Means for a Large Data-Center Project

For a service with a circuit-breaker main device, CNP additionally expects relay and equipment bills of material, breaker / recloser schematics, and the electronic setting file with time-current-curve (TCC) coordination data. All required submittals must be approved before construction begins.

7. Motor Starting Review

This document is a technical reference, not a design or a contract. It summarizes and interprets CenterPoint’s published requirements as they apply to a large data-center load and organizes them for quick reference by the project team. Where specific clauses are cited (for example, “Article 4.8.1”), they refer to the corresponding CenterPoint specification listed in Section 13.

The scope covers: the primary-service configuration and protection rules; the harmonic distortion limits and the customer’s modeling and mitigation obligations; the requirements for emergency / backup generation and secondary load transfer; the motor-start review process; the Article 9.0 submittal package required before energization; and CenterPoint’s design-stage process, including how and when the need for a substation is determined. It does not reproduce the specifications in full, and in any conflict the governing CenterPoint documents and CNP-approved project drawings control.

3. CenterPoint’s Role & the Governing Standards

4. Primary Service Requirements

CenterPoint Energy (CNP) governs how large electrical loads connect to its distribution system through a set of published specifications. For a large data-center development, four of these documents are central: the specification for MV primary service ,the specification for limitation of harmonic distortion , the specification for customer emergency generation and secondary load transfer , and the Motor Start Review process. Together they define how the service is configured and protected, how harmonic distortion is limited, how backup generation is interconnected, and how large motors are reviewed.

This white paper distills those requirements into a single, project-oriented reference for a large data-center development in the CenterPoint Energy service area. It explains each requirement, highlights the thresholds and rules that most affect a large, electronically-fed data-center load, and closes with a detailed FAQ, a glossary, and a register of the governing documents. The intent is to give a project team a clear, shared understanding of what CenterPoint expects and why — so the engineering, schedule, and budget decisions can be made on a sound basis.

Key takeaway:

A data-center load is large enough to trigger CenterPoint’s service-splitting and circuit-breaker thresholds, and it is dominated by non-linear electronic equipment that makes harmonic compliance (IEEE 519-2014) a real design driver rather than a formality. These two facts shape nearly every requirement discussed below.

6.1 Emergency / Backup Generation

Customer-owned emergency generation must connect to the customer’s load through either an open-transition transfer switch or a key-interlocked two-breaker / switch arrangement that prevents any electrical or physical connection between the CNP system and the customer’s generator bus. Standby generators must sit on the load side of the meter, behind the main device; no transfer switch is allowed on the high (CNP) side of the meter. The generator and its exhaust must be located where they do not interfere with CNP personnel or equipment, and the location must be approved by CNP.

Operating a generator in parallel with the CNP system (closed transition) is an exception that is granted only when the customer installs protective devices in accordance with PUCT Substantive Rule 25.212 and CNP’s customer-generation specification (007-231-76) at the interconnection point, and when generator voltage and current distortion comply with IEEE 519. Customers served from a CNP secondary network may not parallel generators with CNP at all. Recall also that, under the harmonic specification, generation equipment is held to the most stringent harmonic-distortion category.

6.2 Secondary Load Transfer

Where a customer’s split load is served from two CNP feeds, the buses from each feed must be electrically isolated from one another, and load transfers must be performed manually using open-transition transfer switches or breakers — automatic transfer by the customer is not permitted without CNP Engineering approval. Before transferring, the customer must shed an equal or greater amount of pre-approved load on the receiving feed so the normal load level is never exceeded. Secondary load transfer is offered only under separate contractual arrangements, requires submission of a load-transfer schedule, and is documented in an Operational Agreement.

Any motor rated 250 HP or larger must be reviewed by CNP’s power-quality engineer using the Motor Start Review Form. The form captures service information (service name and address, circuit and section, transformer GLN, service voltage, transformer type), present and future kVA demand, and — for each motor or motor group — a data table including the number of motors, horsepower, phase, voltage, full-load amps, locked-rotor amps, code letter (locked-rotor kVA per HP), required starting frequency, whether the motor starts under load, and the starting method. Recognized starting methods are full-voltage / across-the-line (FVS), auto-transformer reduced-voltage (ATRF), capacitor start (CAPS), star-start delta-run (SSDR), and other (OTHR). Where motors start simultaneously, a description of the scheme must be attached. For a data center , large chillers and pumps are the most likely candidates to trigger this review.

Before any construction to serve the customer may begin, CenterPoint requires a submittal package to be reviewed and approved. CNP typically identifies the following items (per Article 9.0 and project-specific guidance):

CenterPoint has advised that the design stage is approximately a three-to-six-month process. During this stage, CNP develops the primary overhead design into the site and the major underground utilities on site that connect to the overhead system. Critically, if a substation is required to serve the load, that requirement is identified during this stage. In parallel, CNP’s power-quality engineer reviews and approves the customer’s major equipment and prepares the Operational Agreement. Only after these steps are complete does CNP finalize its construction costs; payment then allows construction to commence.

Because a data-center-scale load sits well above the single-service thresholds, the central engineering question of the design stage is whether the load can be served from existing infrastructure through multiple primary services or whether dedicated substation infrastructure is the controlling path. Producing a defensible answer early — grounded in the load analysis, service-voltage selection, and CNP’s feedback — is one of the most valuable outcomes of this phase.

Operational Agreement triggers. An Operational Agreement is required when the customer uses a circuit breaker as the main protective d evice, does not install a bypass switch on the Primary Service Pole, operates generation in parallel with CNP, or receives service from two CNP feeds with the ability to transfer load. A large data-center service is likely to trigger more than one of these.

Bringing the requirements together, several implications follow directly for a large data-center load in the CenterPoint service area:

Service configuration: at roughly an order of magnitude above the single-service splitting thresholds, the facility should anticipate multiple primary services and/or dedicated substation infrastructure; 35 kV service merits evaluation to reduce current and service count.
Protection: a circuit-breaker main device (not fuses) will be required, with customer-set, CNP-verified relay coordination, no automatic reclosing, and one or more Operational Agreements.
Harmonics: the UPS / rectifier / VFD load makes IEEE 519 compliance a real design driver; a harmonic model and, likely, a mitigation strategy will be needed, and any capacitors must be coordinated to avoid resonance.
Transformers: winding connections must respect the prohibitions (no grounded-wye / delta on the customer system; no ungrounded primary at 35 kV).
Backup generation: open-transition or key-interlocked connection on the load side of the meter, with the stricter generation harmonic category, and additional protection if parallel operation is ever desired.
Large motors: chillers or pumps rated 250 HP and larger will trigger the Motor Start Review.
Process: the substation determination, equipment approval, and Operational Agreement are CNP-controlled and occur over the three-to-six-month design stage before final costs are set.

Large data-center programs are frequently advanced in phases — for example, an initial build served at the distribution-primary level through CenterPoint’s design process described here, and a larger build that crosses into transmission-level service and the ERCOT large-load study process. The requirements in this paper apply principally to the distribution-level work; a transmission-level interconnection introduces additional ERCOT modeling deliverables (load data, dynamic, and PSCAD / EMT models) that are addressed separately.

This specification places responsibility for limiting harmonic flow into the CNP system on the customer and reflects IEEE 519-2014 and PUCT Rule 25.51. CNP may refuse to energize a non-compliant service and may measure the customer’s harmonic injection at the point of common coupling (PCC) at any time. Measurement instruments follow IEC 61000-4-7 and IEC 61000-4-30.

5.1 Why It Matters for a Data Center

A data center is dominated by non-linear electronic loads — UPS rectifiers, switch-mode power supplies, and variable-frequency-drive cooling — that draw current in a non-sinusoidal manner. CenterPoint notes that electronic power converters typically create current distortion on the order of 20–30% of rating for three-phase equipment. Left unmanaged, that distortion can exceed CNP’s limits, disrupt other customers, and prevent energization. Harmonic compliance is therefore a substantive design task, not a check-the-box item.

5.2 Current Distortion Limits (Iₛₜ/Iₗ and TDD)

Allowable harmonic current injection is expressed as a percentage of the maximum demand load current (Iₗ), and the allowance grows as the short-circuit-to-load ratio (Iₛₜ/Iₗ) at the PCC grows. The table below reproduces CNP’s current distortion limits (from IEEE 519-2014, Table 2). Even-order harmonics are limited to 25% of the odd-order limits, and — importantly for backup generation — all power-generation equipment is held to the most stringent (“< 20”) row regardless of the actual Iₛₜ/Iₗ. Distortions that produce a DC offset (e.g., half-wave converters) are not allowed.

11. Frequently Asked Questions

5.4 Modeling, Measurement & Mitigation

6. Emergency Generation & Secondary Load Transfer (Spec 082, Rev. 1)

CenterPoint requires the customer to model its system to project harmonic levels and to determine whether limits will be exceeded. Supporting data the customer must provide includes a one-line showing ratings and connections, capacitor data, conductor sizes and impedances, the location and type of non-linear loads, the overall plant load and the non-linear portion, and transformer ratings, connections, and impedances. On written request, CNP supplies system data such as available fault duty, system impedance, and transformer ratings. If limits would be exceeded, the customer is responsible for designing, installing, and maintaining mitigation — current-limiting reactors, passive traps and filters, or active conditioners — subject to approval by CNP’s Manager of Power Quality. Mitigation should be designed to operate independently of the CNP supply characteristics, because CNP reconfigures its system over time.

9. The Design-Stage Process & the Substation Question

PSS®/E Python Automation Guide | Keentel

Fri, 26 Jun 2026 16:16:47 GMT

JJune 26, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

There are two distinct modes. The first is in-GUI recording: you press record, perform actions in the interface, stop, and PSS/E writes the equivalent Python (or response-file) script, which you can replay. It is the fastest way to discover the right API call for a task you already know how to do by hand — but it caps your capability at whatever the GUI exposes.

The second mode is external control : you launch Python first, import the PSS/E libraries, and initialize the engine headlessly. Now PSS/E runs underneath your script. You can iterate over thousands of elements, branch on results, and write outputs anywhere — the full power of the language is available. The rest of this guide focuses on this mode, with GUI recording used as a learning aid to find the calls you need.

Version-to-Python mapping (know this before you start)

PSS/E ships with a bundled Python interpreter, and the version pairing matters — calling the wrong interpreter is the single most common reason an import fails. The current landscape:

Siemens PTI’s PSS®/E is the workhorse of bulk-system planning and interconnection studies, and its graphical interface is excellent for interactive, one-off analysis. But the moment a study involves hundreds of buses, dozens of contingencies, or a base case that must be solved, screened, and reported over and over, point-and-click stops scaling. The answer is automation — driving PSS/E from Python so the software becomes a callable engine inside a repeatable, auditable workflow.

This guide walks through the automation path our team relies on every day: how to reach PSS/E from outside its own GUI, how to open and solve a case programmatically, how to extract results the software will not record for you, and how to scale single-value queries into bulk data pulls. It closes with two anonymized case studies showing where this discipline pays off in real interconnection work.

Automating PSS/E with Python — A Practical Engineering Workflow

WHY AUTOMATE AT ALL?

Recording macros inside the GUI is a fine first step, but it is limited — you can replay a fixed sequence and little more. Accessing PSS/E externally through Python unlocks the full language: loops, lists, conditionals, file I/O, data structures, and integration with the rest of your toolchain. That is the difference between a recorded button-press and a genuine engineering pipeline .

1. Two Ways to Use Python with PSS/E

2. Reaching PSS/E from Outside the GUI

Start from a command prompt and launch the matching interpreter (for example, py -2 for the Python 2.7 that ships with v33/34, or py -3 for the Python 3.9 in v35). The classic, explicit setup tells Python where PSS/E lives, registers that location on both the interpreter and the OS search paths, then imports the core API module and initializes the engine:

The explicit method — full control, but several steps to remember.

import os, sys

# Point at the PSSBIN folder of your installation

psse_path = r"C:\Program Files\PTI\PSSE33\PSSBIN"

# Register PSSBIN on the interpreter and the OS environment

sys.path.append(psse_path)

os.environ['PATH'] = psse_path + ";" + os.environ['PATH']

# Import the core API and initialize PSS/E headlessly

import psspy

psspy.psseinit()

Those path steps are easy to forget. A cleaner approach uses a small helper module that performs the path wiring for you. Install it once with pip, then a single call selects the version and prepares the environment:

The convenience method — the helper handles sys.path, PATH, and environment wiring.

# One-time install from the command prompt:

# pip install psse_path

from psse_path import add_psse_path

add_psse_path(34) # selects v34; omit the argument if only one version is installed

import psspy

psspy.psseinit()

KEENTEL PRACTICE NOTE

Frequently Asked Questions

We wrap initialization in a single, version-aware startup routine kept under version control. Every study script imports the same routine, so the environment is identical on every engineer’s machine and on the build server. Reproducibility starts at psseinit().

Pair the interpreter to the PSS/E version: launch Python 2 for v33/34, Python 3 for v35.

For v35, the workflow is simpler still: from a Python 3 prompt, importing the version package wires the environment automatically — no manual path handling or extra install required — after which you import psspy and call psseinit() as usual.

3. Opening and Solving a Case Programmatically

With the engine live, the OS library handles navigation to your working directory and the case API loads the saved case (.sav). From there you can solve it exactly as the GUI would:

Open a case folder, load the .sav, and solve — entirely from script.

import os

work_dir = r"C:\Program Files\PTI\PSSE33\EXAMPLE"

os.chdir(work_dir) # move to the case folder

print(os.getcwd()) # confirm current working directory

savfile = "savnw.sav"

psspy.case(savfile) # open the saved case

# Solve with fixed-slope decoupled Newton-Raphson (FDNS), default options

psspy.fdns([0, 0, 0, 1, 1, 0, 0, 0])

How do you discover the exact solve call and its option flags? Record it. Perform the power-flow solution once in the GUI with the record function active, stop, and PSS/E writes the precise API line — including the default option vector — which you paste straight into your script. Recording is the fastest route from “I know how to do this by hand” to “I have the API call.”

4. Extracting Results: Activities Are Recorded, Results Are No

This is the concept that trips up newcomers. In PSS/E, activities can be recorded but results cannot. An activity is an action — solving a power flow, building a subsystem, running a contingency. The voltage profile, branch flows, and mismatches that come out of that action are results , and there is no record button for them. To get results into Python for further analysis, you query them deliberately through the data-retrieval APIs documented in Chapters 7–9 of the API reference (api.pdf, in the installation’s doc folder).

System-level and single-element queries

Most retrieval APIs return a tuple whose first element is an error code (0 means the call succeeded) followed by the value. Complex returns carry real and imaginary parts — for power, real is MW and imaginary is MVAr. A representative set:

Each call returns (error_code, value); change the keyword to fetch a different quantity.

ierr = psspy.solved() # 0 = met convergence tolerance

ierr, base_mva = psspy.sysmva() # system base MVA (e.g., 100.0)

ierr, mismatch = psspy.sysmsm() # largest bus mismatch (MVA)

# System totals take an argument that selects the quantity

ierr, demand = psspy.systot('LOAD') # complex: MW + jMVAr

demand_mw = demand.real

demand_mvar = demand.imag

# Single-element retrieval

ierr, vbase = psspy.busdat(101, 'BASE') # nominal kV

ierr, vkv = psspy.busdat(101, 'KV') # solved kV

ierr, vpu = psspy.busdat(101, 'PU') # per-unit voltage

ierr, charg = psspy.brndat(151, 152, '1', 'CHARG') # line charging

ierr, agen = psspy.ardat(1, 'GEN') # area 1 generation (complex)

The pattern is consistent across element types — busdat for buses, brndat for branches, ardat for areas, and their relatives for zones, machines, plants, and loads. The API call stays the same; the keyword (PU, KV, CHARG, RATEA, GEN, and so on) selects what you get back. Learn the pattern once and the whole library opens up.

5. Scaling Up: Lists and Iteration for Bulk Retrieval

Querying one bus at a time is fine for a spot check, but interconnection studies need results for hundreds of elements. Python lists — ordered, iterable, mutable collections — turn a single-value query into a bulk operation. Define the elements of interest, loop over them, and collect the results:

A for-loop plus a results list scales one query into many; zip() pairs inputs with outputs.

# Buses of interest (could be the entire bus list from the case)

buses = [101, 151, 201, 301, 3004, 3008]

bus_pu = [] # start with an empty list

for bus in buses: # iterate element by element

ierr, vpu = psspy.busdat(bus, 'PU')

bus_pu.append(vpu) # grow the results list

# Pair each bus with its voltage for clean reporting

bus_data = zip(buses, bus_pu)

for record in bus_data:

print(record[0], record[1]) # index 0 = bus, index 1 = pu voltage

Two details matter in practice. First, Python is zero-indexed — the first element of a record is at index 0, not 1. Second, indentation is syntax, not decoration: the body of a loop must be indented consistently or the interpreter will reject it. With those in hand, the same template extends naturally to branches, transformers, and contingency sets — the foundation of any automated screening routine.

6. Response Files, the CLI, and Progress Output

Python is the most powerful automation surface, but PSS/E offers two complementary mechanisms worth knowing.

Response files (.idv) and recorded scripts

When you record an activity, you can save it either as a Python script or as a response file (.idv) — a batch format that replays a fixed sequence of activities. Recorded option flags appear as 0s and 1s (again reflecting zero-based indexing). Response files are ideal for deterministic, repeatable solve-and-report sequences; Python is the choice when you need logic, loops, and data handling around those steps.

The Command Line Interface (CLI)

PSS/E’s CLI, reachable from the command-line input section of the output bar, lets you load cases and run reports without touching the GUI — for example, opening a case with the CASE activity (entry is case-sensitive), listing data by category, locating elements with FIND, or screening voltages with a limit check (VCHECK) that flags buses outside a defined band (with Vmax greater than Vmin). The full CLI command set spans roughly twenty documented chapters and is a fast path for interactive, no-GUI investigation.

Capturing the progress report (PDEV)

The iterative solution log — the progress output — is more than a status readout; it is a diagnostic record. Through Input/Output Control, the Direct Progress Output (PDEV) option redirects that log from the terminal to a file. This is invaluable when reconciling a case migrated between platforms (for example PSLF, PSCAD, or DIgSILENT into PSS/E): capture the progress output before and after each adjustment, then compare logs to see exactly which changes drove convergence. You can append successive runs to one file and set the lines-per-page for clean review.

THE ENGINEERING PAYOFF

Automation is not about replacing engineering judgment — it is about removing the manual, error-prone steps that stand between a solved case and a defensible result. Scripted retrieval is repeatable, auditable, and fast, which is exactly what regulators, interconnection customers, and internal QA all expect of a modern study workflow.

Case Studies

The following case studies are drawn from Keentel Engineering project experience. In keeping with our confidentiality practice, all client names, project names, locations, and proprietary identifiers have been removed; figures are representative and rounded to illustrate the engineering approach rather than any specific engagement.

CASE STUDY 01

Automated Voltage-Compliance Screening Across a Multi-Point Interconnection Portfolio

Challenge.

A developer client advanced a portfolio of utility-scale generation and storage project sharing a congested study area, requiring steady-state voltage screening across more than 1,500 monitored buses under a large matrix of dispatch and contingency conditions. The original workflow relied on manual, bus-by-bus inspection in the PSS/E GUI after each solved scenario. The effort was slow, difficult to reproduce between analysts, and prone to transcription error when results were copied into spreadsheets for the compliance narrative.

Approach.

Keentel replaced the manual review with a Python layer driving PSS/E externally. A version-aware startup routine initialized the engine identically on every run; the monitored-bus set was loaded once into a Python list; and a screening function iterated the list with busdat, retrieving per-unit voltage for each bus after every solved scenario. Every API return was error-code-checked before use, and results were paired with bus identifiers using zip() and written directly to a structured output file — eliminating manual copying entirely. Buses outside the applicable voltage band were flagged automatically, mirroring a CLI-style limit check but applied programmatically across the full scenario matrix.

Result.

Per-scenario review time fell from hours of manual inspection to a script run measured in minutes, and the screening became fully reproducible: any analyst could regenerate identical results from the same inputs. The structured outputs fed directly into the compliance documentation, and the error-code checks surfaced two silently failed retrievals in an early batch that manual review would likely have missed — reinforcing the audit value of scripted retrieval in an interconnection-study context.

Engineering takeaway.

List-driven iteration over single-element APIs turns a tedious manual task into a repeatable screening tool. The discipline that makes it defensible — standardized initialization, error-code checking, and direct-to-file output — is what distinguishes production automation from a one-off macro.

CASE STUDY 02

Cross-Platform Case Reconciliation During a Power-Flow Model Migration

Challenge.

A client needed a large base case migrated from a different power-system platform into PSS/E for an interconnection study , then validated to confirm the migrated model behaved consistently with the source. After import, the case did not converge cleanly, and the team needed a structured way to identify which adjustments restored convergence — without resorting to undocumented trial-and-error that would be impossible to defend later.

Approach.

Keentel combined PSS/E’s Direct Progress Output (PDEV) with scripted data extraction. Before and after each modeling adjustment, the iterative progress log was captured to file via PDEV, building a documented trail of how each change affected the solution path and tap-stepping behavior. In parallel, a Python routine extracted system totals (systot for load and generation) and key bus voltages (busdat) from both the source-platform export and the PSS/E case, then compared them element by element to quantify discrepancies and confirm alignment as the model was tuned. Successive progress runs were appended to a single log for side-by-side comparison.

Result.

The captured progress logs isolated the specific adjustments that drove convergence, and the scripted comparison confirmed that migrated system totals and voltage results matched the source model within an acceptable tolerance once tuning was complete. The result was a defensible reconciliation package — progress logs plus a quantitative discrepancy report — that documented not just that the migrated case converged, but exactly why, and that it represented the same system as the original.

Engineering takeaway.

Progress output is a diagnostic asset, not just a status readout. Pairing PDEV capture with scripted, element-by-element comparison converts an opaque convergence struggle into a documented, reviewable migration — the kind of traceability that holds up under study review.

Disclaimer & Notice

This document is provided by Keentel Engineering for general informational and educational purposes only. Code patterns are illustrative and simplified to convey workflow; API signatures, option vectors, and behavior vary by PSS/E version, and you should consult the API and CLI references shipped with your installation for authoritative usage. Examples reference the standard sample case distributed with PSS/E.

The case studies are anonymized composites of project experience; all client, project, and location identifiers have been removed and figures are representative. Nothing herein constitutes engineering, legal, or procurement advice. PSS®/E is a registered trademark of Siemens; PSLF, PSCAD, DIgSILENT, Python, and other product names are the property of their respective owners. Their use here is solely for factual technical context and does not imply any affiliation with, endorsement by, or sponsorship of Keentel Engineering.

Switching Overvoltage Studies in PSCAD/EMTDC Guide

Tue, 23 Jun 2026 11:58:44 GMT

jun 23, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

A switching study is only as trustworthy as the model behind it. The guiding principle is that the area of interest must be represented faithfully enough to capture the real travelling-wave behaviour, with the rest of the network reduced to equivalents that do not distort it.

How Far Out to Model

As a working rule, model the system in full detail to at least two buses away from the point of study in every direction. Beyond that boundary, the rest of the network is collapsed into equivalent sources. Buses carrying generation can be represented as equivalent sources, and the boundary buses themselves are terminated with equivalent sources so the reduced network still presents the correct impedance and short-circuit behaviour to the study area.

Each network-boundary equivalent is defined by its bus voltage magnitude and angle, the real and reactive power at that bus, and its positive- and zero-sequence impedances — the same quantities you would draw from a load-flow and short-circuit case.

Choosing the Right Transmission-Line Model

PSCAD/EMTDC offers three line representations, and the choice is decisive for a switching study because switching overvoltages are a travelling-wave phenomenon.

Modeling, Statistical Point-on-Wave Analysis, and Surge-Arrester Energy Assessment for Slow-Front Transients.

An in-depth technical explainer, FAQ, and three field-style case studies.

Insulation Coordination: Switching

Overvoltage Studies in

PSCAD/EMTDC

Why Switching Studies Matter

Every time a breaker or switch operates on a transmission system, it does more than connect or isolate a line. The switching action launches travelling waves down the line, and those waves excite the inductance and capacitance of the network into oscillation. The result is a high-frequency voltage transient a switching overvoltage that can momentarily push system voltage well above its nominal value. On long extra-high-voltage lines these transients are among the most demanding events the insulation will ever see.

A switching study answers four practical questions that sit at the heart of insulation coordination: what overvoltage levels the network actually produces during switching; what insulation withstand level the equipment therefore needs; whether the surge arresters can absorb the energy these events deposit; and whether any network resonance will amplify the problem. Get these wrong and you either under-insulate — risking flashover and equipment damage — or over-insulate and overspend. Insulation coordination is the discipline of getting that balance right, and the switching study is where the numbers come from.

WHERE SWITCHING OVERVOLTAGES SIT

Switching transients are slow-front transients international standards (IEC) place them in a frequency band reaching up to a few kilohertz, between the power-frequency temporary overvoltages below them and the fast-front (lightning) and very-fast transients above them. The study method here targets that slow-front band.

Building a Credible Model

The pi-section model captures only the fundamental and represents the line as a lumped RLC circuit, so it cannot reproduce the travelling waves that create switching overvoltages. The Bergeron model does carry travelling waves but is accurate only at the single frequency it is tuned to. The frequency-dependent (phase) model is valid across the entire spectrum from DC into the megahertz range and reproduces the travelling-wave behaviour faithfully — which makes it the correct default for switching work.

Line Geometry and Transposition

A frequency-dependent model is only as good as the physical data behind it. Model the actual tower geometry: conductor positions and arrangement, bundle configuration, ground clearance, and line sag. Enter conductor data accurately — conductor type, radius, DC resistance, bundle details, sag — and include the ground (shield) wires. The tower geometry must match the specific line under study, not a generic template.

For a long transposed line, represent the transposition explicitly. Rather than modelling a 200 km line as a single uniform span, divide it into segments — for example four 50 km segments — and transpose them externally between segments so the phase rotation matches the real installation. An untransposed line of the same length can be modelled as a single segment of the correct length.

Transformers, Shunt Devices, Generators

Transformers — use the general rating and impedance data, and include saturation data so the nonlinear core behaviour is represented. Both are typically available from manufacturer datasheets.

Shunt devices — line reactors and shunt capacitors near the study line are modelled as equivalent inductance or capacitance, computed from the bus voltage level and the device MVAr rating. These matter enormously for switching, as shown later.

Generators — define the bus voltage and angle, the real and reactive power at the machine, and the positive-sequence (subtransient, X″d) impedance; include the zero-sequence impedance if available, otherwise accept the default.

Surge Arresters: V-I Curve and Energy

The surge arrester is the protective heart of the study, so its model needs three things: the arrester rating, its nonlinear voltage-current (V-I) characteristic, and its energy-absorption capability. When the manufacturer datasheet does not provide a complete V-I curve, the lower portion of the characteristic must be tuned so that it corresponds to very small currents — the high-resistance leakage region — while the upper portion follows the published data. If full factory test results for the V-I curve are available, use them directly for the most accurate result.

Cables

Cables follow the same logic as overhead lines: use a frequency-dependent model built from the actual cable design data. Current PSCAD/EMTDC versions add cable representations that capture real installations more faithfully — for instance pipe-type cables and three-single-core arrangements in flat or trefoil formation — which is important because cable geometry strongly influences the travelling-wave response

Validate Before You Simulate

Here is the crux of switching analysis: the overvoltage magnitude depends primarily on where, on the voltage waveform, the breaker poles actually make contact and the arc establishes. That instant — the point on wave — varies statistically from one operation to the next and differs pole to pole. A single deterministic run tells you almost nothing about the worst case.

Point-on-Wave: Deterministic and Statistical

Two complementary techniques capture this. In the deterministic sweep, the nominal breaker closing time is advanced by one one-hundredth of a power-frequency cycle on each successive run, stepping through the whole cycle (T₁, T₂, … over hundredths of a 60 Hz period) and recording the peak overvoltage at each instant. In the statistical approach, each scenario is run 100 times or more, with the pole closing times drawn randomly from a statistical breaker model that represents pole scatter and pre-strike. The breaker opening and closing times are chosen to maximise trapped charge, which drives the worst-case overvoltage. PSCAD/EMTDC's Multiple Run component automates this sweep; a second component can be added when more signals need to be recorded than one component allows.

The statistical pole parameters — the mean of the minimum and maximum time delays, the standard deviation, and the per-pole random scatter — are set in the configuration so the simulated breaker behaves like a real one. Because switching overvoltages are a kilohertz phenomenon, each run need only be one to two seconds of simulated time to capture them fully, and a simulation time step on the order of 10 microseconds is generally sufficient for slow-front work (shorter lines or cables may require a smaller step).

Trapped Charge and the Line-Reactor Effect

When a line with no reactor or shunt device is de-energised, charge remains trapped on the line. Reclosing soon after opening, onto that trapped charge, produces the highest overvoltage — so the worst case is a quick reclose. But when a shunt (line) reactor is in service, the picture changes: the reactor and the line capacitance form a resonant pair, and after opening, the trapped charge oscillates at a dominant natural frequency that is considerably slower than the system frequency. To find the worst case in that situation, the reclose must be timed against that slow natural oscillation, not the power-frequency cycle — closing at the instant the oscillating line voltage opposes the source gives the maximum overvoltage.

Scenario Matrix

A real study sweeps network states, not just switching instants. For a double-circuit line between two stations with line reactors, the scenarios multiply quickly: reactors in or out of service; one circuit in or out; energising from the first station or the second with the far end open; and neighbouring lines into each station in or out of service. Contingency conditions such as N-1 and N-2 are included to capture the network's worst case. Each combination is then subjected to the point-on-wave sweep.

Reading the Results

Why This Matters for Interconnection Engineering

At Keentel, we treat grid interconnection as a first-order design input, and switching studies are exactly where that philosophy earns its keep. New EHV lines, cable additions, reactive-compensation schemes, and generator or storage interconnections all change the network's travelling-wave and resonance behaviour. An interconnection that looks clean in load flow can still impose switching overvoltages that drive arrester ratings, breaker pre-insertion resistor decisions, or even the equipment basic insulation level at the point of interconnection.

Doing the insulation-coordination study early with a validated frequency-dependent model and a proper statistical point-on-wave sweep turns those risks into specified, defensible numbers before equipment is procured. That is the difference between discovering a resonance problem on paper and discovering it during commissioning.

Case Studies

The following three scenarios are anonymized and generalized to illustrate the method in practice. They contain no client, project, or location identifiers and are presented for educational purposes.

Case Study 1 EHV Double-Circuit Line Energization and Arrester Adequacy

Focus:Statistical point-on-wave energization; surge-arrester energy duty

Subject:A long (~200 km) EHV double-circuit overhead line with arresters at each line end

Challenge

A long extra-high-voltage double-circuit line had to be energised under conditions that would produce the highest credible switching overvoltage, and the team needed to confirm that the line-end surge arresters could withstand the resulting energy without exceeding their capability — and that the overvoltage stayed below the equipment's switching impulse withstand level.

Approach

The line was represented with a frequency-dependent model built from detailed tower geometry, conductor and bundle data, and sag, with the surrounding network modelled two buses out and boundary buses reduced to equivalent sources. Arresters were modelled with their full nonlinear V-I curve and energy capability. A Multiple Run point-on-wave study stepped the breaker closing instant across a full power-frequency cycle in hundredth-of-a-cycle increments and repeated the operation 100 times, with opening and closing timed to maximise trapped charge. Voltages at the switching and open ends and the per-arrester energy were recorded for every run.

Outcome

This explainer synthesizes widely accepted practice in insulation coordination and electromagnetic transient simulation. Readers seeking primary technical detail may consult:

IEC 60071 series — Insulation co-ordination (definitions, principles, and application guidance), including the classification of slow-front (switching) overvoltages.
IEEE standards and application guides on insulation coordination, surge arresters (metal-oxide arrester application), and transmission-line switching.
Published literature on frequency-dependent transmission-line and cable modeling and travelling-wave representation for transient studies.
Manufacturer datasheets and factory test data for surge-arrester V-I characteristics and energy-absorption capability, and for transformer saturation data.
Vendor documentation for the PSCAD/EMTDC simulation environment regarding line/cable models, the statistical breaker model, and the Multiple Run component.

About Keentel Engineering

The Statistical Switching Study

The study reports voltages at the switching end and the open (receiving) end of the line, the induced voltage on any out-of-service parallel circuit, and the energy absorbed by each surge arrester. Two patterns are typical: the open end sees the highest overvoltage (the travelling wave reflects and roughly doubles there), and the arrester at the open end therefore absorbs the most energy. An out-of-service parallel circuit usually shows only a small induced voltage on the order of a tenth of a per-unit and its arresters absorb almost nothing.

Results are presented statistically. For each monitored quantity the output file lists the minimum, maximum, mean, and standard deviation, along with the 2% and 98% statistical levels. Those levels are defined relative to the mean and standard deviation, and the 98% level is the value used for design margin against the equipment withstand:

Frequently Asked Questions

Against an arrester capability on the order of 13 kJ/kV-MCOV, the worst-case open-end duty of roughly 1.4 kJ/kV-MCOV left a wide margin, and the ~2.2 per-unit overvoltage sat well below the equipment switching impulse withstand level. The energization scheme was confirmed with no insulation-coordination violation, and the out-of-service parallel circuit showed only a small induced voltage with negligible arrester energy.

TAKEAWAY

A frequency-dependent model plus a statistical point-on-wave sweep converts “will the arresters cope?” into a defensible margin — here, worst-case energy roughly a tenth of the arrester's rated capability.

Case Study 2 Shunt-Reactor-Compensated Line: Resonance-Aware Reclose Timing

Focus:Trapped-charge reclosing with a line reactor; natural-frequency timing

Subject:A reactor-compensated double-circuit line switched from either terminal

Challenge

On a line equipped with shunt (line) reactors, a conventional quick-reclose assumption did not produce the true worst case. After opening, the reactor and line capacitance resonated, leaving the trapped charge oscillating at a dominant natural frequency well below system frequency. Reclosing on the power-frequency assumption alone risked under-predicting the overvoltage.

Approach

The reactors were modelled as equivalent inductances derived from their MVAr rating and bus voltage. The study first identified the dominant natural frequency of the reactor-line combination, then timed the reclose against that slow oscillation so the closing instant coincided with the line voltage opposing the source. A scenario matrix was swept — reactors in and out of service, each circuit in and out, and energization from either terminal with the far end open — and each combination was run through the statistical point-on-wave sweep.

Outcome

Timing the reclose to the reactor-line natural frequency, rather than the power-frequency cycle, revealed the genuine worst-case overvoltage that a naive quick-reclose study would have missed.
The scenario matrix isolated which network state (reactor status and circuit configuration) governed the design, so attention could be focused there.
With the worst case correctly identified, arrester energy was confirmed within capability and the overvoltage within the equipment withstand level.

TAKEAWAY

When a line reactor is present, the worst-case reclose is set by the reactor-line natural frequency, not the system frequency. Find that frequency first, then time the switching against it.

Case Study 3 Validation-First Workflow for a Mixed Overhead/Cable Circuit

Focus:Model validation against load-flow and field data before the SOV study

Subject:A tie circuit combining overhead line and underground cable sections

Challenge

A tie circuit combining overhead-line and underground-cable sections needed a switching study, but the consequences of an inaccurate model — wrong arrester ratings or an unnecessarily high insulation level — were costly. The team needed confidence that the EMT model genuinely represented the as-built network before trusting its overvoltage predictions.

Approach

Both the overhead and cable sections were built with frequency-dependent models — the cable from its actual design data using the appropriate single-core (flat or trefoil) representation — with the network modelled two buses out and transformers including saturation. Before any switching scenario, the model's load-flow solution and fault levels were compared against an independent load-flow program and against available field measurements. Only after the steady-state and fault behaviour matched within tolerance was the statistical point-on-wave study run.

Outcome

Power-flow and fault levels agreed with both the independent load-flow tool and field measurements within an acceptable range, establishing the model's credibility.
The validated model produced switching-overvoltage and arrester-energy results that could be defended to reviewers and used directly for equipment specification.
Treating the cable sections with proper frequency-dependent models avoided the travelling-wave errors a lumped representation would have introduced at the overhead-to-cable transitions.

TAKEAWAY

Validation is not a formality. Matching known load-flow and fault levels against an independent tool and field data is what lets the unknown switching overvoltages be trusted for design.

Keentel Engineering is a power systems and grid interconnection consulting firm with offices in Tampa, Florida and Austin, Texas. Our work spans EMT modeling, point-of-interconnection (POI) interconnection engineering, substation and transmission-line design, utility-scale renewables and battery energy storage (BESS) engineering, power system studies, and NERC operations and planning compliance. We approach every engagement interconnection-first — treating grid interconnection as a first-order design input rather than a downstream utility formality.

Author: Sandip R. Patel, P.E. —

Founder & Principal Engineer, IEEE Senior Member. Contact Keentel Engineering to discuss accelerating your EMT, interconnection, or parametric study workload.

DISCLAIMER & TRADEMARKS

This document is an independent educational and technical commentary produced by Keentel Engineering. PSCAD® and EMTDC™ are trademarks of their respective owner; InfiniBand® is a trademark of its respective owner. Keentel Engineering is not affiliated with, endorsed by, or sponsored by any such trademark holder.

Any case studies are anonymized, generalized, and presented for educational purposes; they do not identify any specific client, project, or location. Performance figures are illustrative of the methods described and will vary with hardware, network model, and configuration. © Keentel Engineering. All rights reserved.

Sources & References

Before any switching scenario is run, the model must be validated — otherwise the overvoltages it predicts are meaningless. Validation is done at the steady-state and fault level the model should already know how to reproduce: check the load-flow solution and the fault (short-circuit) levels against an independent source. That source can be a separate load-flow program such as PSS/E, or actual field measurements. When the power-flow and fault levels fall within an acceptable tolerance of the reference, the model has earned the right to be used for transient work.

VALIDATION IN ONE LINE

If the model cannot reproduce the known load-flow and fault levels, it will not be trusted to predict the unknown switching overvoltages. Validate against an independent load-flow tool or field data first.

The overvoltage at the open end is compared against the equipment's switching impulse withstand level (its insulation level), and each arrester's absorbed energy is compared against its capability conveniently normalised as kilojoules per kV of MCOV (maximum continuous operating voltage). If the predicted overvoltage sits comfortably below the withstand level and the arrester energy below its rated capability, the scenario passes with no insulation-coordination violation.

Concurrent EMTDC: Parallel EMT Simulation Explained

Tue, 23 Jun 2026 10:26:22 GMT

jun 23, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

There are two fundamentally different ways to spread an EMT workload across processors, and they answer different questions.

Data Parallelism Many Runs, Same Model

Data parallelism keeps a single solver process but runs it across multiple sets of data. This is the natural fit whenever you need the same simulation under a sweep of different variables: a parametric study, a Monte-Carlo screen, a control-tuning loop, or a fault-position scan. Because any one run does not depend on any other, the runs can execute simultaneously rather than one after another.

Traditionally a "multiple run" was strictly serial — execute, read the result, change a condition, execute again. Data-parallel execution breaks that chain. Each case is dispatched to its own EMTDC instance, and because the runs are independent, communication between them is minimal: values are exchanged only at the very end and beginning of each run, not on every time step.

Two architectures are useful here. The simplest is a pure parallel sweep, where many identical copies of a case run concurrently. The more elegant is a master–slave arrangement:

A master case holds only the logic that decides what to run next an optimizer, a sweep generator, or a search routine. It is built as an EMTDC case but does almost no power-system computation; it runs a single step to compute new conditions from incoming results.
Slave cases are copies of the actual EMT model. Each receives a condition from the master, runs the real simulation for a period, and returns its objective value.

Values move between projects using an existing PSCAD transfer component (a radio-link / inter-project transfer) configured to carry data from one project namespace to another. The master seeds initial conditions to all slaves; the slaves return results; the master computes the next set and re-dispatches — repeating until it has enough information, then signalling every slave to terminate.

KEY BUILDING BLOCKS

Simulation sets you must tell PSCAD which projects belong together and how they interconnect. A set defines the complete collection of EMTDC instances to generate and launch together. You run the set , not the individual case.

Volley count the number of identical slave instances to launch in parallel (e.g., a volley of three runs three solvers plus PSCAD a clean fit for a four-core machine).

Rank each parallel run is assigned a rank from 1 to N. That index lets each instance look up its own row in a parameter table, so a single model can self-select the right conditions for its slice of the study.

For very large sweeps, PSCAD's Parallel Multi-Run (PMR) executes a pure multiple-run scenario entirely in parallel — on the order of thousands of runs. A run profile such as "500 × 24" means 500 total runs executed at most 24 at a time, so the machine stays saturated without being overwhelmed. A practical rule of thumb on a hyper-threaded desktop is to run roughly twice the number of logical cores concurrently.

A second optimization is the snapshot. One simulation set runs the model up to a chosen instant — say 0.45 seconds — and captures a snapshot of the network state. The bulk study set then starts every run from that snapshot instead of re-simulating the identical pre-event window thousands of times. Between sets you can change project properties so one stage serves the next.

BOTTOM LINE ON DATA PARALLELISM

It is essentially completely scalable. The more cores you have, the more runs execute at once and the faster the study completes, because communication between runs is negligible.

Task Parallelism One Network, Many Cores

Task parallelism answers a different question: how do you accelerate one single large network? Here you take one electrical problem and split it into pieces, where each piece is a subsystem of the same network. Each subsystem becomes a separate project running on its own CPU core, and the pieces are tied back together through a communication interface so they behave as one coherent run.

The split is made across transmission lines. Travelling-wave line models naturally decouple the network in time — a sufficiently long line means each end can be solved a time step apart without loss of accuracy — which is exactly why line ends are the classic place to break a system for parallel or real-time simulation. The server side of the split holds the line model; the client side simply names the project and line it connects to, and PSCAD ties them together at runtime once both are placed in a simulation set.

The catch is communication. Unlike a parametric sweep, these subsystems are tightly coupled: they must exchange interface quantities on every time step. So where data parallelism trades data once per run, task parallelism trades data millions of times per run. That makes communication latency — not raw compute — the dominant factor, and it makes load balancing essential. If subsystems are unevenly sized, the fast ones simply wait on the slow ones, and your speed-up evaporates.

How Data-Parallel and Task-Parallel Execution and High-Performance Interconnects Push EMT Studies Toward Network Scale

Parallel Simulation with Concurrent EMTDC

The Need for Speed: Why EMT Simulation Goes Parallel

Electromagnetic transient (EMT) simulation is the most detailed tool in the power engineer's kit. It resolves the network at microsecond time steps, capturing the switching transients, control interactions, and fast dynamics that phasor-domain tools simply average away. That fidelity has a cost: EMT is computationally heavy, and as networks fill with inverter-based resources, HVDC, and large electronic loads, the cases we are asked to run keep getting bigger and more numerous.

The hardware sitting under most engineers' desks has changed to meet that demand. Multi-core processors with hyper-threading are now ordinary, and a typical workstation exposes eight, twelve, or more logical cores. The question is no longer whether parallel hardware exists — it is whether our simulation tools actually use it. A single EMTDC solver running on a single core leaves most of that silicon idle.

The right way to think about acceleration is holistic. The goal is rarely "make one run faster" in isolation; it is "finish the engineering task faster." And the task usually involves running the same model under many conditions — sometimes hundreds or thousands of conditions — or running one very large network that no single core can chew through in a reasonable wall-clock time. Concurrent EMTDC addresses both. This paper explains how, where the limits sit, and how high-performance interconnects move those limits.

THE CORE IDEA

Concurrent EMTDC runs multiple EMTDC solver instances at the same time, coordinated by PSCAD. Two distinct strategies data parallelism and task parallelism map cleanly onto two very different engineering problems.

Two Axes of Parallelism

The Communication Bottleneck

Because task-parallel performance lives or dies on communication, it is worth measuring carefully and the case itself can contaminate that measurement. To isolate the cost of communication alone, a clean experimental method is to take standard IEEE benchmark networks, tie copies of them together with transmission lines, and watch what happens as you add more pieces.

A Controlled Scaling Experiment

Using IEEE 14-, 39-, 78-, 118-, and 300-bus cases on an eight-core machine, copies were chained into a ring from one piece up to seven, each piece on its own core. In a perfect world, seven pieces on seven cores would finish in the same wall-clock time as one piece, since they run concurrently. Reality departs from that ideal by exactly the amount communication costs you. Even over the local host with no network wire involved at all — the pattern is striking:

Breaking the Bottleneck: High-Performance Interconnects

RDMA and InfiniBand

Standard TCP/IP is slow for this job for two reasons: the operating system kernel has to handle every transfer, and the data is copied multiple times between buffers. Those copies and kernel transitions burn time and CPU cycles that should be spent solving the network.

Remote Direct Memory Access (RDMA) over a high-performance fabric takes a different path. The transfer moves data directly from one machine's memory to another's, bypassing the kernel stack and avoiding the intermediate copies. On a fabric such as InfiniBand, a dedicated processor on the network adapter performs the transfer in hardware, which offloads the operating system entirely leaving more CPU time for EMTDC while the adapter moves data far faster.

Measured on a two-machine test rig (six-core, 3.5 GHz, hyper-threaded workstations — twelve logical cores each), the latency gap is large:

The Road Ahead: A Unified Communication Fabric

The destination is a hybrid that picks the fastest available path automatically. On the same machine, instances talk through shared memory; across machines, they talk through RDMA over a high-performance fabric via the host channel adapter. A communication fabric layer inside each EMTDC instance hides those details from the engineer entirely.

THE ENGINEER'S-EYE VIEW

You should never have to think about sockets, adapters, or copies. You declare that this line connects to that line and the fabric figures out how to move the data as fast as the hardware allows, shared-memory or RDMA, local or remote.

The ambition is to scale to sixty, seventy, even a hundred subsystem splits across multiple machines, and have the whole thing run as if all those cores lived in one box under your desk. With communication latency driven into the low-microsecond range, the granularity limit drops, more cores become usable, and EMT stops being constrained to what one workstation can hold.

Why This Matters for Interconnection Engineering

Case Studies

At Keentel, we treat grid interconnection as a first-order design input, not a downstream formality and EMT studies are increasingly where interconnection is won or lost. Point-of-interconnection studies for inverter-based resources now demand EMT-domain models to capture control interactions, sub-synchronous behaviour, and weak-grid stability. Those studies multiply quickly: many fault positions, many dispatch conditions, many what-if configurations across an entire interconnection cluster.

Concurrent EMTDC maps onto that work in two ways. Data-parallel execution turns a multi-day parametric or fault-scan study into an overnight or same-day deliverable which directly compresses interconnection study timelines. Task-parallel execution, paired with high-performance interconnects, makes it feasible to model a wide-area network in EMT detail as a single coherent case rather than stitching together reduced equivalents. For utility-scale renewables, BESS, and large-load interconnections , that is the difference between a representative answer and an approximate one.

This explainer is part of Keentel Engineering's ongoing technical series on EMT modeling, interconnection engineering, and power system studies. The FAQ and case studies that follow distill the practical questions we hear most, and three anonymized, field-style scenarios that show the methods at work.

The following three scenarios are anonymized and generalized to illustrate the methods in practice. They contain no client, project, or location identifiers and are presented for educational purposes.

Case Study 1 — Automated Control Tuning via Master–Slave Data Parallelism

Method:Data parallelism (master–slave optimization)

Scale:Three parallel slaves on a four-core workstation

Challenge

A control system inside an EMT model required tuning across a parameter space. Done the traditional way run, read the objective, adjust, run again — the optimization was strictly serial and slow, leaving most of the machine's cores idle while a single solver worked through one candidate at a time.

Approach

The optimization logic was lifted out of the simulation and placed in a dedicated master case. The master takes incoming objective values on one side and generates the next set of conditions on the other, running only a single step to do so. The actual EMT model became the slave, replicated as a volley of three so that three solver instances plus PSCAD ran together on a four-core machine. Radio-link transfer components carried objective values back to the master and new initial conditions out to the slaves. Objective values were plotted on a trend graph whose horizontal axis represented run number rather than time, and the master was configured to terminate automatically once the objective crossed a convergence threshold.

Outcome

The optimization converged and self-terminated at roughly 74–75 runs, with no manual intervention.
All available cores stayed busy throughout, replacing a serial loop with concurrent candidate evaluation.
Waveform output on the slaves was suppressed to maximize throughput, since only the objective mattered during the search.

TAKEAWAY

When a study is really a search, separating the decision logic (master) from the physics (slaves) turns an inherently serial optimization into a parallel one — and lets the master stop the moment it has learned enough.

Case Study 2 High-Volume Parametric Study with Snapshot-Initialized PMR

Method:Data parallelism (Parallel Multi-Run + snapshot)

Scale:500 runs in waves of 24; six-core hyper-threaded desktop

Challenge

A switching study required running a network many hundreds of times repeatedly taking transmission lines in and out of service and observing the response. The portion of the network under study was small, but the run count was very high, on the order of hundreds to thousands. A serial campaign of that size was impractical, and re-simulating the identical pre-event window on every run wasted enormous compute.

Approach

The study used Parallel Multi-Run with a two-stage simulation-set design. The first set ran the model to a fixed instant (approximately 0.45 seconds) and captured a snapshot of the network state. The second set started every run from that snapshot, using a run profile of 500 by 24 500 total runs, at most 24 executing concurrently — chosen because roughly twice the logical core count keeps a six-core, hyper-threaded desktop saturated without overload. A rank component assigned each run an index from 1 to 500, and that index drove a table lookup so each run automatically selected its own switching parameters. Data output was suppressed after the snapshot so the runs produced only the required output files.

Outcome

Five hundred-run batches executed as tightly packed parallel waves rather than a long serial queue.
Snapshot initialization eliminated repeated pre-event computation across every run in the batch.
Rank-based table indexing let one unmodified model cover the entire parameter set, with results written straight to output files for post-processing.

TAKEAWAY

For large sweeps, combine three levers: parallel waves sized to your logical cores, a snapshot to skip redundant pre-event simulation, and rank-indexed tables so a single model self-configures each run.

Case Study 3 Region-Scale EMT Decomposition with a High-Performance Interconnect

Method:Task parallelism + RDMA over high-performance fabric

Scale:~2,500+ buses, ~900 lines, ~50 mesh interconnects, 10-way split

Challenge

A large interconnected transmission network more than 2,500 buses and roughly 900 lines — needed to be simulated in full EMT detail as a single coherent case. Run whole on one solver, the all-in-one wall-clock time was about 56 minutes, far too slow for iterative study work, and the network was too large to hold comfortably in a single EMTDC instance.

Approach

The network was decomposed by zones into about ten subsystem projects, split across long transmission lines. The interconnections formed a mesh rather than a simple ring roughly fifty interface points with each subsystem assigned to its own CPU core. For configurations that exceeded the cores available on one machine, RDMA over a high-performance fabric replaced standard TCP between hosts to keep interface latency in the low-microsecond range. A companion sub-case a DFIG wind farm of 22 machine models producing 23 executables was used to confirm cross-host behaviour, since it could not fit on a single host at all.

Outcome

This explainer synthesizes publicly understood concepts in parallel and high-performance computing as applied to electromagnetic transient simulation. Readers seeking primary technical detail may consult:

IEEE standard benchmark network models (14-, 39-, 118-, and 300-bus systems) commonly used to characterize simulation scaling.

Published literature on travelling-wave transmission-line decoupling for parallel and real-time electromagnetic transient simulation.

General references on Remote Direct Memory Access (RDMA) and high-performance interconnect fabrics, including InfiniBand architecture and verbs.

Operating-system scheduling and lock-free inter-process communication concepts (FIFO buffers, shared memory) as applied to compute-bound concurrent processes.

Vendor documentation for the PSCAD / EMTDC simulation environment regarding simulation sets, Parallel Multi-Run, snapshots, and inter-project transfer.

About Keentel Engineering

The throughline is simple. The communication cost per interface is roughly a fixed quantity. When each subsystem does very little computation, that fixed cost is huge relative to the work, and parallelizing actually loses. As each subsystem grows, the same fixed communication cost becomes a smaller and smaller share of the total, until the parallel benefit wins decisively.

The Granularity Insight

THE CENTRAL TRADE-OFF

The key to performance is smaller grains of computation split the network into more, smaller pieces to use more cores but only as far as you can stay ahead of the communication limit . Below a certain grain size, faster communication is the only thing that lets you keep splitting.

That reframes the whole problem. If 78 buses is the smallest grain that breaks even on a given interconnect, then the only way to profitably split into smaller pieces and thus use many more cores is to make communication faster. And communication gets dramatically worse the moment you leave a single machine. On the local host, a round trip is on the order of 20 microseconds. Cross the wire to another workstation over standard TCP/IP and that jumps to roughly 230 microseconds about ten times slower. Since a single workstation does not have enough cores to hold a network broken into dozens of pieces, you must eventually cross that wire. So the wire has to get faster.

Measured Results

The payoff shows up directly in coupled EMT cases. Across a range of split models a trivial AC system, a DC-link case, DFIG wind-farm models, and mixed IEEE-bus combinations the high-performance fabric kept cross-host timing almost indistinguishable from local-host timing, even on the smallest, fastest cases that stress communication hardest. Over standard TCP, the same cases degraded sharply the moment the split crossed from one machine to another.

One case is especially telling. A doubly-fed induction generator (DFIG) wind farm built from 22 machine models produced 23 executables — more than would fit on a single host's cores. It simply could not be run all on one machine; it had to span hosts. Over the high-performance fabric, wall-clock time stayed essentially flat between the local-host and cross-host configurations on the order of 27 seconds whether at 10 or 22 generators — with no measurable degradation from crossing machines. That is the door opening: networks can now be split into many more, much smaller pieces, and spread across many machines, without paying the old cross-host penalty.

Shared Memory on the Local Host

Within a single machine, there is an even faster path than any network protocol: shared memory. The design uses first-in-first-out (FIFO) buffers behind an abstraction layer between EMTDC instances. A deliberate choice here is to avoid semaphores or other locking mechanisms — because if the operating system ever sees an EMTDC instance blocking on a lock, it may de-prioritize that process and starve it of CPU cycles. Keeping every solver visibly "busy" ensures the scheduler keeps all of them running at full tilt.

Frequently Asked Questions

The decomposed run finished in roughly six minutes against a 56-minute baseline close to a ten-to-one, near-linear improvement on ten cores. Over the high-performance fabric, the wind-farm sub-case showed essentially no penalty for crossing machines, where standard TCP would have degraded sharply. The combined result demonstrates that very large networks can be modeled in EMT as a single entity by splitting them across cores and, with a fast enough interconnect, across machines.

TAKEAWAY

Task-parallel decomposition turns an hour-long EMT case into a minutes-long one, and a high-performance interconnect removes the cross-machine penalty together unlocking wide-area EMT modeling that no single workstation could hold.

Author: Sandip R. Patel, P.E. —

Founder & Principal Engineer, IEEE Senior Member. Contact Keentel Engineering to discuss accelerating your EMT, interconnection, or parametric study workload.

DISCLAIMER & TRADEMARKS

Sources & References

Substation Electrical Design: 30%, 60%, 90% & IFC Guide

Tue, 23 Jun 2026 08:22:01 GMT

jun 23, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

A substation is a tightly coupled system of electrical, civil, structural, and protection-and-control elements occupying a confined footprint. The bus has to clear the steel; the steel has to land on foundations; the foundations have to miss the ground grid and the duct bank; the relays have to match the CT ratios; and all of it has to satisfy the interconnecting utility, the applicable codes, and the owner’s budget. No engineer designs all of that correctly in one attempt.

The milestone model solves this by making the design reviewable at defined levels of completeness. At each stage the owner, the utility, and the reviewing engineers can confirm the design is heading the right direction before more hours and more committed procurement are spent. The percentages are not a measure of effort spent; they describe how complete and how firm the documentation is, and what decisions are locked versus still open.

1.1 Interconnection-First: the input that shapes every milestone

For grid-connected substations utility-scale solar and storage, wind, and large new loads such as data centers the single most consequential external input is the interconnection study process. The System Impact Study (SIS) and the subsequent facilities study establish the point-of-interconnection (POI) requirements, the available and post-project fault duties, metering and protection requirements, and the schedule for utility-side work.

Preliminary design values are exactly that preliminary. They are explicitly written to be validated and updated during detail engineering once the interconnection fault model is finalized. A design that treats interconnection as an afterthought routinely discovers, at 60%, that equipment selected at 30% is now under-rated for the confirmed fault current. Treating the interconnection as a first-order driver of the 30% package is the discipline that keeps the 60% package from becoming a redesign.

Keentel insight

The cleanest projects we see are the ones where the 30% one-line is drawn against the interconnection requirements, not against an assumed feeder. When the SIS confirms duties at 60%, the only change should be a refinement of margins not a wholesale re-rating of breakers, bus, and grounding.

A practitioner’s guide to milestone-based engineering, drawing maturity, multidiscipline coordination, and construction-ready deliverables written for developers, utilities, EPCs, and large-load owners.

Substation Electrical Design:

The 30% / 60% / 90% / IFC

Executive Summary

Substations are not drawn in a single pass. A safe, constructible, code-compliant substation emerges through a disciplined sequence of design milestones commonly expressed as 30%, 60%, 90%, and Issued for Construction (IFC) in which the design matures from concept to a sealed, build-ready package. Each milestone has a distinct purpose: the 30% package establishes intent and confirms the station fits the system; the 60% package proves the disciplines coordinate and supports a credible construction estimate; the 90% package is fully checked and ready for final issue; and the IFC package is the sealed, comprehensive set a contractor builds from.

This paper explains what actually belongs in each substation electrical package, how the design develops between milestones, what the licensed Professional Engineer’s seal means and when it is applied, and which industry standards govern the work. It closes with five anonymized case studies drawn from real project patterns and an in-depth FAQ. Throughout, Keentel Engineering’s position is consistent: grid interconnection is a first-order design input, not a downstream formality — and the milestone process is where that principle is enforced.

Why this matters

Most substation cost and schedule risk is created or avoided in the gap between 30% and 60%. Decisions made at concept (voltage, fault duty assumptions, footprint, equipment ratings) propagate through every later drawing. A rigorous milestone discipline catches the expensive problems while they are still lines on a drawing, not steel in the yard.

1. Why Substations Are Designed in Milestones

2. The 30% Design Package Concept & Preliminary

Purpose: demonstrate understanding of the project and the system it connects to, confirm the station concept is viable, and define what the finished design will contain. The 30% package is where the owner and reviewers decide whether the design is meeting expectations, whether a course correction is needed, or whether project requirements should be re-evaluated while changes are still cheap.

2.1 Key engineering tasks at 30%

Preliminary electrical load and capacity basis — station ratings derived from the generation or load to be served and the POI voltage.
Voltage selection and major equipment sizing transformer MVA and ratio, primary and secondary voltages, basic insulation level (BIL), preliminary continuous current and interrupting ratings.
Preliminary fault-duty basis first-pass three-phase and line-to-ground fault currents (often from the interconnection request and prior studies), explicitly flagged for validation against the SIS / fault model.
Initial single-line diagram (SLD) the spine of the entire design: sources, bus arrangement, transformer, breakers, switches, instrument transformers, surge arresters, and metering.
Design criteria basis the controlling standards, clearances, and parameters captured in a design criteria document so every later drawing inherits a single source of truth.

2.2 Typical 30% substation deliverables

For a collector or interconnection substation, a 30% electrical package typically contains:

Preliminary power one-line diagram (high-level view).
Preliminary general arrangement / equipment layout plan and preliminary site plan.
Preliminary cable-routing and duct-bank concept.
Design criteria document — system parameters (nominal/maximum voltage, BIL, continuous current, fault duties), electrical clearances, grounding basis, lightning-protection basis, AC station-service concept.
List identifying major electrical equipment and components.
Identification of long-lead procurement items (the long-lead clock starts here).
A preliminary construction cost estimate and a complete list of the drawings, specifications, and calculations the final package will contain.

The 30% deliverable list is itself a deliverable

A defining feature of a good 30% package is that it tells the owner what every later package will contain — the full drawing index, the calculation list, and the specification list. This is also where drawing numbers are reserved so they appear consistently from 60% onward. If a reviewer cannot see the shape of the finished design at 30%, the package is not really at 30%.

2.3 What is deliberately NOT finished at 30%

Schematic and wiring diagrams, detailed grounding-grid calculations , foundation design, and final equipment specifications are intentionally immature at 30%. The fault duties are not yet validated. Committing to detailed schematics before the fault model is confirmed wastes hours that the 60% phase will erase.

3. The 60% Design Package Detailed Design & Multidiscipline Coordination

Purpose: the 60% package is the primary coordination milestone. It carries enough detail that a construction estimator can develop a credible class estimate, and that reviewers can see how the design meets the project intent. Where information is intentionally omitted, a placeholder and an explanation take its place rather than a silent gap.

3.1 Key engineering tasks at 60%

Finalize major-equipment specifications transformer, switchgear/breakers, switches, instrument transformers, surge arresters now reconciled against validated fault duties and continuous ratings.
Develop the protection-and-control (P&C) scheme protection one-lines with ANSI device function numbers, relay selection, CT/VT ratios, and the trip/control philosophy.
Detailed cable sizing and routing ampacity with derating, conduit fill, cable-tray and duct-bank layouts.
Preliminary grounding-grid and lightning-protection studies step/touch-voltage analysis and direct-stroke shielding analysis carried to a reviewable result.
Multidiscipline clash review electrical, structural, and civil reconciled spatially, increasingly through a coordinated 3D model, so bus, steel, foundations, grounding, and duct banks physically coexist.

3.2 Typical 60% substation deliverables

Building on the 30% set, the 60% package typically adds:

Updated, detailed one-line diagram(s) and protection one-line(s).
Detailed equipment layout, elevations, and sections.
Schematic and wiring diagrams (relaying, control, AC/DC station service).
Preliminary grounding plan and lightning-protection layout with supporting calculations.
Conduit/cable schedules, conduit routing, and cable-tray/rack layouts.
Near-final equipment schedules and an equipment label schedule.
Construction specifications approaching final form (organized to the CSI MasterFormat structure), and final procurement specifications for long-lead items so orders can be placed.

Reviewers should be able to develop a Class 2-level construction estimate (per AACE International’s cost-estimate classification) from a complete 60% set — which is why a draft of every sheet, specification, and calculation is expected, even if some carry placeholders.

Keentel insight

60% is where projects are won or lost on coordination. The most common avoidable rework we see is a grounding grid and a foundation/duct-bank layout designed in isolation, then colliding. We run the clash review against a single coordinated model precisely so the conflict shows up as a redline at 60%, not as a field change at construction.

4. The 90% Design Package Checked & Review-Ready

Purpose: the 90% package is fully checked and essentially ready for final issue and construction. The distinction from 60% is one of completeness and verification, not new scope. By 90% the engineering decisions are settled; the work is to confirm they are correct, internally consistent, and well drafted.

4.1 What checking actually means at 90%

Checking at 90% goes beyond technical adequacy. A proper 90% review addresses three layers:

Technical adequacy calculations support the ratings shown; protection settings, grounding results, clearances, and equipment selections are verified against the design criteria and the validated study basis.
Drafting-standard compliance sheets follow the drawing standard: borders, title blocks, drawing numbers, symbols, legends, and cross-references are correct and consistent.
Editorial quality and workmanship errors, omissions, mismatched references, and inconsistencies between sheets are resolved. Each calculation carries its checklist and check record.

4.2 Typical 90% substation deliverables

A complete, checked set of every drawing in the index one-lines, protection one-lines, general arrangement, elevations and sections, grounding plan and details, lightning protection, AC/DC station service, control-house layout, conduit/cable, and schedules.
Final grounding-grid and lightning-protection calculations; final cable and conduit-fill calculations; arc-flash, load-flow, and short-circuit results as applicable.
Construction specifications in final form; master submittal register aligned to the specifications.
A design-requirements compliance record each requirement traced to where and how the design satisfies it (drawing/sheet, calculation, or test).

90% is a verification gate, not a design gate

If a reviewer finds genuinely new engineering at 90% — a changed transformer rating, a reworked protection scheme — that is a signal the 60% package was not truly at 60%. The 90% review should surface corrections and refinements, not surprises.

5. Issued for Construction (IFC) The Final Package

Frequently Asked Questions

Purpose: IFC is the comprehensive, sealed package a contractor builds from. It incorporates all 90% review comments and carries no new, unreviewed information. By definition, nothing should appear at IFC that the owner and reviewers have not already seen it is the clean, final, releasable expression of the design.

5.1 What IFC contains

The complete, finalized, and sealed construction drawing set.
Foundation and grounding plans, final protection-and-control schematics, conduit/cable routing details, and the bill of materials (BOM).
Final specifications and the procurement/submittal documentation needed to build and commission the station.
All 90% comments dispositioned and incorporated; drawings carry the IFC revision and the engineer’s seal as required.

5.2 IFC is a status, not the end of engineering

Issuing for construction does not freeze a project in amber. Field conditions, vendor-certified equipment dimensions, and requests for information (RFIs) generate revisions. These are handled through controlled revisions to the IFC set — each change tracked, re-checked, and re-issued — rather than informal markups. The engineering rigor that produced the IFC package is the same rigor that governs every revision after it.

The 100% / record-set relationship

Some owners use a 100% or “final” designation that should contain no information not already reviewed at 90% — it simply incorporates final comments and adds the required professional seals. After construction, the design is updated to an as-built / record set reflecting what was actually installed. IFC, 100%, and as-built are points on the same continuous chain of custody for the design.

Deliverable maturity at a glance

The same document matures across milestones rather than appearing all at once. A representative progression for core substation electrical deliverables:

This is a representative pattern, not a contractual list the exact deliverable set is defined per project by the owner’s scope of work and the interconnecting utility’s requirements.

Construction-level substation drawings carry the seal and signature of a licensed Professional Engineer (P.E.). The seal is not a formality or a logo — it is a legal attestation that a qualified, licensed engineer prepared the work, or directly supervised its preparation, under what licensing law calls “responsible charge,” and takes professional responsibility for it.

6.1 What the seal means

Responsible charge The sealing engineer exercised direct control and personal supervision of the engineering work not a rubber stamp of someone else’s drawings.
Jurisdiction Engineering licensure is granted by individual state boards. The engineer who seals must hold a current license in the state (jurisdiction) where the project is located, regardless of where the firm is based.
Accountability The seal ties the work to a named, licensed individual who is professionally and legally accountable for it. Permitting authorities, owners, and utilities rely on it as the mark of a constructible, code-aware design.

6.2 When the seal is applied

Sealing generally occurs at final issue IFC and/or 100% once the design is checked and complete. Interim packages (30%, 60%) are typically marked PRELIMINARY or NOT FOR CONSTRUCTION and are not sealed for construction, because they are explicitly subject to change. Some jurisdictions and owners require preliminary submittals to be identified as preliminary and signed accordingly; the controlling rule is always the state board’s and the owner’s requirements for that project.

Plain-language summary

A sealed drawing says: a licensed engineer, accountable under that state’s law, stands behind this design as ready to build. That is why IFC drawings are sealed and 30% drawings are not — and why the engineer must be licensed where the substation will actually be built.

6. The Licensed Professional Engineer’s Seal

7. The Standards Framework Behind Each Milestone

Substation design is governed by a layered framework of codes and consensus standards. A practical distinction matters here: codes (such as the National Electrical Safety Code and the National Electrical Code) are adopted into law by jurisdictions and are mandatory where adopted, while most IEEE documents are consensus guides and recommended practices that become binding when a regulator, utility specification, or contract incorporates them. Engineers treat the relevant IEEE standards as authoritative practice; their legal force comes from adoption, not from the standard itself.

7.1 Grounding and personnel safety

IEEE Std 80 Guide for Safety in AC Substation Grounding: soil modeling and tolerable step/touch-voltage and grid-current design.
IEEE Std 81 measuring earth resistivity and ground impedance (e.g., Wenner four-pin testing) to validate the grid model against real soil.
IEEE Std 837 qualifying permanent grounding connections (exothermic welds, compression connectors).

7.2 Protection, automation, and control

IEEE Std C37.2 standard device function numbers (e.g., 50/51 overcurrent, 87 differential) used on protection one-lines.
IEEE Std C37.1 SCADA and automation systems.
IEEE Std 1815 (DNP3) power-system communications protocol for grid automation.
IEEE Std 1686 cyber-security capabilities for intelligent electronic devices (IEDs).

7.3 Physical design, equipment, and site

IEEE Std 605 design of rigid-bus structures (clearances, mechanical loading, ampacity).
IEEE C57 series power transformers, regulators, and instrument transformers (CTs/VTs).
IEEE Std C37.122 gas-insulated substations (GIS).
IEEE Std 69 seismic design of substations.
IEEE Std 525 design and installation of cable systems (tray, ampacity derating, pull tension).
IEEE Std 998 direct-stroke shielding / lightning protection (e.g., rolling-sphere and fixed-angle methods).
IEEE Std 1127 (with IEEE Std 980) environmentally compatible substations: oil containment, acoustic limits, community integration.
IEEE Std 1402 & IEEE Std 1264 physical/electronic security and animal/wildlife mitigation.
ANSI C37.32, NESC, NEC, NFPA, RUS Bulletin 1724E-300 clearances and spacing, code-level safety requirements, fire and life safety, and the standard design-guide framework for substations.

A note on accuracy

It is common, but imprecise, to call IEEE 80 a “law” or a “mandate.” IEEE 80 is a consensus guide. It becomes enforceable when a utility, regulator, or contract requires it — which most do. Keentel states this distinction plainly because owners deserve to understand what is legally binding versus authoritative best practice.

8. Keentel Engineering’s Substation Electrical Design Services

Keentel Engineering provides full-lifecycle substation electrical design from interconnection-driven concept through sealed Issued-for-Construction packages for utility-scale renewables and storage, large new loads including data centers, utility and industrial substations, and transmission and switching stations. We engineer the milestone process itself, not just the drawings, so owners get predictable packages, defensible decisions, and a construction set that builds cleanly.

8.1 What we deliver

Single-line and protection one-line development the backbone of the design, drawn against interconnection requirements from day one.
Physical design general arrangement, equipment layout, elevations and sections, bus and clearance design (ANSI C37.32 / NESC).
Grounding and lightning protection IEEE 80 grounding-grid design and step/touch-voltage analysis; IEEE 998 direct-stroke shielding.
Protection & control relaying schemes, schematics and wiring, CT/VT sizing, and control-house layout.
Power system studies short-circuit, load-flow, coordination, and arc-flash to validate ratings and settings.
Interconnection (POI) engineering aligning the substation design with the SIS/facilities study and the utility’s requirements.
Sealed deliverables 30%, 60%, 90%, and IFC packages, with P.E.-sealed construction drawing .

8.2 Why owners work with Keentel

Interconnection-first engineering. We treat the POI as a design driver, so the 60% package refines the concept instead of rebuilding it.
Milestone discipline. Defined, reviewable packages with no surprises 90% is a verification gate, IFC carries nothing new.
Coordination rigor. Clash review against a coordinated model so conflicts surface as redlines, not field changes.
Standards fluency. Clear about what is code, what is consensus practice, and what the project actually requires.

Engage Keentel

Whether you are starting a 30% concept, recovering a stalled 60% package, or need a sealed IFC set on a deadline, Keentel can lead the design or step in mid-stream. Reach us at keentel.com — offices in Tampa, FL and Austin, TX.

9. Case Studies

The following case studies are composites drawn from real project patterns. All client names, locations, project identifiers, and proprietary details have been removed; figures are illustrative and rounded. They are presented to show how the milestone progression behaves on real work, not to describe any single identifiable project.

Case Study 1 Solar-plus-storage collector substation: the fault-duty correction

Context

A roughly 50 MW PV facility with battery storage required a collector substation stepping 34.5 kV collection up to a sub-transmission POI, with collector-feeder breakers, a step-up transformer, high-side breaker, disconnects, instrument transformers, and revenue metering.

Challenge

The 30% package was built on interconnection-request fault estimates. When the System Impact Study returned, the confirmed line-to-ground and three-phase duties were higher than the preliminary basis, pushing the bus-design and ground-grid fault currents up.

How the milestones handled it

Because the 30% one-line and equipment list were drawn against the interconnection requirements with documented margins, the 60% phase absorbed the higher duties as a refinement: equipment interrupting ratings and momentary withstand were re-checked, the ground-grid design current was updated under IEEE 80, and bus mechanical design was re-verified. No major equipment had to be re-selected.

Outcome

The 60% set supported a credible construction estimate; 90% checking confirmed the grounding and clearance results; IFC issued on schedule with sealed drawings. The correction cost drawing hours, not procurement.

Case Study 2 Hyperscale data-center load interconnection: procurement on the critical path

Context

A large new data-center load required a step-down substation from a transmission-class POI to medium-voltage distribution within the campus, on an aggressive energization deadline.

Challenge

The large power transformers and high-voltage breakers had procurement lead times far exceeding the design schedule. If the order waited for a finished design, energization would slip by quarters.

How the milestones handled it

Long-lead items were identified at 30%. Final procurement specifications for the transformers and breakers were completed at 60% ahead of the rest of the set so orders were placed while detailed design continued. The remaining drawings matured to 90% and IFC around the confirmed equipment dimensions.

Outcome

Decoupling long-lead procurement from full design completion protected the energization date. Vendor-certified dimensions fed the foundation and bus design before IFC, avoiding field rework.

Case Study 3 Brownfield distribution-substation rebuild: designing around a live bus

Context

An aging utility distribution substation required replacement of major equipment while continuing to serve load through portions of the existing yard.

Challenge

The station could not be fully de-energized. Construction had to proceed in stages, each requiring its own safe clearances, temporary protection, and outage windows.

How the milestones handled it

The 30% concept established a phased construction sequence as a design constraint, not an afterthought. The 60% package coordinated demolition and new-work drawings against the live equipment and the existing ground grid. The design was issued for construction as sequenced IFC packages aligned to the outage windows rather than a single release.

Outcome

Staged IFC issuance let the contractor build in step with available outages while load stayed served, with clearances and grounding verified for each phase.

Case Study 4 Wind-collector repowering: re-using steel, re-checking the grid

Context

A repowering effort updated the generation feeding an existing collector substation, with the owner seeking to reuse existing structures and foundations where possible.

Challenge

Updated generation changed the fault contribution and collector configuration. Reusing existing steel and foundations was attractive but could not be assumed safe without verification.

How the milestones handled it

At 30%, existing drawings were inventoried and the reuse-versus-replace question was framed explicitly. At 60%, the ground grid was re-evaluated under the revised fault current per IEEE 80, and bus/structure loading was re-checked; some reuse was confirmed, some elements were upgraded. 90% verified the mixed old/new design held together.

Outcome

Documented reuse where the math supported it and targeted upgrades where it did not capturing cost savings without compromising the grounding safety case.

Case Study 5 Industrial substation on a constrained, seismically active site

Context

A manufacturing facility needed a new substation stepping a sub-transmission supply down to plant medium-voltage, on a tight footprint in a region with meaningful seismic demand.

Challenge

Limited land pressured the layout, and seismic requirements raised structural and equipment-qualification concerns for rigid bus, insulators, and bushings. Arc-flash energy at the plant interface also needed control.

How the milestones handled it

The 30% phase evaluated a compact arrangement (including gas-insulated equipment per IEEE C37.122 as an option) against the footprint. 60% carried seismic design per IEEE 693 into structure and equipment selection, and the protection scheme was set to manage arc-flash incident energy. 90% checking reconciled the dense layout’s clearances and grounding; IFC issued sealed.

Outcome

A constructible station on a constrained site, with seismic qualification and arc-flash mitigation engineered in from concept rather than retrofitted at the end.

This white paper is provided by Keentel Engineering for general informational and educational purposes and reflects general industry practice as of its publication date. It does not constitute engineering, legal, or regulatory advice, and it does not establish a client relationship. Design milestone definitions, deliverable lists, licensure and sealing requirements, and applicable codes and standards vary by jurisdiction, by interconnecting utility, and by project scope; requirements should be confirmed for each specific project with the appropriate authorities and a licensed engineer.

Keentel Engineering is an independent engineering firm. References to standards-development organizations, codes, software, equipment, and other third parties — including IEEE, ANSI, NEMA, NFPA, AACE International, CSI, and the USDA Rural Utilities Service — are for identification and educational purposes only. Keentel Engineering is not affiliated with, endorsed by, or sponsored by any such organization, and all trademarks are the property of their respective owners. The case studies are anonymized composites drawn from general project patterns; figures are illustrative, and they do not depict any single identifiable client, project, or location.

Disclaimer

Sources & References

IEEE Std 80 — Guide for Safety in AC Substation Grounding.
https://standards.ieee.org/ieee/80/
IEEE Std 81 — Guide for Measuring Earth Resistivity, Ground Impedance, and Earth Surface Potentials.
https://standards.ieee.org/ieee/81/
IEEE Std 605 — Guide for Bus Design in Air-Insulated Substations. https://standards.ieee.org/ieee/605/
IEEE Std 693 — Recommended Practice for Seismic Design of Substations. https://standards.ieee.org/ieee/693/
IEEE Std 998 — Guide for Direct Lightning Stroke Shielding of Substations. https://standards.ieee.org/ieee/998/
IEEE Std C37.2 — Standard for Device Function Numbers, Acronyms, and Contact Designations. https://standards.ieee.org/ieee/C37.2/
IEEE Std 1127 — Guide for the Design, Construction, and Operation of Electric Power Substations for Community Acceptance and Environmental Compatibility. https://standards.ieee.org/ieee/1127/
USDA RUS Bulletin 1724E-300 — Design Guide for Rural Substations. https://www.rd.usda.gov/sites/default/files/UEP_Bulletin_1724E-300.pdf
AACE International — Recommended Practice 17R-97, Cost Estimate Classification System. https://web.aacei.org/
CSI MasterFormat — Construction Specifications Institute. https://www.csiresources.org/standards/masterformat
National Electrical Safety Code (NESC), IEEE C2.
https://standards.ieee.org/products-programs/nesc/

WHITE PAPER | PJM's Expedited Interconnection Track (EIT) — What 250 MW+ Developers Must Know

Sun, 21 Jun 2026 20:39:53 GMT

June 15, 2026 | White Paper

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

PJM’s Expedited Interconnection Track (EIT)

A fast lane for large capacity resources

Interconnection Cost Estimation Before You File

Sun, 21 Jun 2026 18:48:55 GMT

May 21, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Interconnection costs have climbed substantially since the early 2000s. But the most important finding is which projects carry the highest costs.

Among recent (2018–2024) projects, the data show a striking spread by outcome:

Projects that completed all studies averaged about $194/kW.

Active projects (still in the process) averaged about $294/kW.

Withdrawn projects averaged about $671/kW — more than three times the cost of completed ones.

That pattern is the whole story in miniature. High costs don't just correlate with withdrawal; they cause it. And the driver behind the high end is network upgrade costs — the broader transmission reinforcements triggered when a new generator creates reliability or stability problems on the grid. Across all projects, network upgrades grew from roughly 35% of total interconnection costs in the 2000s to about 85% in 2018–2024. Point-of-interconnection (POI) costs — the substation and line work right at the connection — are real, but they are no longer the part that kills deals.

The financial weight is easy to underestimate. The study notes that an interconnection cost of $500/kW translates to roughly $50 million for a 100 MW project. Even where a developer can eventually be credited back for network upgrades over time (as is common in non-ISO territories, sometimes over a period as long as 20 years), the up-front capital and carrying cost still have to be financed first.

New federal research shows interconnection costs are rising, opaque, and the single biggest reason projects walk away from the queue. Here's what the data says — and how Keentel Engineering helps developers understand their cost exposure in weeks, not months.

If you develop generation, storage, or large new loads like data centers, you already know the hardest question in the business isn't "can we build it?" It's "what will it cost to connect it — and will we find out in time to do anything about it?"

A recent research effort funded by the U.S. Department of Energy (DOE) put hard numbers to that problem. The researchers manually assembled more than 2,100 project-level interconnection cost estimates across five non-ISO balancing authorities — Bonneville Power Administration (BPA), PacifiCorp (PAC), and the three Duke utilities (Carolinas, Progress, and Florida). The picture it paints is one every developer should take seriously: interconnection has become a slow, expensive process of cost discovery , and the projects that survive it are increasingly the ones that understood their numbers early.

Below, we summarize what the research found, what it means for your project economics, and how Keentel Engineering's feasibility and pre-application cost services are built to address exactly these pressures.

Why Most Interconnection Projects Never Get Built and How to Know Your Number Before You File

The queues are exploding and almost nothing makes it through

Interest in building is at record highs. Across the five balancing authorities studied, the active interconnection queue tripled between 2020 and 2024, reaching roughly 229 gigawatts of capacity actively seeking a connection by the end of 2024. In some regions the growth is staggering — BPA's active queue grew nearly nine-fold compared to 2017. Put differently, the capacity lined up to interconnect is about four times the combined peak load of those regions.

Yet very little of it gets built. Of the projects that requested interconnection between 2000 and 2019, only about 15% had reached commercial operation by the end of 2024. The rest stalled, suspended, or — most commonly — withdrew. In the cost sample itself, withdrawn requests outnumbered every other category.

The reason matters enormously: most projects withdraw after they receive their cost estimates. The queue, in other words, is functioning less like a construction pipeline and more like an expensive way to find out a project doesn't pencil out.

Costs are rising and network upgrades are the reason

Where you connect, what you build, and how you connect all move the number

The research breaks cost down across several dimensions, and each one is directly relevant to siting strategy:

Market structure

Recent interconnection costs in the non-ISO balancing authorities studied were higher than in the major ISOs — for completed projects, more than double. Network upgrade costs explain most of that gap.

Fuel type

Among all recent requests, natural gas projects were cheapest to interconnect (about $150/kW), while solar ( $509/kW), wind ( $504/kW), and storage (~$437/kW) were considerably higher — largely because they tend to sit at more remote points on the grid with weaker network capacity. Notably, the gap narrows sharply for projects that actually complete studies (solar ~$216/kW, wind ~$103/kW, storage ~$151/kW), which tells you that careful siting separates the viable projects from the ones that withdraw.

Project size

Bigger projects enjoy real economies of scale. Costs fell from about $763/kW for small (1–50 MW) projects to roughly $244/kW for very large (750+ MW) ones, driven mostly by proportional declines in network upgrade costs.

Service type

Firmer "network resource" service (NRIS) carried higher network upgrade costs than "energy resource" service (ERIS) — about $467/kW vs. $353/kW across all recent requests. Choosing the right service level for your commercial goals is a cost lever in its own right.

Location

High network upgrade costs cluster geographically — projects strung along the same constrained transmission corridors tend to share the same painful bills. Two sites a short distance apart can have radically different interconnection economics.

The real problem: you can't see any of this before you commit

Tie all of this together and a single theme emerges, one the researchers state plainly: developers usually don't have a clear sense of expected interconnection costs before they enter the queue. They use the interconnection study process itself as the way to discover those costs — a process that is slow, carries rising milestone payments, and frequently ends in withdrawal once the number finally lands.

The data transparency simply isn't there yet. The researchers describe having to hand-extract cost figures from study PDFs at an average of 20 minutes per project — roughly 700 hours of manual work — precisely because this information isn't published in any accessible, machine-readable form. If a national lab has to do that to see the picture, an individual developer staring at a single site has almost nothing to go on at the moment the decision matters most.

That is the gap Keentel Engineering exists to close.

How Keentel helps: see the number before you spend the money

Keentel Engineering 's services are built around a simple principle that this research validates: the projects that succeed are the ones that understand their feasibility and cost exposure before they file an interconnection request or commit capital. We deliver that understanding up front, typically in a matter of weeks rather than the months a formal study process takes.

First-pass site feasibility screening

Before you commit to a site, our engineers assemble the factors that actually determine whether it works — nearby transmission infrastructure and its capacity, substation headroom on both the injection and withdrawal side, land ownership, competing generators already in or entering the queue, natural gas pipeline flows where relevant, and long-range energy price forecasts — onto a single, location-specific picture. Instead of chasing every lead with equal effort, you can rank sites, rule out the non-starters, and concentrate capital where there's a genuine path. Given how sharply costs cluster by location, this screening is often the difference between a $103/kW outcome and a $671/kW one.

Pre-application interconnection cost estimation

This is the heart of it. We quote likely network upgrade and POI costs before the application is filed — turning the queue from a tool of cost discovery into a decision you make with the number already in hand.

For generation projects , our analysis is fully configurable. Our engineers set their own queue-withdrawal assumptions, adjust competing project sizes, and model different fuel types to build a realistic, defensible view of what the queue will look like when your project reaches its first phase — rather than the generic output of a standard study.

For load projects data centers , industrial facilities , large-scale manufacturing we deliver a current-state read on exactly what the grid can absorb at a given bus today. When a site-selection decision hinges on hundreds of millions of dollars, that precision is the point.ceter

Operator-grade credibility

Our modeling runs on the same class of power-flow methodology that grid operators use for their own interconnection studies, so the cost estimates you take into a financing conversation or a site committee carry real weight.

Service-type and sizing strategy

Because the research shows NRIS/ERIS choice, project size, and location all move the cost number materially, our engineers don't just hand you a figure — we help you understand the levers, so you can structure the project to improve its odds before you're locked into a queue position.

The bottom line

Frequently Asked Questions

The DOE-funded research is, in effect, a national-scale confirmation of what experienced developers feel intuitively: interconnection cost uncertainty is the chokepoint, and the cost of finding out too late is measured in millions of dollars and years of lost time. At a moment of unprecedented load growth and queue congestion, the advantage goes to whoever can see the grid clearly before they commit.

That's what we do. If you're evaluating a site for generation, storage, a data center , or an industrial facility, talk to Keentel Engineering before you file — and find out whether your site works, and what it will cost to connect, in weeks rather than months.

Power System Resilience Metrics: A Practical Guide

Sat, 20 Jun 2026 14:27:06 GMT

May 20 , 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

2. The resilience lifecycle and its attributes

Resilience is best understood as a sequence of capabilities exercised before, during, and after an event — conventionally grouped into a proactive set (anticipate, prepare, adapt) and a reactive set (withstand, respond, recover). Several attribute taxonomies map onto this lifecycle; the most widely cited is the 5Rs.

Beyond Reliability: Why the Grid Needs to Be Measured on Its Worst Day

Measuring the Grid on Its Worst Day A Practical Guide to Power System Resilience Metrics

Ask most utilities how their grid is performing and they will hand you a SAIDI or SAIFI number — the familiar reliability indices that track how often the average customer loses power and for how long. Those numbers are useful, and they have driven real improvements for decades. But they share a quiet blind spot: by design, they look away from the worst days. Under the standard IEEE 1366 methodology, the days with the most extreme outages — the major-event days — are statistically excluded so that ordinary, blue-sky reliability can be compared fairly year over year. The hurricane, the ice storm, the multi-day heat dome, the cascading failure: precisely the events the public remembers, the regulator scrutinizes, and the data center worries about, are the events the headline reliability metric is built to ignore.

That gap is where resilience lives. Resilience is not a synonym for reliability — it is a different question. Reliability asks: under normal conditions and a short, pre-defined list of contingencies, can the system keep the lights on? Resilience asks: when a high-impact, low-probability event overwhelms that list, how far does the system fall, how long does it stay down, and how fast does it claw its way back? One is about avoiding interruptions; the other is about surviving and recovering from the interruptions you cannot avoid.

Three forces have moved this from an academic distinction to a board-level and rate-case priority. Extreme weather is now a planning assumption, not a tail risk — Winter Storms Uri and Elliott, repeated arctic events, and intensifying hurricane seasons have triggered successive federal performance reviews and, in late 2024, a new mandatory NERC standard (TPL-008-1) requiring transmission planners to study extreme heat and cold and develop mitigation plans. Load is exploding — data centers and AI campuses are driving forecasts of roughly a quarter more peak demand, and those loads demand a resilience case before they will site. And regulators want proof — public utility commissions increasingly expect comparative, reproducible metrics before they will let a utility recover the cost of a hardening or restoration-improvement program.

The trouble: everyone wants resilience, nobody agrees how to measure it

Robustness, redundancy, resourcefulness, response, recovery — the attributes of a resilient system are easy to list and hard to quantify. The result, across utilities and jurisdictions, is a patchwork: every program defines resilience slightly differently, so investments cannot be compared and value cannot be demonstrated to ratepayers. Closing that gap requires a shared, defensible way to put a number on resilience — one that works at the transmission level and the neighborhood level, in planning and in real time, and that a regulator can audit.

The performance curve: a number you can defend

The most durable way to quantify resilience is to track system performance — load served, customers up, critical facilities energized — through the full arc of an event, and measure the area of the dip. Two models dominate. The resilience triangle captures the simplest story: performance drops at the moment of impact, then recovers along a line back to normal; the triangular area of lost performance is the resilience penalty — smaller is better. The resilience trapezoid is the grown-up version: it separates the event into the disturbance phase (how far and how fast performance falls), a degraded plateau (how long the system sits at the bottom while crews assess damage and mobilize), and the restorative phase (how quickly it climbs back). Because real outage and restoration processes overlap in time, the same idea generalizes to cumulative outage and restore curves built directly from utility data.

The one-paragraph version

Reliability indices hide the worst days; resilience metrics are about exactly those days. The practical move is to draw the performance curve of an event — the triangle or, better, the trapezoid — and measure the area lost. From there you can attach a dollar value (avoided customer-minutes, energy not served, outage cost) and compare investments on equal footing: this is what turns “we should harden the coast” into “this project buys X fewer customer-hours of outage per dollar.”

That last step is what makes resilience metrics actionable rather than rhetorical. Once an event’s performance curve is quantified, you can re-run history: how much smaller would the triangle have been if a feeder had been undergrounded, if a substation had been flood-walled, if restoration crews had been pre-staged? The difference in area, multiplied by the value of lost load, is the benefit-cost case a commission can approve. Resilience stops being a slogan and becomes an engineering and economic argument.

The sections that follow lay out the framework Keentel uses — how resilience differs from reliability in scope and method, how to quantify it with the performance curve, how to classify metrics so the right one is used for the right decision, how to value resilience in dollars, and what data it takes. Three anonymized case studies then show the framework at work, from a coastal utility’s hardening rate case to a data-center interconnection and a community microgrid scorecard.

A Working Framework for Power System Resilience Metrics

This section is written for the planner, the rate-case witness, and the interconnection engineer who has to choose, compute, and defend a resilience metric. It draws on the established resilience literature and the frameworks published by DOE, NREL, PNNL, and the national labs, and aligns with the direction of NERC’s extreme-weather planning standard (TPL-008-1) and the FERC/NERC post-event reviews.

1. Resilience is not reliability — and the difference is methodological

The two disciplines diverge on almost every axis that matters for analysis. Reliability evaluates a fixed, experience-based list of contingencies (single and a few credible multiples) and demands that adequacy and security criteria be met for every one. Resilience deliberately reaches past that list to the high-impact, low-probability events that conventional design treats as out of scope — and accepts that, under those events, a controlled, temporary degradation of service is a legitimate outcome rather than a failure.

3. Quantifying resilience: the performance curve

Resilience is quantified by tracking a performance measure (load served, customers energized, critical loads up) through an event and measuring the area of lost performance. Two canonical models and one data-driven generalization are used.

Resilience triangle

Performance sits at its nominal level, drops at the moment of impact, then recovers linearly to normal. The triangular area between the nominal line and the depressed performance trajectory is the resilience loss; a smaller triangle means a smaller drop and/or a faster recovery. It is intuitive and a good first cut, but it collapses the whole event into a single drop-and-recover.

Resilience trapezoid

The trapezoid resolves the event into three phases and is the more faithful model for utility decision-making: (I) the disturbance progress phase — from onset to the performance nadir, capturing depth and speed of the drop; (II) the post-disturbance degraded state — the plateau at the bottom while damage is assessed and crews mobilize, a phase the triangle ignores entirely; and (III) the restorative state — the climb back to normal. The area under the curve, phase by phase, is the metric, and each phase points to a different intervention: robustness flattens Phase I, logistics and pre-staging shorten Phase II, and automation/crews accelerate Phase III.

From models to data: outage and restore curves

In real utility records the phases overlap — restoration begins before the last customer goes out. So the trapezoid is generalized into two cumulative curves built directly from event data: the outage process (cumulative customers/load lost) and the restore process (cumulative restored). Every trapezoid metric — depth, duration, area — applies to these curves, and they can be assembled at scale from time-stamped outage management system data, which is what makes the framework practical rather than theoretical.

4. Classifying metrics so the right one fits the decision

No single number answers every resilience question. The practical skill is matching the metric to the decision context along a few axes.

Planning vs. operations. Planning metrics are probabilistic and forward-looking (expected annual resilience loss across a hazard distribution); operational metrics are fast, scenario-specific, and support real-time response.

Transmission vs. distribution. Transmission resilience is dominated by cascading and N–k events and bulk-system stability; distribution resilience is dominated by localized weather damage to overhead assets and feeder-level restoration.

Grid vs. community. Community-scale metrics add microgrid islanding success, critical-facility service continuity, and socio-economic and public-health impact — dimensions that pure grid metrics miss.

Performance, consequence, and non-performance. Performance metrics measure load/duration/frequency of loss; consequence metrics measure the downstream economic and social cost; non-performance metrics score inherent attributes (hardening level, redundancy, crew readiness). Composite indices blend them — but the aggregation must stay transparent and auditable.

System-wide vs. localized. Local metrics rarely aggregate cleanly into system metrics because interconnecte d networks exhibit emergent, non-linear behavior — a localized disruption can propagate system-wide in ways a sum of local scores will not predict.

5. Modeling vulnerability: two complementary routes

Model-based (fragility) approach. Component fragility curves give the probability a line, pole, insulator, or substation fails as a function of stress (wind, ice, flood, debris). Aggregated across the system, they predict performance — and, crucially, let you evaluate a specific hardening investment on a specific asset. The cost is the difficulty of cataloguing every component, age, design, and exposure, and of calibrating and validating the combined result.

Statistical (area outage-rate) approach

Pairing detailed utility outage records with weather records yields area outage-rate curves — the observed system outage rate as a function of a measured stress such as wind speed. These capture the system’s real, combined response without modeling every component, and they translate directly into historical resilience metrics. Their limit is that they describe the past: they extrapolate to a changing climate or a hardened future only with supporting engineering analysis of how the outage rate itself would shift.

6. Putting a dollar on resilience

Metrics drive decisions only when they connect to value. The performance-curve area becomes avoided customer-minutes or energy not served; multiplied by the value of lost load (or sector-specific interruption cost), it becomes avoided outage cost. Around that core sit several valuation methods: probabilistic risk assessment (probability × consequence across the hazard set), cost-benefit analysis comparing program cost against avoided-loss benefit, resilience-adjusted resource adequacy that folds extreme-event risk into adequacy planning, and macroeconomic impact studies that capture the ripple effects of prolonged regional outages. The discipline is to keep the valuation transparent enough that a regulator can reproduce it.

7. Data: the binding constraint

Resilience metrics are only as good as the event data behind them. The framework needs time-stamped, geolocated outage and restoration records, paired with hazard data (weather, flood, fire) and asset/exposure information. The familiar obstacles are incomplete or inconsistent outage reporting, proprietary-data sensitivity, and the rarity of the largest events (small samples for the highest-consequence cases). Public and curated datasets — national outage trackers and curated outage repositories among them — increasingly supplement utility records, but a clean internal outage management system feed remains the most valuable input.

8. The standards and framework landscape

There is no single mandated resilience metric, but the scaffolding is converging. National-lab and DOE efforts — resilience analysis and planning tools, resilience metrics toolkits and rating systems, the Grid Deployment Office metrics framework, and the NREL and PNNL resilience frameworks — offer structured starting points. On the mandatory side, NERC’s TPL-008-1 now requires extreme-weather (benchmark heat and cold) transmission planning studies and mitigation plans, and the recurring FERC/NERC post-event reviews are steadily codifying “what works” into performance expectations. The common gap these frameworks still leave — and where engineering judgment earns its keep — is selecting a coherent, decision-appropriate metric set and grounding it in defensible data.

Reference: key metrics and where they fit

Case Studies

Project identities are confidential

The following studies are drawn from representative Keentel engagements. Client names, exact locations, service territories, and proprietary data have been removed or generalized. Figures are illustrative of the engineering approach and do not disclose any specific client asset or program.

Case Study 1 — Turning Storm-Outage Data into a Defensible Hardening Rate Case

Client

Investor-owned distribution utility (confidential)

Location

Confidential — hurricane-exposed coastal service territory

Challenge

Justify a multi-year hardening + restoration program to the state commission with reproducible benefit-cost evidence

Keentel scope

Outage-data resilience metrics, area outage-rate curves, hardening/restoration ROI, rate-case technical support

Challenge

The utility had a hardening program it believed in but could not defend in dollars. Its SAIDI/SAIFI numbers, with major-event days excluded, made the system look healthy precisely on the storm days that drove customer complaints and the commission’s questions. The commission wanted a reproducible, comparable benefit-cost case before approving cost recovery.

Approach

Reconstructed historical events from time-stamped outage management system records, building cumulative outage and restore curves for each major storm rather than relying on the major-event-day-excluded indices.

Computed performance-curve resilience metrics (trapezoid depth, degraded-plateau duration, restoration rate, and area) for each event, and converted lost performance to customer-minutes and energy not supplied.

Paired outage records with wind data to fit area outage-rate curves — the observed system outage rate as a function of wind speed — establishing the statistical baseline of how the system actually responds to storm stress.

Ran the counterfactual: applied engineering estimates of how targeted undergrounding, pole reinforcement, and pre-staged restoration would shift the outage rate and the restore curve, then re-computed the historical metrics to quantify avoided customer-minutes and ENS per dollar.

Translated avoided ENS into avoided outage cost using sector value-of-lost-load, producing a transparent, reproducible benefit-cost ratio the commission staff could audit.

Outcome

A rate-case exhibit that quantified resilience in the regulator’s own currency — avoided customer-hours and dollars per program dollar — separating the highest-return hardening segments from the marginal ones and giving the utility a defensible, data-grounded basis for cost recovery rather than a narrative argument.

Case Study 2 — A Resilience-by-Design Case for a Large-Load Interconnection

Client

Hyperscale data-center developer (confidential)

Location

Confidential — extreme-weather-exposed interconnection

Challenge

Quantify campus resilience to HILP events and value on-site generation, BESS, and dual-feed options for the interconnection case

Keentel scope

HILP scenario assessment, N–k/cascading screening, performance-curve targets, resilience valuation of mitigation options

Challenge

A large concentrated load was siting into a pocket with real exposure to extreme heat and cold and to cascading transmission events. Standard interconnection studies cleared the routine contingency list, but the developer’s uptime requirements and the host utility’s questions both demanded something the conventional study did not produce: a quantified picture of how the campus would fare during the rare, severe events — and what mitigation was actually worth.

Approach

Defined a high-impact, low-probability scenario set reaching beyond the standard NERC P0–P7 list — benchmark extreme heat/cold consistent with the new TPL-008-1 direction, multi-element N–k transmission losses, and prolonged degraded-supply conditions.

Screened cascading and interdependency pathways (shared transmission corridors, fuel and water dependencies) that could turn a single initiating event into a system-wide one affecting the site.

Built performance-curve (trapezoid) targets for the campus — maximum acceptable drop depth, degraded-state duration, and restoration time for critical load — translating an uptime requirement into engineering acceptance criteria.

Evaluated mitigation options — on-site firm generation, battery storage sizing, controlled islanding/microgrid operation, and a second independent feed — each scored by how much it reduced the performance-curve area under the HILP scenarios.

Valued the options against avoided unserved critical load, giving the developer a ranked, cost-aware resilience menu to take into interconnection and design.

Outcome

A resilience-by-design package that complemented the standard interconnection study with a quantified HILP performance case — letting the developer right-size on-site generation and storage to a defined uptime target and present the host utility a substantiated resilience story rather than an assertion.

Case Study 3 — A Community Microgrid Resilience Scorecard for Critical Facilities

Client

Municipality / critical-facility cluster (confidential)

Location

Confidential — disaster-exposed community

Challenge

Build localized resilience metrics blending technical and social factors to support a microgrid grant and regulatory filing

Keentel scope

Localized resilience metric design, microgrid/DER modeling, critical-load scorecard, funding-application technical support

Challenge

A community wanted to protect its critical facilities — a hospital, water treatment, emergency shelters — through a microgrid, and needed to demonstrate the resilience benefit to win grant and regulatory funding. System-wide grid indices were the wrong tool: they could not capture critical-facility service continuity, islanding performance, or the socio-economic stakes that the funding decision turned on.

Approach

Designed a localized resilience metric set tailored to the community: time to restore each critical facility, microgrid islanding success rate, critical-facility service-continuity hours, speed of reconnection to the grid, and socio-economic/health exposure of the served population.

Modeled the proposed microgrid — distributed generation, battery storage, and provision for mobile storage — and simulated its islanding and critical-load support across representative disaster scenarios.

Computed performance-curve metrics for the critical-load subsystem with and without the microgrid, quantifying the reduction in critical-load outage area for each scenario.

Assembled a transparent scorecard combining the technical metrics with community-specific social factors, structured so reviewers could trace every number to its source — the auditability that grant and regulatory bodies require.

Outcome

A defensible, reproducible resilience scorecard that expressed the microgrid’s value in terms the funder cared about — critical-facility hours kept alive and islanding reliability — rather than generic grid indices, strengthening the community’s case for resilience investment.

About Keentel Engineering

Keentel Engineering is a power systems and grid-interconnection firm serving utilities, developers, and large-load customers from offices in Tampa, Florida and Austin, Texas. Our work spans power system studies across EHV, HV, and MV networks; point-of-interconnection engineering for utility-scale renewables, storage, and large loads ; substation and transmission-line design; protection and controls; owner’s engineer services; and NERC operations & planning compliance.

On resilience engagements we cover the full analytical scope: outage-data resilience metrics and performance-curve analysis, model-based and statistical vulnerability modeling, HILP and N–k scenario assessment, resilience valuation and benefit-cost analysis, and rate-case and grant technical support — for transmission, distribution, large-load interconnection, and community microgrid applications.

Disclaimers

Independence and non-affiliation. Keentel Engineering is an independent engineering firm and is not affiliated with, sponsored by, or endorsed by IEEE, NERC, FERC, DOE, NREL, PNNL, any national laboratory, or any regional transmission organization or independent system operator. References to standards (including NERC TPL-008-1 and IEEE 1366) and to published resilience frameworks are made solely to identify the governing technical context. All trademarks and standard designations are the property of their respective owners.

Confidentiality. All project examples in this paper are anonymized. No client names, project identifiers, exact locations, service territories, or proprietary data are disclosed. Figures and outcomes are illustrative of engineering approach only.

No engineering or regulatory advice. This paper is general technical information, not project-specific engineering, financial, or regulatory advice. Resilience assessments and metric selections must be developed for the specific system, hazards, data, and decision context by a qualified engineer, and verified against the applicable standards and jurisdictional requirements.

Frequently Asked Questions

Synchronous Condenser Protection: Complete Guide 2026

Sat, 20 Jun 2026 12:15:17 GMT

May 11, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Technical Deep-Dive: A Protection Philosophy for Synchronous Condensers

Case Study 2 — Dual-Mode Hydro Unit Operating as Generator and Condenser

Challenge

A large concentrated load (data-center class) was interconnecting alongside utility-scale renewables into an electrically weak pocket with low short-circuit ratio. The developer needed condensers to lift system strength enough to certify IBR ride-through and stabilize voltage — but the same weak network, with nearby series compensation and dense inverters, raised sub-synchronous interaction concerns, and a portion of the facility tapped the transmission system through effectively a single path.

Approach

Quantified the system-strength deficit at the point of interconnection (SCR / effective SCR) and sized the condenser fault-current contribution to meet the minimum strength needed for IBR stability and ride-through certification.

Screened sub-synchronous oscillation risk with EMT models of the IBR and series-compensation controls; because the machines carry flywheels (added inertia, larger effective mass), specified torsional monitoring and protection where the study flagged interaction — conventional generator schemes do not detect SSO.

Built the condenser protection on the lean baseline (single-mho LOF, 87, 46, 24, ground-fault, backup, breaker-failure), explicitly omitting 78 / 32 / 81 with documented justification.

Applied inadvertent-energization protection as a control interlock for machines intended to be started in a controlled sequence, preventing human-error across-the-line starts.

Addressed the single-line / weak-system case with anti-islanding and out-of-step logic at the transmission-line terminal — not on the condenser — to trip the whole facility cleanly on separation, since an islanded condenser merely follows the dominant machine.

Outcome

A condenser fleet sized to the measured strength deficit, an SSO screening that turned a latent torsional risk into a monitored-and-protected one, and a protection architecture that places stability/anti-islanding logic where it actually belongs — at the interconnection — giving the developer a defensible interconnection package for a weak-grid, large-load site.

The Synchronous Condenser Is Back and Its Protection Is Not Just “Generator Protection”

Where the condensers are coming from

Two supply paths dominate. The first is the purpose-built machine: a new synchronous condenser, often with a flywheel, specified explicitly for grid strength. The second — frequently the more economical — is conversion. A retired steam turbogenerator, an idle combustion turbine, or a hydro unit with surplus capability can be repurposed into a condenser by decoupling the prime mover. The generator that used to make megawatts becomes a machine that makes vars and fault current instead.

Conversion approaches vary with how often the unit will switch modes. A hydro machine can simply close its wicket gates and spin in air (with a blowdown/air-evacuation system if the runner sits below tailwater). A combustion turbine can declutch its engine after synchronizing. A steam unit is typically decoupled by a clutch for frequent switching, or by a bolted shaft flange when the change is effectively permanent, with the turbine blades sometimes removed entirely to avoid windage losses. Auxiliary systems — lube oil, turning gear, stator cooling — often must be retained, and the unit needs a starting means (static starter or pony motor) because there is no longer a prime mover to bring it up to speed.

So why isn’t the protection just “copy the generator settings”?

Because one physical fact changes the dynamics of the whole machine: a synchronous condenser has no source of mechanical input power. It does not export real power. In steady state it actually draws a small amount of real power from the system to overcome friction and windage — it lives almost entirely on the reactive (vertical) axis of the capability curve, slightly into the motoring region. Electrically, the stator winding, rotor, field winding, and excitation system look just like a generator’s, so many protection functions carry over unchanged. But the functions that depend on mechanical input power — and there are several — behave very differently, and a few generator protections become unnecessary or even counterproductive.

The one-sentence version

Treat the stator, rotor, field, and ground-fault protection like a generator—but rethink loss of field, loss of synchronism, inadvertent motoring, and frequency protection from first principles, because a machine with no accelerating power source cannot misbehave the way a generator does.

The headline example is loss of field. On a generator, the apparent-impedance trajectory after a loss-of-excitation event depends on how much real power the unit was carrying, which is why conventional schemes often use two mho zones to stay both secure and dependable across loading conditions. A condenser carries no real power, so its loss-of-field signature is a single, predictable impedance swing — and a single mho element, set from the machine’s own reactances, is sufficient. Simpler, and more secure.

The mirror-image example is pole slipping. A generator slips a pole when a fault traps mechanical energy in the rotor and the rotor angle runs past 180 degrees. A synchronous condenser has no prime mover pushing the shaft, so there is no accelerating energy to store. It can lose voltage synchronism, but transient and EMT studies consistently show the rotor angle oscillating within a bounded window — nowhere near a pole slip — so dedicated out-of-step protection is usually unnecessary for the condenser itself.

Get these distinctions right and you end up with a protection package that is leaner, more secure against misoperation, and fully defensible in a NERC PRC review. Get them wrong — by porting a generator template wholesale — and you risk nuisance trips of an asset whose entire job is to stay connected and hold the grid up during the exact disturbances that would trip it. The sections that follow lay out the engineering in detail, walk through three anonymized projects, and answer the questions we hear most often.

Challenge

A unit nearing the end of its life as a generator was structurally sound electrically. The utility wanted to retain its fault-current and inertia contribution in a corner of the system where retirements had eroded short-circuit strength . The turbine was to be permanently decoupled by a bolted shaft flange, with a static starting system bringing the machine to synchronous speed. The legacy generator protection — including dual-zone loss of field, reverse power, and frequency elements — could not simply be carried forward.

Approach

Re-scoped the protection from first principles: retained stator differential (87), stator and field ground-fault, field overcurrent, negative-sequence (46), V/Hz (24), backup distance (21) and voltage-restrained overcurrent (51V), and breaker-failure — implemented across two independent multifunction relays for redundancy.

Replaced the dual-zone loss-of-field scheme with a single negative-offset mho element (diameter = Xₑ, offset = X′ₑ/2, ~0.5 s delay), verified against the underexcited capability edge, the SSSL, and the underexcitation limiter.

Removed inadvertent-motoring (32) and abnormal-frequency (81) protection as inapplicable; documented the engineering basis for each removal to support the compliance file and future audits.

Retained an inadvertent-energization scheme because the decoupled machine was not rated for an across-the-line induction start, with interlocks disabled during the static-start sequence.

Built the PRC-019-2 coordination plots (LOF vs UEL/SSSL, V/Hz relay vs excitation limiter) and PRC-024-4 ride-through documentation in the same pass.

Outcome

A leaner, more secure protection package with two fewer trip paths than the ported generator template, a single defensible loss-of-field circle, and a complete NERC PRC evidence file delivered with the settings — reducing nuisance-trip exposure on an asset whose value is measured by how reliably it stays connected during disturbances.

This section is written for the protection engineer who has to specify relays, calculate settings, and defend them. It maps each function against synchronous-condenser physics and flags where the condenser departs from IEEE C37.102 generator practice. The governing references throughout are IEEE C37.102 (AC generator protection), C37.101 (generator ground protection), C37.106 (abnormal frequency), C37.119 (breaker-failure), C37.2 (device function numbers), and C50.12/C50.13 (machine ratings), together with NERC PRC-019-2 and PRC-024-4 for compliance.

1. The machine, in three properties

No mechanical input power. There is no prime mover delivering torque to the shaft, so there is no accelerating-power source. The condenser absorbs a small real power (friction and windage) and otherwise operates on the reactive axis — overexcited to supply vars, underexcited to absorb them.

Electrically identical to a generator. Same stator winding, same rotor and field winding, same excitation system and limits. Stator, field, and ground-fault protection transfer over directly.

A strong fault-current source. Behind subtransient, then transient, then synchronous reactance, the machine delivers substantial short-circuit current independent of whether anything is coupled to the shaft. This is a feature — it is much of the reason the condenser was installed — but it must be respected in backup and breaker-failure design.

2. Functions that transfer over from generator practice

These follow C37.102/C37.101 essentially unchanged. Dual (redundant) protection systems are recommended for dependability, typically two independent multifunction generator/motor relays. Relay manufacturers have not historically built a dedicated condenser relay, but standard generator or synchronous-motor multifunction relays implement everything needed.

3. Loss of field (40): the defining difference

On a synchronous generator, the post-event apparent impedance depends on pre-fault MW loading, which spreads the trajectory across the R-X plane and motivates dual-zone mho schemes. A condenser carries no MW, so its behavior is singular and predictable.

Map P-Q to R-X. The terminal apparent impedance relates to output by R = V²P/(P²+Q²) and X = V²Q/(P²+Q²). Producing vars places the impedance in the second quadrant; absorbing vars places it in the third. At zero var and a small incoming MW, the impedance sits far out along the negative-R axis — for a 20 kV machine drawing roughly 1 MW at zero var, on the order of –400 Ω.

The swing. When excitation is lost, var output decays and the machine drifts toward induction-motor behavior. Flux collapses, the machine transitions through a purely-resistive point, then begins absorbing vars and the locus migrates toward the third quadrant. Plotted on the loss-of-field scale, the trajectory traces a large, single circular swing — there is no second family of trajectories to guard against.

Recommended single-mho setting (negative-offset method, C37.102 Method 1)

Diameter = Xₑ (synchronous reactance), scaled to secondary ohms by CTR/VTR.

Offset = X′ₑ / 2 (half the transient reactance), the same negative offset used for generators.

Time delay = 0.5 to 0.6 s, which initiates shutdown well inside the rotor and end-iron withstand times for a condenser-scale slip.

A second zone is unnecessary for a dedicated condenser because the final trajectory approaches Xₑ. Coordinate the circle against the underexcited edge of the capability curve, the steady-state stability limit (SSSL), and the underexcitation limiter (UEL).

Worked illustration: for a 13.8 kV, 330 MVA class machine with Xₑ ≈ 1.71 pu and X′ₑ ≈ 0.43 pu (Z-base ≈ 0.577 Ω, CTR 3000:1, VTR 120:1), the diameter works out near 25 Ω secondary and the offset near 3 Ω secondary — a clean, single circle. The calculation method is identical to a generator; only the second zone is dropped.

4. Functions that still matter — and need extra care

Negative-sequence / unbalance (46)

The condenser has the same rotor thermal vulnerability to negative-sequence current (the I₂²t limit) as a generator, even though it carries no balanced MW load. The watch item is the grid trend: as IBR penetration rises, some networks adopt slower fault-clearing schemes. Longer clearing times mean longer negative-sequence exposure for the condenser rotor during unbalanced grid faults, so coordinate the relay’s short-time capability against the actual (possibly extended) clearing times rather than legacy assumptions.

Stator overflux / V/Hz (24)

Flux is proportional to voltage and inversely proportional to frequency. A condenser’s excitation system can push terminal voltage above 1.0 pu through field forcing, overexciting the core even at synchronous speed on a stable-frequency system. Apply a V/Hz or phase-overvoltage scheme per C37.102/C37.106, and coordinate it with the excitation V/Hz limiter for PRC-019 compliance.

Inadvertent energization (50/27) and out-of-phase synchronization

Inadvertent energization (closing onto a de-excited, stationary machine) drives the unit as an induction motor. A condenser’s unloaded rotor demands less starting torque and current than a loaded motor, but the machine may still not tolerate an across-the-line start unless it was designed for one. Default to applying the scheme unless the OEM confirms the machine can withstand an unexpected across-the-line start; for machines that are intentionally induction-started, the scheme can instead be implemented as a supervisory control interlock that is disabled during a deliberate start. For out-of-phase synchronizing, the condenser is subject to the same electromagnetic/torque constraints as a generator because it has an internal voltage source — full sync-check supervision (magnitude, angle, slip) is warranted unless an across-the-line start sequence is used.

Sub-synchronous oscillations (SSO)

A multi-mass shaft near series capacitor compensation, an HVDC converter, or dense IBR can exchange energy at sub-synchronous frequencies with the shaft’s torsional modes. The risk is small for a compact condenser with a pony motor and brushless exciter, but rises for larger machines carrying a flywheel’s additional inertia. Where EMT studies of the IBR/HVDC controls flag interaction, add torsional monitoring and protection alongside any mitigation. Conventional generator schemes will not detect SSO.

5. Functions that are typically not needed — and why

Loss of synchronism / out-of-step (78)

A condenser cannot slip a pole. With no prime mover there is no accelerating power, and a flywheel is only a passive load. It can lose voltage synchronism if its internal voltage magnitude or angle drifts past the system, but transient and EMT simulations show bounded rotor-angle oscillation (on the order of tens of degrees for nearby three-phase faults), with the impedance locus oscillating near the X-axis and never reaching the electrical center. If an application still warrants the function, use a security-biased single-blinder scheme.

Inadvertent motoring / reverse power (32)

The condenser is designed to motor — it absorbs a little real power by definition, and there is no prime mover to protect from motoring. The function is therefore not required to protect the machine. One useful exception: on a hydro unit that naturally motors, a reverse-power element can still serve as a breaker-failure detector if power keeps flowing in after a shutdown command.

Abnormal frequency (81) and stator overcurrent (51)

Frequency protection guards prime-mover blades against resonance and supports load-rejection / anti-islanding logic that hinges on real-power balance. A condenser has no blades and no real-power role, so neither rationale applies. Stator overcurrent is generally unnecessary too:

condensers are built with large overload capability per C50.12/C50.13 and, lacking a prime mover, will not carry sustained balanced three-phase overload.

6. Special case: a single interconnecting line, or a weak system

The “condenser can’t slip a pole” conclusion is about the condenser in isolation. When a much larger synchronous generator shares the bus and slips poles, the condenser’s voltages, currents, and impedance locus get badly perturbed — it simply follows the dominant machine. The sharper risk is a facility tied to the transmission system through a single line, or into a weak system: once islanded, the condenser tracks whatever else is in the island. The mitigation is not more condenser protection but anti-islanding / out-of-step logic at the line terminal that detects separation and trips every unit in the facility.

7. NERC compliance: yes, the PRC standards apply

PRC-019-2 applies to individual synchronous condensers above 20 MVA directly connected to the Bulk Electric System. It requires coordination, reviewed at least every five years, of the voltage-regulating controls and in-service limiters (UEL, OEL, V/Hz limiter) with the machine’s capability and the protection settings — most visibly the loss-of-field mho against the UEL and SSSL, and the V/Hz relay against the excitation V/Hz limiter. PRC-024-4 — updated in 2024 — now explicitly names synchronous condensers: frequency and voltage protection settings must sit outside the no-trip ride-through envelopes, because the condenser is expected to ride through exactly the excursions it was installed to help damp. PRC-025-2 (relay loadability) and PRC-026 (relay performance during stable power swings) round out the review.

Keentel’s protection settings and NERC O&P compliance teams build the LOF / UEL / SSSL / V/Hz coordination package and the PRC-024-4 ride-through documentation as part of the same

Case Studies

Project identities are confidential

The following studies are drawn from representative Keentel engagements. Client names, exact locations, one-line identifiers, and proprietary equipment data have been removed or generalized. Machine ratings and findings are illustrative of the engineering approach and do not disclose any specific client asset.

Challenge

The owner operated a hydro unit that needed to provide active power when water was available and reactive support / system strength when it was not (low water, or off-peak). Closing the wicket gates lets the runner spin in air; an air-evacuation system holds water below the runner. Because the same machine operates in both modes, the protection had to be valid across two very different capability footprints — a generator carrying MW, and a condenser on the reactive axis — without compromising security in either.

Approach

Specified a two-zone loss-of-field scheme: the unit genuinely needs both zones because it loads up as a generator in one mode, while a single larger zone would suffice if it were condense-only. We sized both zones and verified the condense-mode trajectory against the smaller zone.

Ran a dynamic loss-of-field simulation in condense mode (machine supplying vars, field breaker opened) and confirmed the rotor, stator-winding, and stator end-iron withstand times all exceeded the mho-plus-delay trip time — i.e., the negative-offset mho protects without exposing the machine to asynchronous-operation damage.

Coordinated negative-sequence (46) protection against the rotor I₂²t capability, explicitly accounting for the longer fault-clearing times trending into the surrounding IBR-rich network so the rotor is protected during extended unbalanced grid faults.

Applied a reverse-power element not as anti-motoring protection (the hydro unit naturally motors a few tenths of a percent in condense mode) but as a breaker-failure detector — flagging real power still flowing into the machine after a shutdown command.

Outcome

A single relay configuration valid in both operating modes, with simulation-backed evidence that the loss-of-field scheme trips inside the machine’s thermal withstand and that negative-sequence settings hold up under the grid’s evolving (slower) clearing times — protection that follows the machine rather than fighting its mode changes.

Case Study 3 — New-Build Condensers for Grid Strength at a Large-Load / Renewable Hub

Keentel Engineering is a power systems and grid-interconnection firm serving utilities, developers, and large-load customers from offices in Tampa, Florida and Austin, Texas. Our work spans power system studies across EHV, HV, and MV networks; point-of-interconnection engineering for utility-scale renewables, storage, and large loads; substation and transmission-line design; protection and controls; owner’s engineer services; and NERC operations & planning compliance.

On synchronous condenser projects we cover the full electrical scope: feasibility and conversion engineering, system-strength and short-circuit studies, protection philosophy and settings, EMT / sub-synchronous screening, and the PRC-019-2 / PRC-024-4 compliance package — whether the machine is a purpose-bui lt condenser or a retired generator being given a second life as grid-strength infrastructure.

Disclaimers

Independence and non-affiliation.

Keentel Engineering is an independent engineering firm and is not affiliated with, sponsored by, or endorsed by IEEE, NERC, any regional transmission organization or independent system operator, or any equipment manufacturer. References to IEEE standards (including C37.102, C37.101, C37.106, C37.119, C37.2, and C50.12/C50.13) and to NERC Reliability Standards (including PRC-019-2, PRC-024-4, PRC-025-2, and PRC-026) are made solely to identify the governing technical framework. All trademarks and standard designations are the property of their respective owners.

Confidentiality.

All project examples in this paper are anonymized. No client names, project identifiers, exact locations, or proprietary equipment data are disclosed. Machine ratings and outcomes are illustrative of engineering approach only.

No engineering advice.

This paper is general technical information, not project-specific engineering advice. Protection settings must be developed for the specific machine, grounding, and interconnection by a qualified engineer, in coordination with the equipment manufacturer and the interconnecting entity, and verified against the applicable standards and the original-equipment-manufacturer documentation.

For most of the last half-century, the synchronous condenser was treated as a relic. As static reactive devices such as SVCs and STATCOMs matured, many utilities retired their spinning condensers and never looked back. That decision is being reversed across North America right now, and the reason is structural: the grid is losing the very machines that quietly held it together.

As coal and older gas plants retire and inverter-based resources (IBR) — utility-scale solar, wind, and battery storage — take their place, two physical quantities decline that no inverter fully replaces: short-circuit strength and rotational inertia. Inverters follow the grid; they do not, on their own, anchor it. Where penetration is high and the network is electrically weak, the symptoms are familiar to anyone who has run an interconnection study lately: low short-circuit ratio (SCR) at the point of interconnection, control interactions and sub-synchronous oscillations, sluggish voltage recovery after faults, and ride-through performance that is difficult to certify.

Layered on top of that is the largest load-growth event in a generation. Hyperscale data centers and AI campuses are arriving as hundreds of megawatts of fast, concentrated, power-electronic demand — and they want to interconnect into precisely the weak pockets where system strength is already thin. Equipment vendors have noticed; several have recently launched synchronous condenser packages aimed squarely at stabilizing the grid around large data-center loads. The condenser is no longer a museum piece. It is a strategic asset.

A synchronous condenser delivers three things an inverter cannot replicate on demand: a robust, fault-current source that raises local system strength; genuine rotational inertia from a spinning mass that slows the rate-of-change-of-frequency during disturbances; and smooth, continuous dynamic voltage support whose reactive output does not collapse with terminal voltage the way some static devices do. Add a flywheel and the inertia contribution grows further.

Case Study 1 — Retired Steam Turbogenerator Converted to a Dedicated Synchronous Condenser

About Keentel Engineering

Frequently Asked Questions

PSCAD Black Box: Protect EMT Models & Control IP

Sat, 20 Jun 2026 09:42:50 GMT

May 20, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Where Black Box keeps a model inside PSCAD but hides it, the Independent C Code feature does the opposite—it takes a control module out of PSCAD entirely. It translates a PSCAD control module into valid, standalone C that compiles separately from PSCAD for any purpose, including running directly on a microcontroller. For teams that prototype control logic in simulation and then need it on real hardware—hardware-in-the-loop rigs, embedded controllers, digital twins—this closes the gap between the model and the device.

A concrete example

Take a small module with two inputs and one output that computes the running ramp-up over time of sin(input 1) × input 2, added to its previous value, with a parameter to reset the accumulated sum. In simulation, the inputs come from a function of time and slider values—a Reset Time slider sets when the sum resets, and a Gain slider sets the multiplier. Running the case shows the output responding to those inputs. The goal is to lift exactly this logic out of PSCAD and run it elsewhere.

Generating the code and choosing data types

Right-click the module and choose Generate → Independent C Code. A dialog lets you pick the data types the generated C will use—important because target hardware often supports only certain types. You configure three families:

How Keentel Applies These Features

Electromagnetic transient (EMT) models have become the currency of modern interconnection. Every time an inverter-based resource, an HVDC link, a STATCOM, or a large electronic load is brought to a point of interconnection, the studies that decide whether it can connect—and under what conditions—increasingly hinge on EMT-domain analysis performed in PSCAD™/EMTDC™. That shift has created two practical problems that have very little to do with the physics of the model and everything to do with how the model is handled: how do you share a model without giving away the proprietary control logic inside it, and how do you take control logic you have validated in simulation and put it to work outside the simulator entirely?

PSCAD’s V5.1.0 release answers both questions directly. The Black Box feature lets you collapse an entire module hierarchy into a single, compiled, non-module component—preserving exact behavior while hiding the source. The Independent C Code feature (often called the C Coder) does the inverse for controls: it translates a control module into portable, standalone C that compiles outside PSCAD, including onto bare-metal hardware. At Keentel Engineering, we work with both sides of this exchange every week—receiving obfuscated vendor models on the study side, and helping clients package and port their own logic on the development side. This guide explains what each feature actually does, where the sharp edges are, and how we use them in interconnection-grade work.

Why this matters for interconnection

ISOs and transmission owners increasingly require validated EMT models for IBR and large-load studies, and OEMs are understandably unwilling to expose their control IP. Black Box is the mechanism that squares that circle—letting a developer hand over a model that runs identically without revealing the schematic underneath. Understanding exactly what an obfuscated model still exposes , and what it does not, is a core competency for anyone running or reviewing those studies.

Need to package an EMT model for interconnection submission, review an obfuscated vendor model, or port control logic to hardware?

Keentel Engineering provides power-system studies, POI interconnection engineering, EMT modeling , and owner’s-engineer services across utility-scale renewables, BESS, and large-load projects. Reach us at keentel.com to discuss your model-delivery or controls-export workflow.

About Keentel Engineering & Disclaimers

PSCAD™ V5.1.0 Black Box and Independent C Code

A Field Guide for Interconnection-Grade EMT Work

Part 1 — The Black Box Feature: Obfuscation Without Compromise

The Black Box algorithm automates the complete conversion of a PSCAD page-module hierarchy into a user-defined component. It is not a cosmetic wrapper; it regenerates the model as compiled machine code and rebuilds the component around it. Three things happen under the hood, and it is worth understanding each because they determine what the recipient of a black-boxed model can and cannot see.

What the algorithm actually produces

1. Fortran source generation

Black Box emits Fortran source formatted specifically to run with EMTDC as an external file. This is the same fundamental code path EMTDC uses internally, but the algorithm formats it with the subroutine structure needed for multiple-instance compatibility—so the resulting component can be dropped into a schematic more than once without symbol collisions.

2. Automatic object or library file creation

Optionally, the generated source is compiled into an object file. If the module being black-boxed contains nested modules, every resulting object file is bound together into a single static library. This compiled artifact is where the real protection lives: once the logic is machine code in a .lib, the original source is no longer recoverable from the deliverable.

3. Automatic component creation

Finally, a component definition and instance are built from the contents of the hierarchy—automatically carrying over every port, parameter, graphic, and relevant script segment. The end result looks and connects like any native PSCAD component, but its internals are sealed.

Two restrictions you must plan around

Runtime objects are not supported. Switches, sliders, dials, and radio links cannot survive the Black Box operation, because they imply live user interaction that a compiled component cannot expose in the same way. Before you run the tool, you have to deal with them, and you have exactly two choices:

Bake in a fixed value. Replace the dial or slider inside the module with a constant tag set to a preset value. This is clean and simple, but it permanently locks that value into the component—there is no real-time variability left once it is sealed.

Preserve real-time control. Move the control or radio link out of the module entirely. If the object is buried deep in the hierarchy, you must promote its signal upward through every intermediate level using ports or parameters until it reaches the top. The control then lives outside the schematic and drives the finished Black Box component as an external input—retaining live adjustability.

Keentel practice note

Decide the bake-in-vs-promote question before you build the hierarchy, not after. Promoting a deeply nested radio link through five module levels late in a project is tedious and error-prone. If a parameter needs to remain operator-adjustable in the delivered model—a fault-ride-through enable, a control mode selector, a gain trim—design its signal path to the top level from the start.

Transmission lines and cables: new in V5.1.0, and worth understanding

V5.1.0 extended Black Box to systems that include transmission lines and cables—a meaningful addition, because line and cable models carry solved segment data that has to go somewhere. The algorithm embeds the solved transmission-segment information inside the automatically generated subsystem script. The tower and cable configurations themselves stay fully hidden, but a portion of the solved data remains visible, and exactly how much depends on the line model you chose.

This distinction matters for IP protection, so it is worth being concrete about what leaks through each model type:

The takeaway is straightforward: if hiding the transmission-segment design is a priority, the frequency-dependent phase model exposes far less usable information than Bergeron. We factor this into how we advise clients on model packaging when line topology is itself considered sensitive.

Inside the Options panel: the settings that actually change your model

When you generate a Black Box, an Options dialog appears. You can also set these preferences globally in PSCAD’s main application options, where they save to your user profile—worthwhile if you black-box frequently. Most settings are self-explanatory; the ones below are the ones that change behavior, results, or protection, and they are the ones we pay attention to.

Code Script Usage Compatibility

This ensures the component’s internal definition script is compatible with a target PSCAD version. The critical caveat: it affects only the definition script, not the generated Fortran source. If you build a Black Box in 5.1 using subroutines from the 5.1 master library, setting compatibility back to 4.5 will not make it run in 4.5—that older version simply lacks the newer EMTDC master-library models. Compatibility flags do not back-port capability.

Feed Forward Signal Storage

Leave this on Force Disabled unless you have a specific reason not to. Choosing Assume Current Project Setting inherits the project’s “Store Feed Forward Signals for Viewing” state; if that is enabled, every feed-forward signal is read from and written to storage on every time step inside the generated Fortran. That can hurt code efficiency and simulation speed. However, there is a real failure mode here: some custom models read and write storage improperly, and Force Disabled can change simulation results once the Black Box replaces the original module.

A quick diagnostic before you seal a module

Run the project once with “Store Feed Forward Signals for Viewing” enabled, then again disabled (or vice-versa). If the simulation results change between the two runs, the module is storage-sensitive—select Assume Current Project Setting before you Black Box, then re-run the tool. This five-minute check routinely prevents a black-boxed component from quietly behaving differently than the module it replaced.

Inner component definitions — a bottom-up algorithm

Black Box processes a hierarchy from the bottom up. Picture a top module Top containing A, B, and C; A contains D; C contains E and F; and E contains G. The algorithm walks upward like this:

Black-box G first; its instance inside E is replaced by the new component.
Move up to E and F; both are processed and replaced inside parent C.
Black-box D, replacing its instance inside A.
Black-box the intermediate modules A, B, and C, replacing each inside Top.
Finally, black-box Top itself.

The Inner Component Definitions option decides what happens to those intermediate definitions (A through G). Keep them all if you want to reuse the individual children later; otherwise set Do Not Retain and PSCAD deletes the intermediates once the top-level operation finishes, keeping your project tree clean.

Component instance, parameters, and electrical ports

Component Instance Creation — choose Create New Instance to attach the finished component to your cursor for placement, or Generate Definition Only to build it quietly in the project tree for later manual placement.

Parameters — Obtain from Source Module Instance copies every preset value from the original module onto the new component (so a module with 20 custom-set parameters transfers seamlessly and runs immediately). The alternative reverts each parameter to its original default.

Fixed and Switched Electrical Port Names — internal electrical nodes are exposed as fixed and switched ports. Ten fixed nodes, for example, create a ten-element fixed-port array (NBBF). You may rename these internal ports for clarity, or leave the defaults.

Definition Labels — comma-separated labels that categorize the component on the Models ribbon tab and integrate with the V5.1 Library Viewer pane—useful for keeping a growing library of custom tools organized.

Global substitutions

If your hierarchy calls global substitutions, you choose how they are handled:

Embed Values — bakes the current substitution values directly into the component as literals. Once sealed, they are permanently locked.

Make Accessible as Parameters — converts them into component parameters under a new “Global Substitutions” category in the component properties dialog, so you can keep adjusting them exactly as before.

Passive Elements Obfuscation — the heart of circuit protection

This setting is the one that most directly hides proprietary circuit design. Enabling Replace with Runtime Configurable Passive Branch tells PSCAD to use a master-library component model to secure your data: the algorithm strips the specific R, L, and C values out of each electrical branch and moves them into the Fortran file as literal values. Because that Fortran is then compiled into an object or library file, the true values are completely protected. In the final compiled data file, the real component values are replaced by dummy values. The recipient can still see that an RL branch exists in the graphic—but the actual values are hidden.

Transmission segment names, resources, and compilation

Transmission Segment Names — replaces real line and cable names with generic placeholders such as Tline1 or Cable1 to further obscure proprietary topology. The base prefix is customizable (default: “Segment”).

Resource Folders — if unspecified, generated files land in the active project folder and clutter it fast. Point this to a dedicated directory (for example a Blackbox folder under your local temp path) to keep generated output organized.

Binary Files Compilation — generate source only, compile with your current default compiler, or generate objects for every installed compiler at once. Because object, library, and .o files are bound to the exact compiler that produced them, building for all installed compilers is smart practice if you switch compilers regularly.

Append Resource Files — automatically attaches generated binaries and source to the project’s resources branch, so you can black-box a module, swap it into the schematic, and run immediately. Choose No Action to skip the automation.

Clear Existing Resources — a maintenance utility that purges the resources branch before appending new files, preventing a pile-up of stale libraries if you black-box often.

Naming Prefix — a strict requirement. The prefix is prepended to generated subroutines so they stay globally unique. Without it, two libraries that each define a subroutine named, say, “x” will trigger a multiple-source error and the Fortran compile will fail. A distinct prefix removes that risk entirely.

Walkthrough and verification: trust, then verify

Consider a simple RectifierAC page module containing a basic source and an RL load. Right-click, choose Generate Black Box, set a destination folder, and confirm. One best practice governs everything here:

Compile clean first

Black Box uses the exact same netlist compiler PSCAD uses to build a project. So compile the project normally and resolve every existing error before you black-box anything. Sealing a module that already has build errors only drags those errors into the generation process.

After the operation, the destination directory holds a new folder named for the original definition module. Inside is a Fortran source file with the Black Box name appended—structured like a standard EMTDC-generated file but formatted with the subroutines needed for multiple-instance use. The libs folder contains compiler-dependent output: with GFortran, for instance, an object file is produced and automatically bound into a static .lib, which PSCAD labels with the compiler name for later compatibility.

The finished component is where verification happens. Its graphic shows a three-element port array, matching the three fixed nodes in the original module. The underlying script generates the DSDYN and DSOUT calls into the compiled rectifier subroutine and places all branch and node numbers into storage. The original network’s nine branches each appear as a single line in the branch segment—and because passive-branch hiding was enabled, dummy values stand in for the real R, L, and C. With no transformers or model-data components present, those segments are simply empty. The component behaves exactly as the module did, while its design is sealed.

Part 2 — The Independent C Code Feature: Controls Beyond the Simulator

The complex-number options exist because many C compilers do not support complex types at all (hence No Complex ), some support the C99 standard, and others—the Microsoft Visual C compiler among them—implement complex values through a structure rather than a native type. Once the types are set, click OK to generate.

What gets generated, and where

The output lands in the case’s current temporary build directory, organized into three folders:

Libs — contains pscadlib.h and pscadlib.c, holding source for all C-export-enabled PSCAD components, with every data-type version of each needed function. The source is provided so you can modify or strip out code you don’t need for a particular compiler.
Headers — auto-generated header files for calling the module code—one per module compiled (including sub-modules), each named to match its module-definition source.
Source — the C implementation for each module, again one file per module, named to match the de finition.

Using the generated code

A typical integration—say, a Visual Studio 2022 program that reads input from standard-in and writes output to standard-out—follows a consistent pattern. Include every header and source file from the generated output in your project, then wire up the call:

Include the main header. Its name is the exported module-definition name with a .h extension.
Declare a void pointer, initialized to null, for application memory. Even though it is a pointer, memory is statically allocated by default—so the code runs on hardware that forbids dynamic allocation. The pointer is bound to the static storage structure the first time the code is called.
Use the module’s I/O structure. Imported from the module header, its name is the module-definition name followed by _main; it carries inputs in and outputs out.
Assign inputs, then call the run function. Its name is always the module-definition name followed by _run. It takes a pointer to the void application-memory pointer, an integer run-type, and a pointer to the I/O structure.
Read the output from the I/O structure, then loop. On the first call, a null memory pointer triggers allocation, value initialization, and the first time step.

The run-type integer takes one of three values:

When the end time is not known in advance—as in a console program that runs until it reads a “Q”—you typically use only 0 during the loop and a single 2 at shutdown to release resources cleanly. (Passing 1 instead would run one final time step on the current I/O before cleanup.) By default the generated algorithm runs in real time , querying the system clock on each call—worth remembering when you benchmark or when deterministic timing matters.

Static versus dynamic allocation

Static allocation is the safe default for constrained hardware, but some situations call for dynamic memory—most commonly when you need multiple instances of the generated code running at once, or when the target allows dynamic allocation but caps static allocation. A small, well-defined edit to the generated source switches modes:

In the run function, remove the statically allocated structure (named application_memory) declared just above it.
Replace the two uses of that application memory with a call to malloc, passing the size of the structure.
Add a call to free for the pstate variable in the two places where it is set to null, before it is nulled.

With those changes, the module runs on dynamic memory and multiple instances can coexist in one application. Document the edit—because it is a manual modification to generated code, it must be re-applied whenever the module is regenerated.

Limitations: know what will and won’t export

The C Coder supports a deliberately bounded set of components. Two conditions must hold: the component must have no electrical aspect (it must be purely controls-based), and the mathematics behind it must not be proprietary to PSCAD. Within those bounds:

The unsupported components are excluded either because their implementation is proprietary or because they need additional detail that cannot be expressed generically. To check any component yourself, open its definition, look at the Scripts section, drop down the current-segment box, and select C. If a C script section exists, the component is C-Coder-ready.

Extending the C Coder with custom components

If a module includes a custom component you need to export, you can make it C-Coder-compatible by giving it a C script section. Edit the component definition, open the Segment Manager, add and highlight the C section, then switch the current segment to C. The code you place there is written verbatim by the C Coder, so it must be valid C—semicolons and all—outside of directives and substitutions. It is auto-indented when written to the C files.

The C script section uses the same control directives as other script sections, plus a few specific to C export:

Control flow & structure: if / else / elseif / case, and begin, endbegin, top, and bottom—which retain their meaning from the Fortran script sections.

Function placement: dsdyn / enddsdyn force code into the DSDYN function; dsout / enddsout force it into the DSOUT function.

Includes: header declares a C #include for a header file; include pulls in script-section code (with its own directives and substitutions) from an external source.

local — defines a local variable. It works much like the Fortran version, except the variable must be accessed using dollar-sign ($) substitution.

stash — defines a variable that persists across the program’s execution—the C-export equivalent of the storage arrays used in Fortran sections. Stashed variables are also accessed with $ substitution.

struct & symbol — both local and stash accept a struct keyword plus a structure name (you define the structure in the included C code), and symbol defines your own substitution for use within the section.

Once the C script section is defined, regenerate the independent C code; you can inspect the result directly in the generated module .c file. The component is then a first-class citizen of the C Coder workflow.

These are not abstract capabilities for us. On the study side, we routinely receive black-boxed EMT models from equipment vendors and developers, and a real part of model-quality review is knowing precisely what an obfuscated model still reveals—line lengths, phase counts, the presence (but not the values) of passive branches—and what it has genuinely sealed. That knowledge shapes how we validate a model we did not build and how we communicate residual exposure to a client who is handing their IP to a third party.

On the development side, we help clients package their own intellectual property for distribution to ISOs, transmission owners, and study consultants without surrendering control design. And when validated control logic needs to leave the simulator—for hardware-in-the-loop testing, embedded deployment, or a digital twin running alongside the physical asset—the C Coder path lets that logic move from PSCAD to silicon with its behavior intact. Used together, Black Box and Independent C Code turn a PSCAD schematic into something you can protect, deliver, and deploy.

Talk to Keentel

Technical FAQ

Keentel Engineering is a power systems and grid-interconnection firm with offices in Tampa, Florida and Austin, Texas. Our services span power system studies (EHV/HV/MV), point-of-interconnection engineering, substation and transmission-line design, utility-scale renewables and BESS engineering, owner’s-engineer services, and NERC O&P compliance, with deep EMT-modeling capability across interconnection workflows.

Non-affiliation and trademark notice

This document is an independent technical commentary prepared by Keentel Engineering for educational purposes. Keentel Engineering is not affiliated with, authorized by, sponsored by, or endorsed by Manitoba Hydro International Ltd. PSCAD™ and EMTDC™ are trademarks of Manitoba Hydro International Ltd. All product names, features, and trademarks referenced herein are the property of their respective owners and are used for identification and descriptive purposes only.

Feature descriptions reflect PSCAD V5.1.0 as understood at the time of writing. Software behavior may change across versions; always confirm against the official PSCAD Application Help (including the “Black Boxing Modules” section) and the vendor’s current documentation. For software support, refer to the official PSCAD support channels. This material does not constitute professional engineering advice for any specific project; engagement of a licensed engineer is recommended for project-specific application.

Common questions we field on PSCAD V5.1.0 Black Box and Independent C Code, answered from a study-and-delivery perspective.

Black Box

Independent C Code (C Coder)

Transmission Line Modeling PSCAD: Complete Guide

Sat, 20 Jun 2026 07:14:57 GMT

Jun 20, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Building the Line, Step by Step

Validating with the Ferranti Effect

A practical guide to building a distributed line model in PSCAD—choosing between PI, Bergeron, and frequency-dependent representations, configuring tower geometry, and validating no-load voltage rise.

Transmission-line modeling sits underneath almost everything we do in EMT-domain work. The same disciplined workflow—right model for the phenomenon, correct tower geometry, name-matched interfaces, a firm source, and a deliberate validation case—supports our insulation-coordination and TOV studies, switching and energization analysis, shunt-reactor sizing, protection and relay studies, and the point-of-interconnection EMT studies that increasingly govern renewable, BESS, and large-load projects. A line model is only as good as the choices behind it, and those choices are where study credibility is won or lost.

Talk to Keentel

Need EHV/HV line or cable modeling, energization and TOV studies, shunt-compensation sizing, or interconnection-grade EMT analysis?

Keentel Engineering provides power-system studies, POI interconnection engineering, substation and transmission-line design, and owner’s-engineer services across utility-scale renewables, BESS, and large-load projects. Reach us at keentel.com.

About Keentel Engineering & Disclaimers

Transmission Line Modeling

in PSCAD

From Model Selection and Tower Geometry to

Ferranti-Effect Validation

Why Line-Model Fidelity Decides the Study

A transmission line is never just an impedance between two buses. It is a distributed network of inductance, capacitance, and resistance whose behavior changes with frequency, geometry, and length—and the way you represent it in PSCAD™/EMTDC™ directly determines whether your study results mean anything. Choose too simple a model and you miss the travelling-wave behavior that drives switching surges and temporary overvoltages. Choose the right one, configure its geometry correctly, and a single line model can tell you whether an EHV circuit will overstress equipment at no load, how a breaker will see recovery voltage, or how a corridor will interact with nearby inverter-based resources.

This guide walks through how we build and sanity-check a distributed line model in PSCAD—starting from the choice of representation, moving through the transmission segment, tower geometry, interfaces, and source, and finishing with a classic validation: the Ferranti effect at no load. We use a Bergeron build as the worked example because it is the cleanest way to see every part of the workflow, and because the Ferranti effect is a fundamental-frequency phenomenon that a single-frequency model captures well.

Three Ways PSCAD Represents a Line

PSCAD offers three families of transmission-line representation. They are not interchangeable; the requirements of the study decide which is appropriate.

PI section

A lumped-parameter model. It gives the correct fundamental-frequency impedance, but it cannot represent other frequencies accurately unless many sections are cascaded—which is computationally inefficient—and it does not capture the frequency dependence of line parameters such as skin effect. It is reserved for very short lines, where travelling-wave models cannot be used because the wave would cross the line in less than a single time step.

Bergeron model

A distributed travelling-wave model that represents the line’s L and C in a distributed manner (effectively an infinite number of PI sections), with resistance lumped at the ends. Critically, it is a single-frequency model: all parameters are computed at one specified frequency—typically 50 or 60 Hz for AC lines—so only results at that steady-state frequency are strictly meaningful. That makes it well suited to fundamental-frequency work such as load-flow matching, relay studies, and power-frequency overvoltage checks like the Ferranti effect.

Frequency-Dependent (Phase) model

The most accurate time-domain line model available. Like Bergeron it is a distributed travelling-wave model, but it distributes resistance along with L and C and, decisively, solves the line parameters at many frequency points—so it captures frequency dependence and attenuation across a wide band, including unbalanced geometries. PSCAD’s own guidance is that this should be the model of choice for new line and cable studies unless there is a specific reason to use another. It costs more solve time than Bergeron, which is the trade-off for broadband accuracy.

How we choose

For broadband transient work— switching surges , lightning, harmonic interaction, IBR/ EMT interconnection studies—we default to the Frequency-Dependent (Phase) model. For fundamental-frequency phenomena where a single-frequency representation is exact enough—load flow, relay reach, and the no-load Ferranti check that follows—the Bergeron model is appropriate and faster. The PI section we keep for very short segments only.

The worked example below builds a 100 km, three-phase, single-circuit overhead line as a Bergeron model and prepares it for a no-load validation.

1. Place the transmission segment

From the component wizard, drop in a transmission segment. PSCAD distinguishes between an overhead T-Line segment and a Cable segment; choose the T-Line for an overhead corridor. Rename the segment something descriptive—here, simply T line—and create it. The name matters later, because it is the key PSCAD uses to tie the segment to its interfaces.

2. Edit the segment parameters

Open Edit Parameters and set the electrical and physical basics:

Segment name — must match the name used by the interfaces you will place at each end (see step 5).

Frequency — the steady-state frequency at which the model is solved (50 Hz or 60 Hz depending on the system).

Segment length — 100 km in this example.

Number of conductors — for a three-phase single-circuit line this is 3 (phase conductors only). A double-circuit line is 6; add ground/shield wires and the count rises accordingly—e.g. 7 or 8 with one or two ground wires.

Termination style determines how the segment connects to the rest of the network, and changing it updates the segment symbol:

Coupling and tandem options are left unset for a standalone single corridor.

3. Select the Bergeron model in the definition

Open Edit Definition. The same frequency, length, and conductor count appear here. By default the segment carries a Frequency-Dependent model object; to build a Bergeron line, delete that default object, right-click, choose Transmission Model, and select Bergeron. (The same menu is where you would instead choose Frequency-Dependent for a broadband study.) The Bergeron model options require no further configuration for a standard build.

4. Define the tower cross-section and conductor geometry

With the model selected, PSCAD prompts for a tower cross-section. The available configurations cover the common structures—flat, vertical, horizontal, single- and double-circuit, a universal tower, and manual data entry—because the physical arrangement of the conductors is what sets the line’s electrical parameters. Using the universal tower with default parameters, the geometry in this example is:

Conductor type — Chukar (a standard ACSR conductor).

Tower center — 0 m defines the structure’s centerline; conductor horizontal positions (x) are measured from it, and vertical positions (y) from the ground.

Phase conductors — Y-phase on the centerline (x = 0), R-phase 5 m to the left (x = −5 m), B-phase 5 m to the right (x = +5 m), all at 30 m above ground. Because all three sit at the same height, this is a horizontal configuration.

Ground (shield) wires — two conductors at x = −2.5 m and x = +2.5 m, each at 40 m above ground—above the phase conductors, as expected for lightning shielding.

Why geometry is not a detail

Conductor spacing and height set the series inductance and shunt capacitance of every phase, and therefore the surge impedance, the charging current, and the no-load voltage rise you are about to validate. Two lines of identical length and voltage but different tower geometry will not behave the same. Getting the cross-section right is getting the physics right.

5. Add the T-Line interfaces and match the name

Place a T-Line interface at each end of the corridor and give both interfaces the same segment name you assigned in step 2 (T line). Set the number of equivalent conductors to 3 to match the build. PSCAD links the two interfaces to the single 100 km segment by name—not by physical proximity—so the interfaces can sit anywhere on the schematic and the result is identical. Mirror the second interface so its ports face the opposite bus.

6. Convert three-phase to single-line and connect the buses

The interface exposes three individual phase conductors. To land them on a single-line bus, insert a three-phase-to-single-line converter at each end and wire it through in wire mode, mirroring on the receiving side. The corridor now runs cleanly from Bus 1 (sending end) to Bus 2 (receiving end) as a single 100 km, three-phase line.

7. Add and configure the source

Add a voltage source at the sending end. Starting from a default three-phase source, set it to the study voltage (230 kV in this example, behind a typical 100 MVA rating), set the frequency to match the segment, and configure it as an infinite (ideal) bus so the sending-end voltage is held firm while you observe what the line does to the receiving end. The model is now ready to validate.

The Ferranti effect is the canonical no-load check for a long line or cable. With the receiving end (Bus 2) open—no load connected—the line’s distributed capacitance draws a leading charging current through its series inductance, and that current raises the receiving-end voltage above the sending-end voltage. The longer the line and the higher its capacitance, the larger the rise; cables, with far greater capacitance per kilometer than overhead lines, show the effect even at modest lengths.

To observe it, meter the RMS voltage at both ends—call them Ea1 at Bus 1 and Ea2 at Bus 2—run the case at no load, and compare. A correctly built model will show Ea2 > Ea1, with the magnitude of the rise tracking the line length and geometry. Because this is a steady-state, fundamental-frequency phenomenon, the single-frequency Bergeron model represents it faithfully; you do not need the broadband Frequency-Dependent model to get a trustworthy answer here.

Why this matters beyond the textbook

No-load and light-load voltage rise is not an academic curiosity—it drives real engineering decisions: sizing shunt reactors to absorb the line’s charging reactive power, setting temporary-overvoltage (TOV) withstand for insulation coordination, defining energization and switching sequences, and confirming that equipment at the open end is not overstressed. The same model that demonstrates the Ferranti effect becomes the basis for the compensation and overvoltage studies that follow.

From here the workflow extends naturally: plot Ea1 and Ea2 over time, quantify the per-unit rise, then iterate—adding shunt compensation, testing energization transients with the Frequency-Dependent model, or sweeping line length—to turn a single validated corridor into a full overvoltage and compensation study.

How Keentel Applies This

Technical FAQ

Keentel Engineering is a power systems and grid-interconnection firm with offices in Tampa, Florida and Austin, Texas. Our services span power-system studies (EHV/HV/MV), point-of-interconnection engineering, substation and transmission-line design, utility-scale renewables and BESS engineering, owner’s-engineer services, and NERC O&P compliance, with deep EMT-modeling capability across interconnection workflows.

Non-affiliation and trademark notice

This document is an independent technical commentary prepared by Keentel Engineering for educational purposes. Keentel Engineering is not affiliated with, authorized by, sponsored by, or endorsed by Manitoba Hydro International Ltd. PSCAD™ and EMTDC™ are trademarks of Manitoba Hydro International Ltd. Conductor names and all other trademarks referenced are the property of their respective owners and are used for identification and descriptive purposes only.

Software behavior, model options, and default settings may change across PSCAD versions; always confirm against the official PSCAD documentation and the vendor’s current guidance. Example values (voltage, length, frequency, geometry) are illustrative and not design recommendations. This material does not constitute professional engineering advice for any specific project; engagement of a licensed engineer is recommended for project-specific application.

Common questions on transmission-line modeling in PSCAD, answered from a study-engineering perspective.

Choosing and configuring the model

Segment parameters and geometry

Connections, source, and validation

How Data Centers Work: Power, Cooling & Uptime

Thu, 18 Jun 2026 13:42:27 GMT

Jun 18, 2026 | blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

How Data Centers Actually Work: The Electrical and Mechanical Systems Behind Continuous Uptime

What is T&D Co-Simulation?

Confusing Physical Connections with Logical Nodes in IEC 61850

Three Problems Every Data Center Must Solve

Following the Power: From the Grid to the Server Rack

The Three System Groups

Two stages deserve emphasis. The service entrance switchgear is where the facility takes control of its own power: it provides protection, isolation, metering, and the protective relaying that clears faults quickly and selectively. And the backup chain, generators plus an automatic transfer switch or generator paralleling switchgear, is what carries the building through a utility outage. Because generators need several seconds to start and stabilize while servers cannot tolerate even milliseconds of interruption, the UPS bridges that gap: it conditions incoming power continuously and supplies instantaneous battery power the moment the utility fails, then recharges once the generators are carrying the load. Modern UPS plants increasingly use lithium-ion batteries over legacy valve-regulated lead-acid for their higher energy density, smaller footprint, longer life, and lower maintenance, and they are typically arranged in N+1 modules with static and maintenance bypass so they can be serviced without dropping the load.

The Data Center as an Energy-Conversion Building

Redundancy and Uptime: Why Nothing Is Single-Threaded

Following the Heat: The Cooling Spectrum

Every email, stream, transaction, and AI query lives somewhere physical: a building filled with servers, electrical infrastructure, and cooling equipment running without pause. From the outside a data center looks like a windowless warehouse. Inside, it is something more specific than an IT facility. It is an energy-conversion building. Electricity goes in, computing work is performed, and heat comes out, and essentially every watt delivered to a server reappears as heat that must be removed. The entire facility exists to manage that conversion safely and continuously.

That framing is the key to understanding how these buildings are designed. The two largest engineered systems, power and cooling, are not independent: they are the input and output sides of the same energy flow, and they are sized against each other. This article follows that flow end to end, from the utility connection down to the server and back out as heat, and then looks at the controls and the redundancy that make the whole thing dependable.

Why this matters for power planning

Because nearly all delivered power becomes heat, the cooling system is itself one of the largest electrical loads in the building. Power in and heat out are two views of the same number, which is why the total facility load presented to the grid, and the interconnection that must support it, can only be understood by looking at power and cooling together.

A quick-reference set of answers to the questions that come up most often about how data centers work.

The progression is not a replacement so much as a layering. Room-based air remains viable for lower densities and is familiar to most mechanical contractors. Close-coupled air extends air's useful range and lets different rows be cooled independently when load profiles vary. Direct-to-chip liquid, now becoming standard for AI workloads, exploits liquid's far greater heat-transfer capacity to handle densities air cannot, and because its loop can run warmer than traditional chilled water, it often enables economized heat rejection for much of the year, cutting or eliminating compressor energy. Immersion pushes further still, eliminating high-volume airflow entirely at the cost of specialized hardware. Most high-density facilities now run a mix, with the choice at each location driven by density, climate, water availability, energy cost, redundancy requirements, and the balance between capital cost and operating efficiency. There is no single universal answer.

The density inflection

Rising rack density is the single force reshaping data center design today. It is pulling liquid cooling into the mainstream, changing how facility water systems are planned, and pushing total campus loads, and therefore interconnection requirements, sharply upward. Power and cooling decisions that used to be made separately now have to be made together, against the same density assumptions.

The third system group is easy to overlook because it is mostly invisible, but it is what turns a collection of equipment into a coordinated, self-protecting facility. Building management and automation systems supervise the mechanical plant; electrical power monitoring systems watch the power chain; and data center infrastructure management software ties the whole picture together. These systems continuously track temperatures, power draw, equipment status, and alarms, and they act on what they see, staging cooling capacity up and down, initiating failover when a component drops out, and alerting operators before a developing problem becomes an outage.

In a facility whose defining requirement is continuous operation, this automated awareness is not a convenience; it is a core reliability system. The handoffs that keep the load alive during a utility failure, UPS to generator, primary cooling to backup, happen on timescales too short for human reaction, so the controls layer has to be engineered with the same rigor as the equipment it governs.

These levels map onto the industry tier classifications used to describe reliability. Lower tiers have little or no redundancy; mid tiers add redundant components; higher tiers add independent distribution paths so the facility is concurrently maintainable (any component or path can be serviced without downtime) and ultimately fault tolerant (it can absorb an unplanned single failure with no impact, generally a 2N architecture). Layered correctly across generators, UPS plant, distribution paths, chillers, pumps, towers, and in-room cooling, this is what lets well-run facilities target availability around 99.999 percent, roughly five minutes of downtime a year. The detailed mechanics of the power chain and of the cooling plant are explored further in our companion articles on data center electrical systems and on containment and chilled water.

The drive for continuous operation expresses itself, everywhere, as redundancy: spare capacity and duplicate paths so that a failure or a maintenance event is absorbed without reaching the load. The same vocabulary applies to both power and cooling.

Before the Load Arrives: The Interconnection Dimension

Why Data Centers Don't Behave Like Other Buildings

From the curb, a data center can resemble any commercial building, but operationally it is a different animal. An office building's load rises and falls with occupancy, cooling demand cycles through the day, and equipment switches on and off. A data center's electrical load is essentially constant, its cooling demand is continuous, and its systems rarely shut down. Failure tolerance differs just as sharply: a cooling failure in an office is a comfort problem, while a power or cooling failure in a data center is an immediate operational risk. That is why these facilities are built with redundant equipment, multiple power and cooling paths, and designs that allow maintenance without shutting down, characteristics that would be unusual, and unnecessary, in almost any other building type.

How Keentel Engineering Approaches Data Center Design

There is one more layer, upstream of everything above. All of this internal infrastructure presumes the facility can secure the grid capacity it needs, and at modern scale that is no longer a safe presumption. As campus loads climb into the hundreds of megawatts, driven in large part by the density trends reshaping cooling, service moves from distribution voltages toward sub-transmission and transmission, and the project enters the utility interconnection process: queue positions, system impact and facilities studies, point-of-interconnection engineering, and the protection and modeling obligations of connecting a large load to the bulk power system.

Increasingly, these realities govern site viability and schedule more than any decision made inside the building. The most elegantly engineered power chain and cooling plant are worth nothing if capacity cannot be delivered to the site on a workable timeline. That is why interconnection belongs at the front of the design conversation, not the end of it.

Frequently Asked Questions

Regardless of size, every data center is built to solve three problems at once, and each one drives major design decisions.

Continuous power

Servers cannot simply ride out a power loss. Even a brief interruption can corrupt in-flight data or take services offline for thousands or millions of users. Power must remain available even when utility power is interrupted or individual equipment fails, which is why the electrical system is layered with conditioning and multiple backup sources rather than relying on the grid alone.

Continuous cooling

Because servers generate heat constantly, cooling has to run constantly. If it stops, temperatures climb in minutes, forcing equipment to throttle or shut down to protect itself. Cooling here is not about human comfort; it is about equipment survival, and it carries the same intolerance for interruption as the power system.

Continuous operation

The facility must keep running through maintenance, failures, and repairs. That requirement, more than any other, is what drives the pervasive use of redundancy and multiple paths across both electrical and mechanical systems, so that work can be done and components can fail without taking the IT load down.

Although the infrastructure looks complex, most of it falls into three groups that work as one.

Electrical systems bring power into the building and distribute it safely to the servers, spanning the utility connection, switchgear, transformers, backup power, and distribution equipment. The objective is simple: power must always be available.

Mechanical cooling systems remove the heat the servers produce, using some combination of chillers, cooling towers, pumps, air handlers, and increasingly liquid-cooling equipment located right at the racks. The objective is to keep equipment within safe temperatures at all times.

Controls and monitoring tie the electrical and mechanical systems into a single operating environment, tracking temperatures, power, equipment status, and alarms, and responding automatically to changing conditions or failures. In a modern facility this layer is as important as the physical equipment.

Power inside a data center moves through a sequence of controlled transformations and protected distribution stages, each taking electricity a step closer to the electronics. The chain begins well outside the building.

Utilities transmit power at high voltage and correspondingly low current, because line losses rise with the square of current; moving energy at high voltage is what makes long-distance delivery efficient. That power steps down at substations and is routed to the campus. From there, the on-site chain takes over:

Once power reaches the servers, electrical engineering hands off to mechanical engineering. Every kilowatt delivered becomes thermal energy that has to be removed instantly and continuously. How that is done depends overwhelmingly on one variable: heat density, the power dissipated per rack.

That number has climbed dramatically. Traditional enterprise racks once averaged roughly 3 to 5 kilowatts; many today run 10 to 20; and AI and GPU clusters can exceed 50, 80, or even 100 kilowatts per rack. Air has a relatively low heat capacity, so as density rises, air-based systems must move ever larger volumes faster, consuming more fan energy until they reach physical limits. This is why cooling has steadily shifted from cooling whole rooms to capturing heat at its source, producing four broad methods arranged along a density spectrum.

Controls and Monitoring: The Facility's Nervous System

Keentel Engineering is a power systems and interconnection consulting firm, and we treat a data center as the integrated energy-conversion building it is, connecting the internal power and cooling architecture to the grid-side realities that determine whether and when it can be energized. Our work spans point-of-interconnection engineering across medium, high, and extra-high voltage; power system studies including load flow, short-circuit, protective device coordination, and arc-flash analysis; substation and switchgear design; owner's engineer services; and NERC operations and planning compliance for facilities that interact with the bulk power system.

Whether you are evaluating a site, sizing and securing a large load, or coordinating the electrical and mechanical design for a new or expanding campus, we can help align the internal reliability strategy with the interconnection path that ultimately defines your route to bringing the facility online

Data Center Electrical Design: From Grid to Rack

Thu, 18 Jun 2026 12:40:22 GMT

Jun 18, 2026 | blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

From the Grid to the Rack: Engineering Continuous Power for the Data Center

What is T&D Co-Simulation?

Confusing Physical Connections with Logical Nodes in IEC 61850

The Power Chain: From Utility Service to the Server Rack

Conditioning and Bridging Power: The UPS

Power inside a data center moves through a series of controlled voltage transformations and distribution steps, each one taking electricity a level closer to the sensitive electronics in the racks. Understanding that chain is the foundation for everything that follows.

Utility service and the service entrance

Electricity arrives from the grid at medium voltage, commonly in the range of roughly 13.8 to 34.5 kilovolts for many facilities. As campus loads have grown into the hundreds of megawatts, large hyperscale sites increasingly take service at sub-transmission or transmission voltages instead, which is one of the reasons interconnection has become a gating item for new development. The incoming supply enters through service equipment that meters and terminates the utility connection.

Switchgear: the control center

From the service entrance, power passes into switchgear, which is among the most important pieces of equipment in the building. Switchgear is the control and protection hub: it lets operators isolate circuits, redirect power flows between sources, and protect the system with circuit breakers and protective relays. When a fault occurs, the protective relays and breakers in the switchgear are what clear it quickly and selectively, ideally isolating only the affected section while the rest of the facility keeps running. The quality of that protection scheme is a direct determinant of both safety and uptime.

Step-down transformers

Switchgear feeds step-down transformers that reduce the medium-voltage supply to levels usable inside the building, most commonly 480 volts three-phase for large equipment and primary distribution. Some modern facilities distribute at 400/415 volts to improve efficiency and simplify the path to the rack.

PDUs and rack distribution

Power distribution units (PDUs) take that 480-volt distribution and transform it again to the voltages the IT equipment uses, such as 208 or 120 volts (or 240 volts in a 415-volt scheme). From the PDU, branch circuits feed rack-level power strips, often called rack PDUs, which deliver power directly to the dual power supplies in each server. The net effect is a deliberate cascade, from high-voltage utility supply down through each protected, transformed stage to the electronics themselves.

The Long Haul: Backup Generators

Why Electrical Design Is the Foundation of Data Center Reliability

Designing Against Failure: Redundancy Architecture

Batteries bridge seconds and minutes; generators carry the facility through hours and days. Most large data centers use diesel generators, frequently rated at several megawatts each and paralleled to support the full load with redundancy. When utility power fails, these units start automatically and are typically designed to reach rated output and accept load within roughly ten seconds to under a minute, depending on their emergency-power classification.

An automatic transfer switch (ATS) senses the loss of utility power, signals the generators to start, and transfers the building load to generator power once the units are stable. The UPS covers the brief interval between the outage and the moment the generators are carrying the load. On-site fuel storage commonly provides hours to a few days of autonomy, and fuel deliveries extend that further during prolonged grid events. Once generators are online, they assume the load, and the UPS rectifier, now fed by generator power, recharges the batteries while continuing to condition the output.

Servers and cooling tend to get the attention, but neither matters without one thing delivered continuously and cleanly: electrical power. A data center exists to keep computing available, and availability is fundamentally an electrical promise. A few seconds without power can reboot tens of thousands of servers, corrupt in-flight data, and take services offline across the world. Unlike an office building, a data center cannot tolerate even a momentary interruption.

Meeting that standard takes more than a strong utility feed. It takes a layered electrical system engineered so that no single failure, and no planned maintenance activity, ever reaches the IT load. This guide walks the full path that power takes from the utility connection down to the server, the systems that bridge and back up that power when the grid falters, and the redundancy architecture that ties it all together. It closes on the part Keentel works in most directly: the protection studies and the interconnection engineering that determine whether the facility can be powered safely and brought online at all.

The Keentel lens

Everything inside the fence depends on what happens at the fence. The reliability of a data center's internal power chain is only as good as the capacity secured at the point of interconnection and the protection scheme that lets that capacity be delivered safely. We treat the utility interconnection, the protection and coordination studies, and the internal distribution design as one continuous problem.

Distributing power is not enough; it also has to be clean and uninterrupted. Utility power is imperfect, carrying voltage fluctuations, frequency variations, transients, and occasional complete outages, and sensitive electronics react badly to even brief disturbances. The uninterruptible power supply (UPS) sits between the incoming power and the critical load, acting as both a power conditioner and an instantaneous source of backup energy.

How a Double-Conversion UPS Works

A UPS has three core elements: a rectifier, an energy store (typically a battery system), and an inverter. In the dominant data center design, the double-conversion UPS, incoming AC is converted to DC by the rectifier. That DC bus simultaneously charges the batteries and feeds the inverter, which converts it back into clean, tightly regulated AC for the load.

The key insight is that the load is always supplied by the inverter, never directly by the utility. The incoming supply is fully decoupled from the output, so disturbances on the utility side are filtered out before they can reach a server. This is also why the response to an outage is seamless: if utility power fails, the rectifier simply stops receiving input, but the already-charged batteries continue feeding the same inverter. There is no switchover, no transfer time, no flicker. That zero-interruption behavior is exactly what the name describes.

Double conversion vs. line-interactive

Line-interactive UPS designs regulate voltage and switch to battery only when needed, with a brief transfer. They suit small or edge applications. Mission-critical data centers favor double conversion because it delivers the highest power quality and a true zero-transfer response, which is worth the modest efficiency cost. Some units offer an economy mode that bypasses the conversion to save energy, trading away a degree of isolation; whether to enable it is a reliability-versus-efficiency decision.

Battery Technologies

UPS batteries are sized to carry the full load for a short bridge period, typically a few minutes up to around fifteen. That is not meant to ride out a long outage; it is meant to span the seconds it takes for generators to start and stabilize. Two battery technologies dominate, and the choice balances cost, footprint, and service life

A quick-reference set of answers to the questions that come up most often in data center electrical design.

The Handoff Sequence, Step by Step

When the utility fails, a precisely choreographed sequence plays out automatically, with no human intervention required:

Utility power is lost.
The UPS instantly carries the load from its batteries through the same inverter, with no transfer time.
The transfer scheme detects the outage and signals the generators to start.
The generators start, reach stable voltage and frequency, and accept load, typically within about ten seconds to under a minute.
The transfer switch moves the building load onto generator power.
The UPS, now fed by the generators, recharges its batteries while continuing to condition power to the servers.
When utility power returns and stabilizes, the system transfers back and the generators shut down after a cooldown.

The result is a power source that, from the perspective of the servers, never went away.

N+1 is the most common level in enterprise facilities because it offers a strong balance of reliability and cost: a UPS plant that needs four modules is built with five, so any one module can fail or be serviced while the load stays fully supported. 2N goes further by duplicating the entire infrastructure, two utility feeds, two switchgear lineups, two UPS systems, and two distribution paths, each sized for the full load . Beyond these, distributed-redundancy schemes (for example, arranging three systems so that any two can carry the load) can achieve fault tolerance with less than full duplication, trading some design complexity for capital efficiency.

Redundancy is the design principle that ties the whole system together. Rather than depending on any single component, the electrical system installs spare capacity and duplicate paths so that a failure, or a planned maintenance outage, can be absorbed without affecting the load. The same vocabulary describes redundancy across both electrical and cooling systems.

Tiers, Concurrent Maintainability, and Five-Nines

A/B Distribution, Dual-Corded Loads, and Transfer Switches

In a 2N design, the duplication extends all the way to the rack. Two independent power paths, conventionally called the A side and the B side, run from separate sources through separate UPS systems and PDUs to the racks. Servers with dual power supplies are corded to both paths, so the loss of either side, whether from a fault or planned maintenance, leaves the equipment running on the other without interruption.

For equipment that has only a single power input, a static transfer switch (STS) provides a similar benefit by switching that load between the A and B sources in milliseconds. Together, dual-corded distribution and static transfer switches are what let an entire power path be taken out of service for maintenance while the IT load continues uninterrupted, the practical expression of concurrent maintainability.

Two distinctions matter most. Concurrent maintainability (the threshold for Tier III) means any single component or distribution path can be removed from service for maintenance or replacement without affecting the IT load; it requires both redundant components and independent distribution paths. Fault tolerance (Tier IV) is stronger still: the facility can absorb an unplanned failure of any single component or path with no impact, which generally implies a 2N architecture with multiple simultaneously active paths and physical separation between them.

These design choices are what allow well-run facilities to target availability of 99.999 percent or higher, the so-called five-nines level, which corresponds to only about five minutes of downtime per year. Reliability at that level is not achieved by any single device; it is the product of layered distribution, instantaneous backup, and disciplined redundancy working together.

Redundancy levels map closely onto the data center tier classifications used to describe overall reliability. Higher tiers carry more redundancy and, importantly, more independence between paths.

The Studies Behind a Safe, Reliable Power System

Before the Load Arrives: The Interconnection Dimension

Frequently Asked Questions

Redundant equipment only delivers reliability if the system that connects it is engineered correctly, and that is established through power system studies rather than assumed. Several are central to a sound data center electrical design:

Load flow analysis confirms that voltages and equipment loadings stay within limits across normal and contingency configurations, including when the load runs on generators.

Short-circuit analysis establishes the fault currents the system must withstand, so breakers and switchgear are rated to interrupt them safely.

Protective device coordination tunes relays and breakers so that the device closest to a fault clears it first, isolating the smallest possible section and preserving uptime everywhere else.

Arc-flash analysis quantifies the incident energy at each location, driving labeling and personal protective equipment requirements that keep maintenance work safe, which is a prerequisite for true concurrent maintainability.

Done well, these studies are what make the difference between a system that is redundant on paper and one that actually rides through faults and maintenance without dropping the load.

There is a layer upstream of everything described so far. All of this internal infrastructure assumes the facility can secure the grid capacity it needs in the first place, and at the scale of modern data centers that assumption is no longer safe to make. Loads now routinely reach into the hundreds of megawatts, which pushes service from distribution voltages up toward sub-transmission and transmission, and brings the project squarely into the utility interconnection process: queue positions, system impact and facilities studies, point-of- interconnection engineering , and the protection and modeling requirements that come with connecting a large load to the bulk power system.

These realities increasingly determine site viability and schedule more than any decision made inside the building. The most carefully engineered internal power chain is of no value if capacity cannot be delivered to the site on a workable timeline. That is why we treat interconnection not as a downstream utility formality but as a first-order design input, evaluated before the rest of the electrical design is locked in.

How Keentel Engineering Approaches Data Center Power

Keentel Engineering is a power systems and interconnection consulting firm, and data center electrical design sits at the center of what we do. We connect the internal critical-power architecture to the grid-side realities that govern whether and when a facility can be energized: point-of-interconnection engineering across medium, high, and extra-high voltage; system studies including load flow , short-circuit, protective device coordination, and arc-flash analysis; substation and switchgear design; owner's engineer services; and NERC operations and planning compliance for facilities that interact with the bulk power system.

If you are siting a new data center, sizing and securing a large load, or coordinating the electrical design for a new or expanding campus, we can help align the internal reliability strategy, the protection scheme that makes it safe, and the interconnection path that ultimately determines your route to energization.

Data Center Cooling Design: Containment & Chilled Water

Thu, 18 Jun 2026 11:10:19 GMT

Jun 18, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Rack Layout Fundamentals: The Hot-Aisle / Cold-Aisle Arrangement

Behind the Air: The Chilled Water System

How airflow containment and chilled water systems work together to remove heat, protect uptime, and shape the electrical load a facility ultimately draws from the grid.

A reasonable question is why large facilities go to the trouble of a chilled water plant instead of distributed direct-expansion (DX) units, such as packaged rooftop systems, which are simpler per unit. At small scale DX is often the right answer. At data center scale the balance reverses:

Efficiency. Centralized chillers and evaporative heat rejection are markedly more efficient per unit of heat removed than many small DX compressors.

High heat loads. Water carries far more heat per unit volume than air, making it better suited to the megawatt-scale loads modern halls produce.

Centralized control. One plant is easier to monitor, optimize, and stage than dozens of independent units.

Flexible redundancy. It is far easier to build N+1 or 2N reliability into a centralized plant with shared spares than to duplicate every distributed unit.

For these reasons most large facilities rely on chilled water rather than packaged DX, reserving DX for smaller spaces or specific edge applications.

This redundancy is applied across the whole plant: multiple chillers, backup pumps, and redundant piping loops so that no single failure interrupts heat removal. Two related design goals come from the Uptime Institute tier framework. Concurrent maintainability means any component can be taken offline for service without disrupting operation, which requires at least N+1 and redundant distribution paths. Fault tolerance means the system can absorb an unplanned failure of any single component or path without impact, which generally implies 2N. Matching the cooling topology to the electrical topology, so that one is not the weak link of the other, is a core part of designing critical facilities.

Why this mirrors the power design

The same N+1 and 2N vocabulary governs UPS systems, generators, and distribution paths on the electrical side. A facility designed for concurrent maintainability or fault tolerance on power, but with a single point of failure in the cooling plant, is only as resilient as its weakest system. Power and cooling redundancy have to be designed to the same standard.

Designing for Reliability: Redundancy and Concurrent Maintainability

Containment and Chilled Water: Engineering Thermal Management for High-Density Data Centers

Why Thermal Management Is a First-Order Design Decision

Everything described so far assumes air can carry heat from the chip to the coil. For a long time that held, but AI and high-performance computing have pushed rack power densities past the point where air alone is practical. Where a conventional rack might draw a handful of kilowatts, accelerated-compute racks now reach many tens to over a hundred kilowatts, and air simply cannot move that much heat through the rack volume.

The response is to bring liquid closer to the heat source. Several approaches now coexist, often within the same facility:

Rear-door heat exchangers replace the rack's back door with a chilled water coil that captures exhaust heat right at the rack, a natural extension of contained-aisle thinking.

Direct-to-chip cooling circulates coolant through cold plates mounted on the hottest components, removing heat at its source rather than from the surrounding air.

Immersion cooling submerges entire servers in a dielectric fluid, transferring heat directly to the liquid with no server fans at all.

These technologies do not replace the chilled water plant; they extend it. A coolant distribution unit (CDU) typically separates the clean technology loop serving the chips from the facility chilled water loop, and that facility loop still relies on the same chillers, pumps, and heat rejection described above. The hot-aisle/cold-aisle and containment principles remain relevant for the air-cooled equipment that continues to share the hall. The practical takeaway for designers is that thermal strategy is now mixed by default, and the facility water system has to be planned to serve both air-cooled and liquid-cooled loads.

In a data center, the servers are the reason the building exists, and the electrical system is what keeps them running. But neither matters if the heat those servers produce cannot be removed continuously. Every watt of power delivered to IT equipment is converted, almost entirely, into heat that must be carried back out of the building every second of every day. A facility can have flawless power infrastructure and still fail within minutes if its cooling stops.

That makes thermal management a first-order engineering problem rather than a secondary mechanical detail. It directly governs reliability, the temperatures equipment is held at, the energy a site consumes for non-computing functions, and ultimately the total electrical load presented at the point of interconnection. Two design domains carry most of that responsibility: airflow management inside the data hall, and the chilled water system that removes the heat the air picks up. This guide treats them as the connected system they are.

The Keentel lens

Cooling is not just a mechanical line item. The cooling plant is one of the largest electrical loads in the building. Decisions about containment, supply temperatures, and heat-rejection method change a facility's effective load profile and its total interconnection request, which is where power systems engineering and mechanical design meet.

The Core Problem: Heat, Airflow, and the Cost of Mixing

Most server-class equipment is front-to-back air cooled. Cold air is drawn in through the front intake, passed across heat-generating components, and exhausted out the rear at a substantially higher temperature. The cooling system's job, at its simplest, is to deliver cold air to those intakes and collect the hot exhaust for heat rejection.

The single most damaging inefficiency in this loop is the mixing of hot exhaust air with the cold supply air before it reaches the servers. When the two streams blend, the air arriving at the intakes is warmer than intended, so the cooling plant must overcool the supply to compensate. The consequences compound quickly:

Higher energy consumption, because the plant works harder and runs colder than it needs to.
Localized hot spots, where recirculated exhaust raises inlet temperatures above equipment limits and triggers throttling or thermal shutdown.
Reduced usable cooling capacity, since a meaningful fraction of the cold air never does useful work.
Greater strain and shortened service life on cooling units forced to operate outside their efficient range.

There are two failure modes worth naming precisely. Recirculation is hot exhaust looping back to an intake. Bypass is cold supply air returning to the cooling unit without ever passing through a server. Both waste capacity, and both are symptoms of the same root issue: the hot and cold air are sharing the same volume. Every airflow strategy that follows is fundamentally an effort to keep those two streams apart.

Containment is what makes aggressive temperature strategies safe. Once hot and cold streams are separated, the operator can run warmer than the cautious, overcooled set points of a mixed room without risking equipment.

Industry thermal guidance, most commonly the ASHRAE TC 9.9 recommendations, defines a recommended inlet-air envelope of roughly 18 to 27 degrees Celsius (about 64 to 81 degrees Fahrenheit), with wider allowable ranges for short excursions. Many operators historically ran far colder than necessary out of caution. Effective containment lets them move set points up toward the warmer end of the recommended band with confidence, and warmer supply air is cheaper to produce.

The larger prize is economization, or free cooling. When the system can supply acceptably warm air, the chilled water plant can lean on outdoor conditions to reject heat for large parts of the year, reducing or eliminating mechanical refrigeration during cool weather. A waterside economizer uses the cooling towers to make chilled water directly when ambient wet-bulb temperature is low enough; an airside economizer brings in filtered outside air. Either way, the savings only materialize because containment allows the warmer operating temperatures that make outdoor conditions useful. Containment, supply temperature, and economization are a single linked decision, not three separate ones.

The baseline discipline that makes everything else possible is orienting racks into alternating hot and cold aisles. Rows are arranged so that rack fronts face each other across one corridor and rack backs face each other across the next.

Cold-aisle orientation

The corridor between two rows of rack fronts becomes the cold aisle. Conditioned air is delivered into it, classically through perforated tiles in a raised floor or through overhead ducting, and the servers draw it in. This organizes supply air where the intakes are, rather than dumping it into the room at large.

Hot-aisle orientation

Behind the racks, the backs face each other into the hot aisle, which concentrates exhaust into a defined corridor that can be directed back toward the cooling units. Collecting the hot air in one place raises the return air temperature, and a warmer, more concentrated return stream is precisely what makes cooling equipment more efficient.

This alternating layout is a major improvement over a randomly arranged room, but on its own it is incomplete. With nothing sealing the aisles, hot and cold air still mix freely over the tops of the racks and around the ends of the rows. The layout sets the stage; containment closes the gaps.

Containment: Sealing the Boundary Between Hot and Cold

Containment takes the hot-aisle/cold-aisle arrangement and physically encloses one of the two aisles so the streams cannot mix at all. The enclosure is built from ceiling panels or vertical baffles above the racks, doors at the ends of the rows, and blanking panels that seal the open rack units and gaps between cabinets. The principle is simple: give the cold air exactly one path, from supply to intake to exhaust to return, with no shortcuts. Two strategies dominate.

Cold-Aisle Containment

Here the cold aisle is enclosed. Supply air is trapped in the sealed corridor in front of the racks, so every server sees consistent, low-temperature intake air. The hot exhaust is allowed to fill the rest of the room, which becomes a large warm-air return plenum.

Strengths: It is generally the simpler and lower-cost approach, especially as a retrofit to an existing raised-floor hall. It delivers tightly controlled, uniform inlet temperatures, which is exactly what equipment warranties care about.

Trade-offs: The general room volume runs warm, which can be uncomfortable for staff and harder on equipment located outside the contained aisle, such as network gear and power distribution. The approach also tends to lose its efficiency edge at very high rack densities, where the volume of hot air the room must absorb becomes difficult to manage.

Hot-Aisle Containment

Here the hot aisle is enclosed instead. Exhaust is captured immediately and ducted, often through an overhead return plenum, straight back to the cooling units. The rest of the room is flooded with cool supply air and becomes the comfortable, equipment-friendly space.

Strengths: It typically achieves higher overall efficiency and performs better in high-density environments. Because the return air is captured hot and undiluted, the cooling coils see a larger temperature difference and the plant can run warmer supply temperatures, which unlocks economization (discussed below). The bulk of the room stays cool, which is easier on ancillary equipment and personnel.

Trade-offs: Design and installation are more involved, upfront cost is higher, and the enclosed hot aisle must be coordinated carefully with fire detection and suppression because the ceiling plane and ducting change how smoke and suppression agents move.

The distinction in intent is worth stating plainly. Cold-aisle containment is organized around protecting the equipment by guaranteeing a consistent intake temperature. Hot-aisle containment is organized around maximizing plant efficiency by managing the return air. In modern high-density and AI-oriented builds, hot-aisle containment is generally the preferred strategy, because control of return-air temperature is what makes the rest of the cooling system efficient at scale.

Side-by-side comparison

Tuning the Air Side: Temperature Envelopes and Economization

Air moves heat around inside the hall, but it does not remove heat from the building. That work is done by a mechanical system, and at scale that system is almost always a chilled water plant. Conceptually it is straightforward: water is an excellent, dense heat-transport medium, and the plant uses it to carry heat from the data hall to the outdoors and dump it there.

The loop step by step

A chiller produces cold water.
Pumps push that chilled water out to the data hall.
Inside the hall, the water flows through coils in air-handling units; room air passes over the coils and is cooled.
The cooled air is delivered to the server intakes through the contained cold aisle.
Servers absorb the cool air and exhaust it hot.
The hot return air passes back over the coils, transferring its heat into the water.
The now-warm water returns to the chiller.
The chiller rejects that heat to the outside environment, and the cycle repeats continuously.

The temperature difference between supply and return water, the system delta-T, is a key efficiency metric: a larger delta-T means each gallon of water carries more heat, so the plant can move a given load with less pumping energy. Much of chilled water design is, in effect, the pursuit of a healthy delta-T, and good containment supports it by keeping return air hot and undiluted.

Major Components

Chillers

Chillers are the heart of the plant. They use a refrigeration cycle, conceptually a very large air conditioner, to pull heat out of the water. They come in two broad families that differ in how they reject heat.

Air-cooled chillers reject heat directly to the outdoor air, typically through integral condenser coils and fans. They are simpler, use little or no water, and suit smaller plants or water-constrained sites, but they are generally less efficient and lose capacity in hot weather.

Water-cooled chillers reject heat to a condenser-water loop served by cooling towers. They are more efficient and scale better for large facilities, at the cost of a more complex plant and significant water consumption.

Cooling towers

In a water-cooled plant, cooling towers reject the system's heat by evaporating a portion of the condenser water, one of the most effective large-scale heat-rejection mechanisms available. Their performance tracks the outdoor wet-bulb temperature rather than the dry-bulb, which is why they remain effective even in warm climates and why they enable waterside economization. The trade-off is water: evaporative rejection consumes water continuously, which is why water-use effectiveness has become a sustainability metric alongside energy.

Pumps

Pumps keep water moving through the loops. Plants commonly separate primary pumps, which circulate water through the chillers at a stable flow, from secondary pumps, which deliver water out to the building loads and vary their flow to match demand. This separation protects the chillers from flow swings while letting the distribution side modulate efficiently.

Air-handling units in the hall

Inside the data hall, computer room air handlers (CRAH units) contain the chilled water coils. Hot return air from the contained aisle passes over the coils, gives up its heat to the water, and the cooled air is supplied back to the cold aisle. CRAH units rely on the central chilled water plant rather than housing their own refrigeration, which is the defining difference between a chilled water approach and a direct-expansion one. The air side and the water side meet at this coil, which is exactly why containment and chilled water design have to be considered together.

Why Chilled Water Beats Direct Expansion at Scale

How Keentel Engineering Approaches Data Center Thermal Design

Frequently Asked Questions

The Integrated View: Power, Airflow, Water, and the Grid

Three systems have to work together flawlessly for a data center to function. The power system keeps the servers energized. The airflow system governs how cooling reaches and leaves the racks. The chilled water system removes the heat that airflow collects. They are deeply interdependent: if power fails the servers stop; if airflow is poorly designed cooling becomes ineffective regardless of plant capacity; and if chilled water fails the whole hall overheats. None of the three can be designed well in isolation.

There is a further connection that is easy to overlook and central to how Keentel approaches these projects. The cooling plant is itself a major electrical load, often among the largest non-IT loads in the facility. Chillers, pumps, tower fans, and CRAH fans all draw power, and how efficiently the cooling system is designed directly shapes the facility's overall energy use, commonly expressed as power usage effectiveness (PUE). A more efficient thermal design lowers the total facility power for the same computing capacity.

That total facility power is what a developer must ultimately secure from the grid. The cooling design therefore feeds directly into the size of the load presented at the point of interconnection, the studies required to bring that load on, and the timeline to energization. Thermal decisions made on the mechanical side ripple straight into the electrical and interconnection scope, which is why we treat data center cooling as part of an integrated load and power-delivery picture rather than a standalone mechanical exercise.

A quick-reference set of answers to the questions that come up most often in data center cooling design.

Keentel Engineering brings a power-systems and interconnection perspective to data center development, and we treat thermal design as inseparable from the electrical and load-delivery picture. Our work spans the questions this guide raises: how containment strategy and supply-temperature targets shape plant efficiency; how chiller type, heat-rejection method, and pumping topology affect both reliability and the facility's electrical draw; how cooling redundancy should be matched to the electrical redundancy so neither becomes the weak link; and how the resulting total facility load translates into an interconnection request, the supporting power system studies, and a credible path to energization.

If you are evaluating a site, sizing a load, or coordinating mechanical and electrical design for a new or expanding data center, we can help connect the thermal strategy to the power delivery and interconnection realities that ultimately determine whether and when the facility can come online.

Cooling in a data center is designed with the same uncompromising reliability philosophy as the electrical system, because the two share the same intolerance for downtime. If cooling stops, the consequences arrive in minutes. So chilled water plants are built with deliberate redundancy, described in the same N-based language used across critical infrastructure.

The High-Density Inflection: From Air to Liquid

Power Siting Feasibility Study | Know Before You Commit

Wed, 17 Jun 2026 11:42:12 GMT

Apr 17, 2026 | blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Know Before You Commit: How Keentel Engineering De-Risks Power Siting Decisions

1. First-Pass Feasibility in Weeks — Not Months

Built on Operator-Grade Accuracy

What is T&D Co-Simulation?

Confusing Physical Connections with Logical Nodes in IEC 61850

Our workflows use the same class of power-flow software employed by regional transmission operators during formal interconnection studies.

The result is engineering analysis grounded in the same methodology used by the organizations ultimately evaluating the project.

In validation exercises against actual operator outcomes, our modeling has achieved better than 99% agreement with grid-operator results.

That level of accuracy provides confidence for:

Site-selection committees
Executive leadership teams
Investors and lenders
Development partners

Why the Software Isn't the Story — The Engineering Is

Turning months of grid uncertainty into a feasibility answer in weeks — for generation, load, and everything in between

Ready to Find Out If Your Site Works?

Technology and Innovation: AI in the Interconnection Process

2. Know the Interconnection Cost Before You File

Powerful software can generate data.

It cannot generate judgment.

What clients actually buy from Keentel Engineering is the engineering expertise behind the analysis.

Our engineers know:

Which assumptions matter
Which scenarios are realistic
Which results deserve confidence
How to translate technical outputs into actionable business decisions

Anyone can run a study.

Far fewer can tell you what the queue may realistically look like eighteen months from now, which projects are likely to drop, or whether available substation headroom will still exist when you need it.

That is the difference between receiving data and receiving an answer.

If you're evaluating a location for:

Utility-scale generation
Battery energy storage
Data centers
Manufacturing facilities
Large industrial loads

Talk with Keentel Engineering before committing capital.

We'll help you understand site feasibility, available capacity, and interconnection costs before you enter the queue.

Every large power decision starts with the same unanswered question.

Contact Keentel Engineering today to scope a feasibility assessment for your next project.

A developer eyeing a 200 MW data center wants to know whether the grid at that location can actually carry the load.

A renewable developer scouting a parcel wants to know whether there's room left on the bus before competitors lock it up.

An industrial site selector wants to know what it will cost to connect — before committing capital, not eighteen months into a queue.

For most teams, getting to those answers means a dozen browser tabs, calls to multiple utilities, and weeks of waiting on studies that arrive too late to influence the decision.

By the time the numbers come back, the opportunity has often already moved.

At Keentel Engineering, we've rebuilt that entire front end of the process.

Our analysts can come to a business-development conversation already knowing whether a site is viable, what the grid can absorb today, and what an interconnection is likely to cost — in a matter of weeks rather than the months a traditional study takes.

The Question We Answer

Is this site even worth pursuing?

Before anyone files anything or commissions a formal study, our team produces a first-pass feasibility assessment on virtually any site in the country.

We bring together the factors that actually determine whether a location works :

Nearby transmission infrastructure and available capacity
Substation headroom for both injection and withdrawal
Land ownership around the parcel
Existing and queued generation projects
Natural gas pipeline flows for gas-dependent projects
Long-range energy price forecasts through 2045

We can load this analysis directly from a coordinate set, KMZ file, or shapefile your team already has.

What used to take months of utility outreach and fragmented research can now be completed in weeks.

Why It Matters

Instead of pursuing every opportunity equally, your team can:

Rank sites quickly
Eliminate non-viable locations early
Focus capital on projects with a realistic path forward
Enter development discussions with confidence

You walk into the room already knowing whether the grid can support your project — and what it may take to get there.

The Question We Answer

What will it actually cost to connect?

A feasibility assessment tells you a site deserves further evaluation.

The next question is often the most important:

What will interconnection cost, and will the project still make financial sense?

Keentel Engineering estimates network upgrade costs before an interconnection application is filed.

That means clients aren't discovering deal-breaking numbers after investing months of time and significant development capital.

For Generation Projects

Siting new generation means planning around an interconnection queue that is constantly changing.

Our engineers build customized scenarios that account for:

Potential queue withdrawals
Competing project sizes
Different technology types
Market-specific development trends

This creates a realistic view of what the queue may look like when your project reaches the next study phase.

The result is a more defensible cost estimate than a standard off-the-shelf analysis.

For Load Projects

Data centers, industrial facilities, and large manufacturing projects face a different challenge.

These clients need to know what the grid can support today.

They are not looking for broad ranges of hypothetical outcomes.

They need a current-state answer that supports a real investment decision.

Keentel delivers a focused assessment of:

Available capacity at the target bus
Potential upgrade requirements
Expected interconnection challenges
Current grid limitations

When hundreds of millions of dollars depend on a location decision, precision matters.

WHITE PAPER | Connecting to the Grid Under IEEE 2800

Mon, 15 Jun 2026 17:33:37 GMT

June 15, 2026 | White Paper

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

Connecting to the Grid Under IEEE 2800

Data Center Site Viability | Power & Grid Risks Guide

Mon, 15 Jun 2026 12:03:57 GMT

May 15, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Flexibility is now a viability lever, not a footnote

Representative scenarios

Why this matters now

The buildout will reshape the energy and land-use landscape of every high-growth market in the country. The question for a developer entering it is not whether the opportunity is real — it is whether the site can actually be powered, cooled, flexed, and approved on a timeline and at a cost that works. Power, water, flexibility, and approvals are not problems to solve after site selection. They are the criteria by which a site should be judged in the first place, and every one of them rewards early, engineering-grounded attention and punishes the developer who meets it for the first time during entitlement. The teams that bring that seriousness early are the ones that move fastest. The rest spend the next several years finding out what they committed to.

Before the Load Arrives

The site-viability equation

Site viability turns on four interacting variables, and they have to be assessed together, early, before money is committed:

Power and interconnection — whether the grid can serve the load, on the developer's timeline, at a known and bounded cost.

Water and cooling — whether a reliable, permittable water supply exists for the cooling architecture the compute actually requires.

Flexibility and ride-through posture — whether the load is willing and engineered to be partially curtailable, which increasingly determines how fast and how much it can connect.

Community and approvals — whether the local and political environment will permit the project without turning a routine entitlement into a multi-year fight.

Most first-time entrants underestimate at least two of these. The teams that move projects forward on schedule treat all four as day-one priorities — and they treat the first three as engineering problems with quantifiable answers, not line items to resolve later.

Power: interconnection is an engineering problem, not a utility formality

A data center draws power that dwarfs almost any other single facility on a transmission system. Getting that power to the site means navigating an interconnection process that is multi-year, capital-intensive, and carries no guaranteed outcome — and the entity you negotiate with shapes everything. In ERCOT you deal with the grid operator plus a transmission service provider such as Oncor or AEP, or, in Austin and San Antonio, a municipally owned utility on its own timeline. In PJM, CAISO, NYISO, MISO, and SPP the cast and the procedures change. Knowing which environment you are entering — and what its large-load rules actually say — before the first conversation is the difference between leading the process and reacting to it.

Texas put the large-load question on a statutory footing with Senate Bill 6, enacted in 2025, which directed the PUCT and ERCOT to build interconnection standards for loads of 75 MW or greater. The implementing rule, 16 TAC §25.194, was still in draft as of spring 2026 and must be finalized by year-end. Two features of that draft hit a developer's pro forma immediately. First, the financial commitment is real: the published draft contemplates per-megawatt obligations on the order of $50,000/MW, on top of study fees that SB 6 floors at $100,000, carried through a study process measured in years. A 400 MW request can therefore tie up tens of millions of dollars in non-deployable capital while the study runs. Second, the framework is built to expose speculative demand: developers must disclose duplicate or concurrent requests, and fragmenting a project across parcels to hedge — long a common tactic — now works against you administratively and slows the entire queue.

The mechanism ERCOT is using to bring order to that queue is the coordinated batch study — “Batch Zero,” established through PGRR145 and NPRR1325 — which studies eligible large load s simultaneously against a shared 2032 system case rather than one at a time against a moving target. Keentel has published a separate, detailed engineering walk-through of Batch Zero, PCLR, and BYOG; this paper does not repeat it. The point for site selection is simpler: there is now a hard eligibility cutoff that determines who is studied in this cycle and who waits, and the validity of an earlier study can be overturned when the loads around it change. A large-load study that passed in February can require re-study months later, and a load that assumed firm service can be reclassified to partial-firm. Early-stage assumptions about “we'll get our megawatts” are exactly the assumptions this framework is designed to test.

What actually determines whether a site's interconnection cost is manageable or ruinous is the study stack beneath the process — the part most non-engineering site evaluations skip entirely:

Steady-state (power-flow) analysis establishes whether the surrounding network can carry the load under normal and contingency conditions, and where thermal or voltage limits cap how much can be served.

Short-circuit / fault-duty analysis determines whether existing breakers and protection can handle the fault current the new connection introduces, and what must be replaced.

Dynamic and transient stability analysis examines how the system — and increasingly the load itself — behaves through faults and disturbances. With large electronic loads this is no longer a generator-only concern.

Reactive and voltage studies size the reactive support the point of interconnection requires to hold voltage within limits.

Ride-through compliance — in ERCOT, the large-electronic-load requirements advanced through NOGRR282 — governs how the facility must behave during grid disturbances rather than tripping offline and turning a local event into a system event.

Each of these can swing required upgrades from a modest local scope to material backbone work hundreds of miles upstream, with the cost assigned to the developer. A site sitting near a constrained corridor, or downstream of a path that needs reinforcement, can carry an interconnection cost that dwarfs the land. None of that is visible from a plat map. It is visible from a power-flow case — which is why the most reliable thing a developer can do before committing is have the interconnection studied, not assumed.

How Keentel Engineering helps

Community, permitting, and the risks no study can resolve

A developer can complete every interconnection study, secure every water agreement, and still lose the project at a public hearing. Communities are markedly more informed than they were a few years ago, more organized when they feel bypassed, and increasingly willing to use the entitlement process to demand answers about water, infrastructure, and quality of life. Several patterns recur:

Water use draws the most direct policy attention. Arriving at the first community conversation with a reclaimed-water solution already structured is a fundamentally stronger position than promising to address sourcing later in the engineering phase.

Noise now surfaces on nearly every project. Generators, cooling towers, and continuous mechanical systems at data center scale carry into surrounding neighborhoods. The mitigations — acoustic design, sound walls, exhaust routing, building-envelope choices — are well understood and inexpensive when designed into the site plan, and expensive and adversarial when first raised at a hearing.

Economic framing matters because data centers create limited permanent employment but transformational tax base. A facility placed on land that generated little property tax can be structured, through mechanisms such as tax increment financing, to direct new revenue toward local priorities — a far better story to tell early than to defend late.

Precedent can stall an otherwise-supportive official who realizes their approval sets the terms for every similar request that follows. This dynamic is now playing out beyond Texas, and it turns routine entitlements into policy deliberations.

These are not Keentel's engineering disciplines, and this paper does not pretend they are. But they decide outcomes, and they have to be coordinated with the technical work from day one — because a project that handles the engineering flawlessly and the politics poorly does not get built.

Before you commit: a pre-commitment diligence checklist

Everything in this paper rewards early engagement. The developers who achieve the best timelines are, almost without exception, the ones who answered these questions before they were financially committed to a site.

Have you confirmed, in a real conversation with the grid operator and transmission or utility provider, that they can serve your load estimate on your timeline — not just that a line runs nearby?
Do you know whether your project triggers a formal large-load study (the LLIS/Batch process in ERCOT, or its analog in your target region), and have you modeled the financial-security and study-fee exposure under current — and draft — rules?
Has the interconnection actually been studied — steady-state, short-circuit, stability, and reactive — far enough to bound the upgrade cost, including potential upstream backbone work, rather than assuming it?
Is the load engineered for ride-through compliance (e.g., NOGRR282 in ERCOT), and have you evaluated a flexible posture — a PCLR-style firm/flexible split, BYOG co-location, or behind-the-meter generation — as a way to connect faster and cheaper?
What is the complete water picture — municipal capacity, groundwater limits, and reclaimed-water feasibility — and is there a wastewater treatment plant within viable distance?
Does the cooling architecture match the compute? Has liquid or immersion cooling been evaluated as both a water-reduction strategy and a hardware requirement for the systems being deployed?
Has a noise analysis been completed against surrounding land uses, with mitigation built into the site plan rather than deferred to a later design phase?
Have the local and state stakeholders been identified and engaged informally before the formal entitlement clock starts — and is the tax-base story structured before opposition forms around a different one?

The following are illustrative composites, not specific projects, intended to show how the variables interact. Detailed engineering case studies are available in Keentel's Batch Zero analysis.

Scenario A — A hyperscale campus that flexes to connect

A developer plans an 800 MW campus ramping over several years in a zone with moderate transmission and no committed backbone upgrades on its timeline. An early study indicates the full load cannot be served as firm without major upgrades — but a firm allocation supports the early phases, and the balance can be carried as a flexible (PCLR-style) portion that ramps as headroom appears. Designing for that split at site selection, rather than discovering it after a study, preserves the path to full capacity and lets construction begin years sooner than a firm-only assumption would allow.

Scenario B — A load that brings its own generation

A site needing several hundred megawatts sits four to five years from the transmission required to serve it as firm load. By co-locating solar plus storage (and, where appropriate, gas) under a BYOG configuration — studied as grid-supplied for withdrawal, with the generation interconnected separately and the integrated facility assessed for full system impact — the load energizes on the generator's schedule under self-limited withdrawal, then transitions to standard service once upgrades complete. The trade-off is real operational-technology investment to schedule load against generation availability, which has to be planned, not assumed.

Conclusion

Technical FAQ

The grid is in the middle of the largest load-growth event in its modern history. Hyperscale data centers, AI training campuses, cryptocurrency facilities, and newly electrified industrial sites are arriving at the interconnection door in volumes the planning machinery was never built to absorb in a single cycle. The clearest window into what that means for a developer is ERCOT, where the active large-load interconnection queue reached roughly 226 GW by late 2025 — nearly four times the level reported a year earlier — with about three-quarters of it tied to data centers targeting energization by 2030. By early 2026, counting requests not yet reflected in the published queue, the figure was pushing past 380 GW. For perspective, that is several times the system's existing generation, and over the same window only on the order of a couple thousand megawatts of new large load were actually approved to energize. The queue is enormous; the throughput is not.

That gap — between what developers are requesting and what the system can physically deliver and reliably study — is the single most important fact for anyone evaluating a site. It is why a parcel with the right zoning, the right price, and good access can still be an unbuildable data center. The factors that decide viability mostly sit upstream of the property line: in the transmission system, in the water utility, and in a regulatory framework being rewritten in real time. ERCOT is the most advanced proving ground, but the same pressures are now shaping PJM, CAISO, NYISO, and every other high-demand region. The rules differ; the underlying engineering, and the discipline required to get ahead of it, do not.

This paper is written for the developers, design-build contractors, and corporate real estate teams entering this space who need a clear, engineering-grounded picture of what a site actually requires before they commit. The consequences scale with project size — from a 5–10 MW enterprise deployment to a gigawatt-class campus — but the questions do not change.

The old model treated a data center as an inflexible, must-serve block of firm load. The grid can no longer absorb hundreds of gigawatts of that on demand, so the rules — and the economics — now reward loads that can flex. For a developer, a flexible posture is increasingly the difference between energizing in two years and waiting for backbone transmission that is five to seven years out. Three levers matter:

Provisional Controllable Load Resources (PCLR) split a load into a firm portion that can always be served (Low Power Consumption, LPC) and a flexible portion up to the studied ceiling (Maximum Power Consumption, MPC) that is dispatched as transmission headroom allows. A campus can begin ramping on its firm allocation immediately while the flexible megawatts come online as upgrades complete, rather than waiting for the full upgrade plan before consuming anything above firm.

Bring Your Own Generation (BYOG) co-locates on-site generation — gas, solar, storage, or a combination — with the load, letting the facility energize on the generator's schedule and reducing dependence on constrained transmission. It is studied through a coordinated, multi-workstream framework, because a co-located facility looks different to the load study, the generation interconnection study, and the integrated transmission study.

Behind-the-meter generation more broadly — on-site gas, solar, battery storage, or a microgrid — can offset the grid interconnection request and meaningfully change the financial exposure that follows. The trade-off: any generation that injects to the grid pulls the generator-side process (in ERCOT, the Generator Interconnection or Modification process and its Full Interconnection Study under Planning Guide Section 5) into scope, with its own steady-state, short-circuit, stability, and reactive studies, its own dynamic-model and commissioning requirements, and — for thermal generation — its own water-supply obligations.

The strategic implication is that flexibility should be designed in at site selection, not bolted on after a study comes back unfavorable. A load engineered from the outset to ride through disturbances, to curtail a defined flexible portion, and potentially to lean on co-located generation is a fundamentally more interconnectable — and more financeable — project than an identical load presented as rigid firm demand.

Water: the constraint you cannot engineer around as easily

Power capacity can be added — built behind the meter, contracted, flexed. Water supply cannot be manufactured. It has to be secured from existing infrastructure, and many municipalities simply do not have the volume a large facility requires. This is the constraint that surprises developers most, and the one that compounds: most conventional on-site power generation at scale also consumes water, so a site with a water problem cannot always solve it by generating its own power. The two constraints are linked, and the linkage only becomes obvious after the fact unless someone is looking for it early.

Cooling architecture drives the number. Traditional evaporative and air-cooled systems are the most water-intensive; direct-to-chip liquid cooling and immersion cooling sharply reduce consumption and, for current high-density AI hardware, are increasingly a design requirement rather than an option. A facility built around high-density compute has to account for that from the first cooling concept, because it changes both the water demand and the mechanical and electrical design around it.

Groundwater is rarely available at the volumes a major facility needs; conservation districts govern most significant aquifers and permitted extraction usually falls well short. In many markets the most viable path is reclaimed water: a site positioned near both adequate transmission and a municipal wastewater treatment plant can partner with the utility to produce reclaimed water at industrial quality under a long-term purchase agreement — turning a likely objection into a durable community benefit and a new revenue stream for the city. Reclaimed-water infrastructure is still uncommon, which means in most markets this conversation has not happened yet, and the developer who initiates it holds the advantage. Like interconnection, a reclaimed-water arrangement is measured in years from first conversation through engineering, permitting, and offsite infrastructure — which is why water has to run in parallel with power from the outset, not as a second phase.

Keentel Engineering is a power-systems and grid-interconnection engineering firm. The challenges in this paper are the work we do every day, across ERCOT, CAISO, NYISO, and other ISO/RTO regions, and we stay current with frameworks — Batch Zero, NOGRR282, and the evolving ride-through and modeling standards — that change faster than any static reference can capture. We help developers and their partners de-risk a site before commitment and carry it through energization:

Power system studies (EHV/HV/MV): load flow, short-circuit and fault duty, dynamic and transient stability, reactive/voltage, harmonics, and arc-flash — the analyses that turn an assumed interconnection cost into a known one.

POI interconnection engineering support: study coordination with the grid operator and transmission provider, model development and quality, and management of the large-load (and, where generation is involved, generator) interconnection process.

Substation and transmission line design: the physical interconnection from the point of interconnection to the site.

Utility-scale renewables and storage engineering (solar and BESS): the engineering behind behind-the-meter and BYOG strategies that change a project's interconnection profile.

Owner's engineer services: independent design review, commissioning support, and bankability assurance across the project.

NERC O&P (693) compliance: keeping interconnected facilities audit-ready.

Our value is integration. When interconnection engineering, study management, substation design, and generation strategy share a table from the start, the expensive problems get identified before they become commitments. Wherever you are — initial site screening, a stalled project, or a study that came back worse than expected — we engage from there.

Generation Injection & Large Load Withdrawal Studies

Thu, 11 Jun 2026 22:24:59 GMT

Jun 11, 2026 | Blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

1. Understanding Transmission Interconnection

Keentel Engineering provides end-to-end client-side interconnection study services across six core technical disciplines. Each discipline addresses a specific aspect of the interconnection analysis required to select the optimal POI, prepare complete application documentation, and support the client through the utility/ISO process.

3.1 POI Screening & Selection

The starting point of every interconnection engagement is the identification and ranking of candidate Points of Interconnection. Keentel's POI screening process is a structured, multi-stage analysis that evaluates candidate nodes across the applicable transmission network against a comprehensive set of technical and commercial criteria.

Stage 1 — Geographic & Network Screening

Candidate POIs are identified by filtering all substations and transmission nodes within the relevant geographic area at applicable voltage levels. The initial candidate pool is assessed against available capacity headroom, planned network changes, active interconnection queue positions, and estimated network upgrade cost exposure. This stage typically narrows a broad initial pool of 15–25 candidates to a focused shortlist of 5–8 technically viable POIs.

Stage 2 — Power Flow Validation in PSS/E

Shortlisted POIs are validated using PSS/E — the industry-standard transmission planning and simulation software — loaded with current and future-year network data for the applicable ISO/utility system. New facility models for the generation injection and/or load withdrawal are built and injected into the network model at each candidate POI, and preliminary power flow checks confirm basic technical feasibility.

Stage 3 — Ranked POI Recommendation

All evaluated POIs are ranked and presented in a comparative POI Ranking Table scoring each candidate against: available capacity, estimated network upgrade cost, queue position, voltage compliance, system strength, application timeline, and jurisdiction suitability. A clear, justified recommendation for the optimal POI is provided — giving the client a single, well-supported answer rather than a list of options to navigate independently.

3.2 Steady-State Power Flow Analysis

AC power flow simulations are performed in PSS/E across the full range of facility operating scenarios to assess the transmission network's response to the new injection or withdrawal under normal conditions:

Peak load (Heavy Summer) and light load (Light Winter) study cases
Maximum and partial generation export scenarios (for generation injection)
Maximum and partial load demand scenarios (for load withdrawal)
Combined simultaneous operation scenarios (for co-located facilities)
Net import and net export conditions across all combined operating modes

Power flow results identify thermal overloads on transmission lines and transformers, voltage violations at the POI and adjacent buses, and the first limiting system element that constrains load service or generation deliverability. These findings directly inform POI selection and preliminary network upgrade identification.

3.3 N-1 Contingency Screening

N-1 contingency screening is performed in PSS/E to assess the network's response to the loss of a single transmission element under the new facility's operating conditions. This is a client-side pre-screening study to support POI selection — it is not a full NERC TPL-001 transmission planning study, which is the utility/ISO's responsibility.

The contingency screening set includes key transmission line outages, transformer outages, largest generator loss events, and facility-specific contingencies. For each contingency, PSS/E evaluates thermal loading on key transmission elements, bus voltage levels near the POI, and system stability indicators. Results are used to rank POI candidates and identify preliminary network upgrade requirements.

3.4 Short Circuit & Breaker Duty Analysis

Short circuit analysis quantifies the fault current contribution from the new generation or load facility at each candidate POI and assesses its impact on existing network equipment. This is performed as a client-side pre-screening study using PSS/E's short circuit module and serves two purposes: informing POI selection and developing the short circuit models required for the interconnection application package.

For Generation Injection

Generator fault current contribution is modeled using correct subtransient reactance and fault current injection profiles. Three-phase bolted and single line-to-ground fault studies are performed at each candidate POI to assess fault current levels, breaker interrupting duty adequacy, and bus fault duty compliance.

For Load Withdrawal

Large industrial loads — particularly those containing arc furnaces, large motors, and variable-speed drives — can make significant fault current contributions that must be accurately modeled. Composite load impedance and motor contribution characteristics are incorporated into the fault study model to produce a realistic assessment of the facility's fault current impact.

3.5 Reactive Power & Voltage Compliance Assessment

Reactive power management is one of the most critical — and most frequently underestimated — aspects of both generation injection and large load withdrawal interconnections.

For Large Load Withdrawal

Steel plants, data centers, manufacturing facilities, and other large industrial loads typically operate at lagging power factors and impose significant reactive power demands on the transmission network. Keentel assesses reactive power requirements across the full load range, evaluates voltage regulation performance at the POI, and sizes reactive compensation systems — including capacitor banks, STATCOMs, and synchronous condensers — to meet ISO/utility voltage and power factor standards. Where arc furnaces are present, dynamic reactive compensation requirements and flicker emission assessment are also addressed.

For Generation Injection

Generators are required by MISO, TVA, and other ISOs to demonstrate reactive capability compliance through P-Q curve analysis. Keentel assesses the generator's full reactive capability envelope, evaluates voltage control mode options, and prepares the reactive capability documentation required for the interconnection application. Where the generator's native reactive capability is insufficient to meet ISO requirements, supplemental reactive support options are identified and sized.

3.6 System Strength & Short Circuit Ratio (SCR) Assessment

System strength — measured through Short Circuit Ratio (SCR) and Weighted Average Short Circuit Ratio (WASCR) — is an increasingly critical study requirement for generator interconnections as the North American grid transitions to higher proportions of inverter-based resources. A low SCR at the POI can cause voltage instability, active power oscillations, and protection coordination failures that threaten both the generating facility and the broader network.

Keentel's SCR assessment calculates the short circuit capacity at each candidate POI relative to the rated MW output of the generation facility, evaluates interactions with nearby inverter-based resources, and compares SCR values across Heavy Summer and Light Winter study cases. Where SCR thresholds are not met, Keentel identifies mitigation options including alternative POI selection, synchronous condenser installation, and grid-forming inverter control strategies.

A comparative SCR assessment across multiple candidate POIs is one of the most valuable outputs of Keentel's pre-screening study — enabling the client to select a POI with adequate system strength before investing in formal utility study costs.

3.7 PSS/E Dynamic Model Development

PSS/E dynamic models are a mandatory submission requirement for formal interconnection applications with MISO, TVA, and most other ISOs and utilities. These models allow the utility and ISO to simulate the transient behavior of the new facility under system disturbances — a requirement that cannot be waived and must be completed before an application can advance to System Impact Study.

Power Plant — Generation Dynamic Models

Keentel develops complete PSS/E dynamic model packages for generation facilities including:

Machine model: generator electromechanical characteristics (GENROU, GENSAL, or equivalent)

Excitation system model: automatic voltage regulator and excitation control (EXAC1, ESST1A, or equivalent)

Governor / prime mover model: turbine-governor speed control response (GGOV1, IEEEG1, or equivalent)

Power System Stabilizer (PSS): oscillation damping controls where required by the applicable ISO/utility

Inverter-based resource (IBR) models: generic renewable energy system models (REGC_A, REEC_A, REPC_A) or manufacturer user-defined models (UDMs) for solar, wind, and battery storage facilities

Steel Plant / Industrial Load — Composite Load Dynamic Models

Accurate composite load dynamic models are required for large industrial loads and are developed by Keentel to include:

Composite load model (CMPLDW or CMLD): representing the full industrial load mix including arc furnaces, large induction motors, and other load components

Motor load representation: large motor load fraction, inertia, and protection settings affecting post-fault voltage recovery

Arc furnace dynamic model: flicker generation and dynamic reactive power demand characteristics

Load protection models: under-voltage load shedding and motor protection relay representations

All dynamic models are validated through standard PSS/E simulation test cases — including 3-phase fault, SLG fault, and generator trip events — and packaged in the format required by the applicable ISO/utility for formal application submission.

3.8 Interconnection Application Support

Keentel provides complete interconnection application support from pre-application strategy through submission-readiness review. This covers:

Application strategy: advising on the optimal application pathway — including hybrid vs. separate applications for co-located facilities and MISO vs. TVA jurisdiction selection

Application form preparation: completing all applicable ISO/utility application forms including MISO Generator Interconnection Request, TVA Transmission Service and Generator Interconnection forms, and large load application forms as applicable

Technical appendices: facility one-line diagrams, POI description, load/generation profiles, operating assumptions, preliminary protection philosophy, and reactive capability documentation

Pre-application meetings: leading or supporting pre-application meetings with the utility/ISO to confirm POI, discuss study assumptions, and align on application scope

Submission-readiness review: final completeness check against ISO/utility requirements before submission

Post-submission coordination: responding to utility/ISO technical information requests, reviewing study notices, and providing technical advisory support throughout the formal study process

2. The Transmission Interconnection Process

Generation Injection & Large Load Withdrawal

Executive Summary

The choice of Point of Interconnection is the single most consequential technical and commercial decision in the entire interconnection process. It determines the network upgrade costs the project will be required to fund, the interconnection timeline, the system strength environment the generation facility will operate in, and ultimately the economic viability of the project. Yet many developers and industrials underinvest in pre-application POI analysis — proceeding directly to formal utility application without adequate pre-screening, only to encounter unexpected and costly network upgrade requirements at the System Impact Study stage.

4.1 The Cost Impact of POI Selection

Network upgrade costs — which the interconnecting party is typically required to fund as a condition of interconnection — can vary by tens of millions of dollars depending on POI selection. A POI that requires reconductoring of a heavily loaded transmission line will carry significantly higher upgrade costs than an adjacent substation with adequate spare capacity. Without rigorous pre-screening, developers risk entering the formal utility study process at a POI that is commercially unviable — incurring application fees, study fees, and delays before the true cost exposure is revealed.

Keentel's POI screening identifies the lowest-cost, highest-capacity interconnection pathway before a single dollar is spent on formal utility study fees — enabling confident go/no-go decisions early in the project development process.

4.2 The Queue Position Challenge

The interconnection queue is a critical factor in POI selection that is often overlooked. Each ISO and utility maintains a queue of pending interconnection applications, and projects earlier in the queue have priority over network capacity and upgrade cost allocation. A POI that appears technically attractive may be heavily contested by prior queue positions, resulting in high estimated upgrade costs that reflect the project's queue position relative to competing applications.

Keentel's pre-screening process incorporates real-time queue position data for all candidate POIs, enabling the client to identify nodes with minimal queue exposure and better upgrade cost prospects — a factor that can make or break the economic case for a given POI.

4.3 Dual Jurisdiction Considerations

Projects located in areas served by multiple transmission owners or ISOs — such as the boundary between MISO and TVA territory in the Southeast, or areas where multiple utilities overlap — must evaluate POI options across both jurisdictions. The optimal POI may lie in a different jurisdiction than initially assumed, with significantly different upgrade cost profiles, queue dynamics, and application timelines. A thorough cross-jurisdictional POI screening is essential to ensure the best available option is identified regardless of utility boundaries.

Connecting a new generation facility or large industrial load to the transmission grid is one of the most technically complex, commercially consequential, and time-sensitive activities in the energy sector. Whether the project involves a power plant seeking to inject energy into the grid or a large industrial facility requiring substantial load withdrawal — such as a steel plant, data center, or manufacturing complex — the interconnection process demands deep technical expertise, precise engineering analysis, and thorough knowledge of utility and ISO procedures.

This white paper provides a technical overview of the generation injection and load withdrawal interconnection study process, explains the key analytical disciplines involved, and describes how Keentel Engineering's specialized services guide developers, industrials, and utilities through every phase — from initial Point of Interconnection (POI) screening to final application submission.

Keentel Engineering's interconnection study services are designed to deliver one clear outcome: identify the best available Point of Interconnection for the client's asset, prepare all required technical models and documentation, and support a successful interconnection application — faster and more cost-effectively than conventional approaches.

1.1 What Is a Transmission Interconnection?

A transmission interconnection is the physical and contractual connection of a new facility — either a generation source or a large load — to the high-voltage transmission network. This connection allows the facility to exchange power with the broader grid: generators inject power into the network for delivery to consumers, while large loads withdraw power from the network to meet their operational demand.

Unlike distribution-level connections, transmission interconnections involve high-voltage infrastructure (typically 69 kV and above), complex power flow interactions with the broader regional network, and formal regulatory processes administered by Independent System Operators (ISOs), Regional Transmission Organizations (RTOs), and Transmission Owners (TOs). The interconnection process is governed by FERC regulations, NERC reliability standards, and the specific tariffs and procedures of the applicable utility or ISO.

1.2 Generation Injection vs. Load Withdrawal — Key Differences

While both generation injection and load withdrawal interconnections share many common study requirements, they present distinct technical challenges that must be addressed through specialized analysis:

5. Technical Deep Dive — Key Study Concepts

2.1 Overview of the Process

The interconnection process follows a structured sequence of activities — from initial site and POI assessment through formal application, utility study, and ultimately Interconnection Agreement execution. Understanding this process is essential for developers and industrials to plan their project timelines and budgets effectively.

3. Keentel Engineering's Interconnection Study Services

4. Why POI Selection Is the Most Important Decision

5.1 Power Flow Fundamentals for Interconnection

AC power flow analysis is the foundation of all transmission interconnection studies. In PSS/E, the network is represented as a system of buses (nodes) connected by branches (transmission lines and transformers), each with defined impedance, thermal rating, and voltage characteristics. The power flow solver iterates to find the steady-state voltage magnitudes and angles at all buses that satisfy Kirchhoff's laws and the network's physical constraints.

For a new generation injection, the facility is represented as a PV bus — where real power output and terminal voltage are specified, and the solver calculates the reactive power exchange and network voltage profile. For a large load withdrawal, the facility is represented as a PQ bus — where real and reactive power demand are specified, and the solver calculates the voltage response at the POI and throughout the network. The key outputs for POI assessment are:

Thermal loading on all transmission elements as a percentage of their normal and emergency ratings
Voltage magnitudes at all buses — violations of ANSI C84.1 voltage standards are flagged
Reactive power flows and reactive margin at the POI
Real and reactive power losses in the transmission system

5.2 Short Circuit Fundamentals

Short circuit analysis calculates the fault current that would flow in the network following a bolted fault at a specific bus. The two most important fault types for interconnection studies are:

Three-phase bolted fault (3φ): the symmetrical fault that produces maximum fault current at the fault location — used to assess breaker interrupting capability
Single line-to-ground fault (SLG): typically the most severe asymmetrical fault type for protection system design and zero-sequence analysis

For a new generation facility, fault current contribution is determined by the generator's subtransient reactance (X"d) and the network impedance between the generator and the fault location. For a large industrial load containing motors, the motor impedance and inertia determine the magnitude and duration of the motor's fault current contribution — which can be significant for large electric arc furnace and motor drive installations.

The key metric from short circuit analysis is the X/R ratio at the POI — a high X/R ratio indicates a predominantly inductive fault current that stresses breaker interrupting capability. Breaker interrupting duty is assessed by comparing the calculated fault current against the breaker's rated symmetrical interrupting current, adjusted for the X/R ratio per IEEE C37 standards.

5.3 System Strength & SCR — Technical Explanation

Short Circuit Ratio (SCR) is defined as the ratio of the short circuit MVA at the POI to the rated MVA of the connecting generation facility. A high SCR indicates a strong grid — one that is relatively unaffected by the generator's output fluctuations. A low SCR indicates a weak grid, where the generator's output can cause significant voltage fluctuations and where inverter-based resources may experience control instability.

The NERC guideline threshold for minimum SCR is generally 1.5 for inverter-based resources, though specific requirements vary by ISO and utility. MISO, for example, has published specific system strength assessment requirements tied to its interconnection procedures. Where SCR falls below the minimum threshold, the following mitigation options are typically evaluated:

Synchronous condenser: a rotating electrical machine that provides reactive power and increases short circuit capacity at the POI — directly raising the SCR
Grid-forming inverter controls: advanced inverter control strategies that allow inverter-based resources to form voltage references in weak grid conditions — an emerging technology increasingly accepted by utilities
Alternative POI selection: moving the interconnection to a stronger point in the network with higher inherent short circuit capacity
Phased generation commissioning: initially operating at reduced output levels until network strengthening upgrades are completed

5.4 Reactive Power & Power Factor — Why It Matters for Large Loads

Large industrial loads — particularly steel plants with electric arc furnaces — impose variable and often highly reactive power demands on the transmission network. This has two significant consequences: power factor penalties imposed by the utility for operating below a specified minimum power factor, and voltage depression and flicker on the network that can affect other customers served by the same transmission infrastructure.

The power factor of an electric arc furnace load varies dynamically as the furnace operates through its melting cycle, generating rapid and large reactive power swings that manifest as voltage flicker on the network. Flicker assessment — quantified using the IEC flicker severity indices Pst (short-term flicker severity) and Plt (long-term flicker severity) — is a standard utility requirement for arc furnace interconnections and directly drives the sizing of dynamic reactive compensation (STATCOM or SVC) requirements.

For generation facilities, reactive capability compliance is a contractual requirement in virtually all ISO/utility interconnection agreements. The generator must demonstrate the ability to operate within a specified power factor range at its high-side bus across its full real power output range — typically 0.95 power factor leading to 0.90 power factor lagging at rated MW output. Failure to demonstrate this capability during the formal study process can delay the interconnection agreement and trigger requirements for additional reactive support equipment.

The Texas grid is being asked to host more new load in the next five years than it has built in the last fifty. The legacy single-study, firm-service-only interconnection model cannot deliver that capacity in time.

BYOG, CLR, and WLPUN are the three frameworks Texas is using to bridge the gap — letting Large Loads come online with self-supply and as-available grid service while transmission catches up. They are not interchangeable, and the regulatory boundaries between them are still being drawn through the Batch Zero process scheduled for finalization in mid-2026.

For developers, the implication is clear: the framework you choose determines your speed-to-power, your operational obligations, your settlement exposure, and your capital deployment risk. Understanding the distinctions between BYOG, CLR, and WLPUN — and the constructs that sit beneath each — is no longer a regulatory nicety. It's a core part of building large-scale compute infrastructure in ERCOT.

7. Frequently Asked Questions (FAQ)

Final Thoughts

1.3 The Co-Located Facility Challenge

Projects involving both a large generation facility and a large industrial load at the same site — such as a power plant co-located with a steel plant or industrial complex — present a uniquely complex interconnection scenario. The combined injection and withdrawal must be studied simultaneously to understand:

Net import and net export operating conditions across all combinations of generation output and load demand
The facility's aggregate impact on the transmission network under all operating modes
Whether a hybrid interconnection application or separate applications for each facility is the more efficient approach
How the generation injection offsets or exacerbates the load withdrawal impacts on the network

Keentel Engineering specializes in this combined facility interconnection scenario, having developed structured analytical workflows that address the full complexity of simultaneous injection and withdrawal studies within a single, integrated study framework.

The following questions are commonly raised by developers, industrial clients, and project owners during the interconnection study process. Keentel Engineering's answers are provided to help clients better understand the technical and commercial dimensions of the process.

6. How Keentel Engineering Supports Your Project

Keentel Engineering serves as the client's dedicated technical partner throughout the pre-application and application phases of the interconnection process. Our role is to ensure that the client enters the formal utility study process with the best possible POI, the most complete and technically robust application, and the highest likelihood of a successful interconnection outcome.

Keentel Engineering's interconnection study services are built on one principle: the client's best interests come first. We provide honest, technically rigorous analysis — not the answer the client wants to hear, but the answer they need to make informed decisions.

CAISO Appendix H Compliance Guide for IBRs

Thu, 11 Jun 2026 20:42:55 GMT

Jun 11 , 2026 | blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Navigating CAISO Appendix H

What Appendix H Covers

What is T&D Co-Simulation?

Confusing Physical Connections with Logical Nodes in IEC 61850

Keentel Engineering brings over three decades of utility-scale power system experience, with engineers licensed in California and across the U.S., and a practice built specifically around IBR interconnection and NERC/ISO compliance. Here's how that maps to the Appendix H requirements.

Dynamic and EMT Modeling

Compliance demonstration starts with credible models. Keentel develops and validates PSSE, PSCAD, TSAT, and PowerFactory models for power flow, dynamic stability, and electromagnetic transient (EMT) analysis — the foundation for proving ride-through performance to CAISO during model review and interconnection studies .

Ride-Through and Dynamic Stability Studies

Keentel performs the voltage and frequency ride-through studies that show a plant will remain online through the fault and recovery profiles Appendix H specifies — and that its behavior aligns with the current NERC ride-through framework (PRC-029-1 / PRC-024-4) as well as IEEE 2800-2022 where applicable.

Inverter and Plant Controller Tuning

To resolve the tuning trade-offs that derail compliance, Keentel optimizes inverter control parameters and plant-level control logic — including PLL tuning and coordinated low-voltage ride-through behavior — so reactive current injection, recovery ramp rates, and the one-second transition requirement are met without sacrificing stability.

Protection Coordination and Relay Integration

Keentel reviews protection philosophy, single-line diagrams, and relay logic (with deep SEL relay expertise) to ensure protection settings don't conflict with ride-through obligations or trip the plant unnecessarily during recoverable disturbances — a frequent root cause of Appendix H findings.

PRC-029-1 / PRC-024-4 Alignment and Exemptions

For owners reconciling Appendix H with the newer NERC standards, Keentel provides relay tuning and validation to align voltage and frequency trip points with PRC-024-4, prepares PRC-029-1 exemption justification packages for legacy IBRs that cannot meet current criteria, and integrates PRC-028-1 disturbance monitoring for audit readiness.

Disturbance Monitoring and Transient Data Recording

To satisfy Appendix H's transient data recording obligations, Keentel supports the design and validation of GPS-synchronized monitoring, phase angle measuring units, and the recording architecture needed to capture, retain, and deliver event data within CAISO's required windows.

Interconnection Studies and Owner's Engineer Support

From feasibility through commissioning, Keentel acts as an independent Owner's Engineer — validating EPC designs, verifying interconnection compliance, and protecting long-term asset value across CAISO and other ISO/RTO territories.

Audit Support and Documentation

Keentel guides clients through the full NERC Align workflow — pre-submittal reviews, mitigation plans, evidence packages, and Regional Entity correspondence — and delivers audit-ready documentation, validation reports, and coordination checks.

Appendix H and PRC-029-1: A Converging Landscape

Where Projects Commonly Struggle

A common question for California asset owners is how Appendix H relates to the NERC PRC-029-1 standard. The short answer: they share the same intent — keeping IBRs online and supportive during disturbances — but they are distinct regimes that must be reconciled, not assumed to be identical.

PRC-029-1, Frequency and Voltage Ride-through Requirements for Inverter-Based Resources , was approved by FERC in 2025 (Order No. 909) alongside PRC-024-4 for synchronous machines. It introduces performance-based, continent-wide ride-through requirements specifically tailored to IBR behavior, closing gaps left by the older PRC-024 framework. It defines ride-through zones, mandates that IBRs not cease current exchange within those zones, and requires rapid return to normal current exchange after voltage recovery.

The practical implications for a CAISO project:

Overlapping but not identical thresholds

Appendix H and PRC-029-1 both govern ride-through, but specific timing, voltage thresholds, and recovery windows differ in their details. A plant tuned only to one may not automatically satisfy the other.

Different origins and enforcement

Appendix H is a contractual interconnection provision in the LGIA; PRC-029-1 is a mandatory NERC reliability standard with its own compliance, evidence, and audit obligations (including legacy-unit exemption pathways under Requirement R4).

A moving target

Appendix H's frequency provision explicitly points to successor NERC standards, so alignment with the current NERC ride-through regime is part of staying compliant over time.

The takeaway: describe the relationship carefully in compliance documentation. Saying the Appendix H ride-through obligation is addressed under the current NERC ride-through framework is defensible; treating the two as equivalent is not, given the threshold and origin differences.

A Compliance Checklist for California Asset Owners

Conclusion

Why Appendix H Matters Now More Than Ever

Appendix H is where California's clean-energy ambitions meet grid-engineering reality. Meeting it requires more than installing compliant inverters — it takes validated models , carefully tuned controls, coordinated protection, robust data recording, and documentation that holds up under both CAISO and NERC scrutiny. With ride-through standards converging across PRC-029-1, IEEE 2800-2022, and the CAISO tariff, the cost of getting it wrong — interconnection delays, curtailment, or compliance findings — keeps rising.

Technical FAQ

How Generator Owners can meet the technical requirements of the CAISO Large Generator Interconnection Agreement — and where the right engineering partner makes the difference.

Appendix H applies to all Asynchronous Generating Facilities interconnecting to CAISO, with a grandfathering carve-out: existing units already interconnected at the same location are generally exempt for the remaining life of that unit (subject to a narrow exception under CAISO tariff Section 25.4.2). The requirements break into six technical areas.

1. Voltage Ride-Through Capability

This is the heart of Appendix H. The facility must remain online through voltage disturbances rather than tripping offline and worsening a grid event. Key provisions include:

Three-phase faults: Remain online for the lesser of the normal three-phase clearing time (4–9 cycles) or 150 milliseconds, plus subsequent recovery to steady-state voltage. Clearing time is referenced to faults that drive the Point of Interconnection (POI) voltage to 0.2 per unit or below.

Single-phase faults: Remain online through delayed (backup) clearing for single-line faults, accounting for a single point of protection or breaker failure.

Momentary cessation: Pausing current injection during a fault is prohibited unless transient high voltage reaches 1.20 per unit or more.

Reactive current injection: During low-voltage events, inverters must inject reactive current proportional to the voltage dip, reaching full reactive capability when terminal voltage falls to 0.50 per unit or below.

Recovery and ramp: Once voltage returns to the normal band (0.90–1.10 per unit), inverters must transition back to real-power injection, ramping at a minimum of 100% per second, with the full transition completed in one second or less.

Reconnection: After a trip, the inverter must make at least one resynchronization attempt within 2.5 minutes, unless a manufacturer-defined fatal fault code applies.

Phase-lock-loop behavior: Inverters may not trip for momentary loss of synchronism; controls must hold the last synchronized phase and continue injecting current until synchronism is regained (with a 150-millisecond limit before a trip is permitted).

Plant controllers: Plant-level controllers must not impede the inverters' automatic, rapid re-synchronization and ramp-up after a disturbance.

2. Frequency Ride-Through Capability

Appendix H defers off-nominal frequency performance to the applicable NERC Reliability Standard for generator frequency and voltage protective relay settings — "or successor requirements as they may be amended." That forward-looking language is important, because the governing NERC standards have recently changed (more on that below).

3. Power Factor / Reactive Power

Facilities must operate within a power factor range of 0.95 leading to 0.95 lagging, measured at the high side of the substation transformer, to maintain a specified voltage schedule. The precise obligation depends on whether the facility was studied under the Independent Study Process (Appendix DD, Section 4) and what the Phase II Interconnection Study determined was needed for safety and reliability. Dynamic voltage support may also be required in lieu of a power system stabilizer.

4. SCADA Capability

The facility must provide SCADA to transmit data to and receive instructions from the Participating Transmission Owner (TO) and CAISO, with the specific data set scoped to the plant's size, characteristics, location, and reliability importance.

5. Power System Stabilizers (PSS)

Not required for Asynchronous Generating Facilities.

6. Transient Data Recording

Facilities larger than 20 MW must monitor and record detailed plant- and inverter-level data for ride-through events, reactive current injection, momentary cessation, and inverter trips. The data requirements are demanding:

Time-synchronized to 1-millisecond resolution via GPS clock.
Sampled at least every 10 milliseconds (except phase-angle data).
Each record captures at least 150 ms pre-trigger and 1000 ms post-trigger.
Data stored for a minimum of 30 days and delivered within 10 calendar days of a CAISO/TO request.
A phase angle measuring unit (≥16 samples per cycle) installed at the facility entrance or main substation transformer.

How Keentel Engineering Supports CAISO Appendix H Compliance

Before energization and before an audit confirm:

Inverter and plant models validated against field/EMT behavior and accepted by CAISO

Voltage ride-through performance demonstrated for both three-phase and single-phase fault profiles

Reactive current injection proportional to voltage dip, full capability at ≤0.50 pu

[Recovery ramp ≥100%/second, full transition ≤1 second

[Momentary cessation limited to the permitted high-voltage band (≥1.20 pu)

[Plant controller programmed to allow rapid, automatic inverter restoration

Power factor capability of 0.95 leading to 0.95 lagging at the HV side, per Phase II study findings

SCADA scope agreed with the Participating TO and CAISO

Transient data recording in place for facilities >20 MW (GPS sync, sampling, retention, turnaround)

Phase angle measuring unit installed and validated

Ride-through behavior reconciled with current NERC PRC-029-1 / PRC-024-4 requirements

California's grid is being rebuilt around inverter-based resources (IBRs). Utility-scale solar, battery energy storage systems (BESS), and modern wind now make up a dominant and growing share of new interconnections in the CAISO footprint. With that shift comes a hard engineering reality: inverters don't behave like the synchronous machines the grid was originally designed around. Their fault-current contribution, control behavior, and ride-through performance are governed by software and power electronics, not physics and inertia.

That is precisely the gap Appendix H of the CAISO Large Generator Interconnection Agreement (LGIA) is written to close. Appendix H sets the technical interconnection requirements specific to Asynchronous Generating Facilities — the formal term for IBRs in the CAISO tariff. If you are interconnecting a solar, storage, or wind project to the CAISO Controlled Grid, the obligations in Appendix H are contractual, enforceable, and tested against real plant performance during grid disturbances.

This post walks through what Appendix H actually requires, how it interacts with the newer NERC ride-through standards, where projects commonly run into trouble, and how Keentel Engineering supports Generator Owners through each stage.

Across utility-scale solar and BESS interconnections, the same compliance pain points recur:

Inverter current limits vs. reactive injection requirements

Hardware constraints on short-duration and maximum current can make full reactive-current injection during deep voltage dips difficult to achieve.

Controller tuning trade-offs

Faster response settings risk controller instability; slower settings delay voltage recovery. Both create compliance exposure.

Plant controller interference

Plant-level controls that delay or override inverter restoration after a disturbance directly violate Appendix H.

Momentary cessation misconfiguration

Inverters configured to cease current outside the narrow permitted high-voltage band fail the standard.

Model fidelity gaps

Generic or unvalidated inverter models that don't match field behavior undermine both interconnection studies and ride-through demonstration.

Data recording shortfalls

The 20 MW recording threshold, GPS time-sync, sampling rates, and retention/turnaround obligations are frequently underbuilt.

Phase-lock-loop and loss-of-synchronism behavior.

Protection settings that trip on phase-angle shifts can conflict with the requirement to hold synchronism.

A reference for Generator Owners, developers, and engineers working through CAISO Large Generator Interconnection Agreement (LGIA) Appendix H requirements for inverter-based resources. Prepared with engineering support from Keentel Engineering.

Owner’s Engineer Services for BESS & Substations

Thu, 11 Jun 2026 14:17:40 GMT

Jun 11 , 2026 | blog

A Coordinated Electric System Interconnection Review—the utility’s deep-dive on technical and cost impacts of your project.

contact@keentelengineering.com

813-389-7871

Schedule a Consultation

Owner’s Engineer for Utility-Scale BESS and HV Substations

1. Why Owner's Engineering Matters Now

What is T&D Co-Simulation?

Confusing Physical Connections with Logical Nodes in IEC 61850

Design review is organized by discipline. A typical storage-plus-substation engagement covers the BESS and power-conversion electrical design; the medium-voltage AC collection system; the high-voltage substation and main power transformer; the generator tie-line; auxiliary power; civil works and drainage; structural and foundation design; the operations-and-maintenance building; and the SCADA and communications architecture.

3.3 Engineering Studies Review

Independent desktop review of the project's engineering studies confirms that analysis supports the drawings and meets the review requirements. Studies commonly reviewed include load flow and voltage regulation; short-circuit and protective-device coordination; arc-flash (AC and DC); grounding; insulation coordination and surge protection; harmonic and transient analysis; reactive-power and temporary-overvoltage studies; cable ampacity and voltage drop; thermal analysis; battery and charger (DC system) sizing; and loss evaluation.

3.4 Construction Monitoring and Commissioning Support (Optional)

During construction, the OE provides as-needed monitoring — on-site presence to witness key activities and milestones, verification of progress against the design, and review of RFIs, technical submittals, and change orders. During commissioning, the OE reviews and witnesses the commissioning program: the commissioning plan, safety and lock-out/tag-out protocols, test procedures and reports, energization procedures, hot-commissioning and operational testing, and SCADA configuration, along with medium- and high-voltage testing reports. Site visits commonly extend through mechanical completion and substantial completion, including review of performance-test results and turnover documentation.

3.5 As-Built Record Review (Optional)

After construction, the OE performs a desktop review of as-built redline drawings for final compliance with the design criteria, typically one round per discipline. This closes the loop between what was designed, what was approved, and what was ultimately built — an important record for operations, future augmentation, and asset transactions.

2. What an Owner's Engineer Is — and Is Not

3. The OE Scope Across the Project Lifecycle

7. How Keentel Approaches Owner's Engineering

Keentel Engineering structures its OE engagements around disciplined, milestone-aligned review and transparent comment management:

Milestone-aligned reviews synchronized to the contractor's submittal calendar, with defined turnaround windows.

Shared comment logs with unique identifiers, severity, discipline, reference, and status — tracked to closure and reported at each milestone.

Independent senior check of every discipline review before issue, with principal-level QA oversight and a clear escalation path for significant findings.

Single accountable team carrying context from design basis through commissioning and as-built closeout.

This approach keeps the owner informed, keeps the contractor accountable, and keeps the project moving — without compromising independence.

An Owner's Engineer is an independent engineering consultant retained by the project owner (or the owner's representative) to provide technical oversight throughout development, design, construction, and commissioning. The OE reviews the EPC contractor's work products for conformance with the project's requirements and reports findings to the owner.

2.1 The OE's Defining Characteristics

Independent. The OE does not design the plant and is not the engineer of record. Its value is precisely its separation from the party producing the work being reviewed.

Multi-disciplinary. A storage-plus-substation project demands electrical, power-systems, protection and controls, civil, structural/geotechnical, and SCADA/communications expertise under one coordinated review.

Owner-aligned. The OE's obligation runs to the owner's contracts, financing requirements, and long-term asset performance — not to the contractor's schedule or margin.

Lifecycle-spanning. The same team carries context from design basis through commissioning and as-built closeout, avoiding the knowledge loss that occurs when reviews are fragmented.

2.2 What the OE Is Not

The OE does not replace the EPC contractor's engineer of record, who retains design responsibility and professional liability for the work. Nor does the OE supplant the Authority Having Jurisdiction (AHJ), the interconnecting utility, or the independent system operator. Rather, the OE confirms — independently and on the owner's behalf — that the contractor's deliverables will satisfy all of those parties before money and schedule are committed to them.

4. Technical Focus Areas in Storage-Plus-Substation Projects

5. The Governing Standards Landscape

8. Conclusion

Utility-scale battery energy storage systems (BESS) coupled to high-voltage (HV) substations and short generator tie-lines have become central to grid reliability, capacity, and renewable integration. These projects combine fast-moving power-electronics technology with conventional high-voltage engineering, compressed schedules, and significant capital at risk. The Owner's Engineer (OE) exists to manage that risk: an independent, multi-disciplinary engineering team that verifies the design and construction delivered by an EPC contractor conform to the owner's contracts, the interconnection requirements, applicable codes and standards, and good engineering practice.

This paper describes the role of the Owner's Engineer across the project lifecycle — from staged design review (typically at the 30%, 60%, 90%, and Issued-for-Construction milestones) through construction monitoring, commissioning support, and as-built record review. It outlines the disciplines an OE covers, the technical risk areas that most warrant independent scrutiny in storage-plus-substation projects, the governing standards landscape, and the commercial and bankability value the function delivers. A detailed Frequently Asked Questions section follows.

Audience: project developers, independent power producers, asset owners, lenders, and EPC partners evaluating how and when to deploy an Owner's Engineer on energy-storage and substation projects.

OE reviews assess conformance with the owner's contracts and a broad set of codes and standards. While the controlling documents vary by jurisdiction and interconnection, the families below are typically in scope:

Energy storage and fire safety: NFPA 855 (stationary energy storage systems), UL 9540 and UL 9540A (system listing and fire-propagation testing), UL 1973 (batteries), and IEC 62933, alongside the requirements of the local fire authority.

Electrical: the National Electrical Code (NFPA 70), the National Electrical Safety Code (IEEE C2), and IEEE standards for grounding (IEEE 80), arc-flash (IEEE 1584), interconnection (IEEE 1547 where applicable), and transformers (IEEE C57 series).

Structural and civil: the International Building Code, ASCE 7 for loads, ACI 318 for concrete, and AISC for steel, together with local permitting and stormwater requirements.

Interconnection and commercial: the generator interconnection agreement, the independent system operator's / utility's technical requirements, and the performance obligations of the power-purchase or offtake agreement.

A frequent early-stage check is confirming that the design references the correct interconnection authority and standards for the project's location — a mismatch here propagates into metering, protection, and telemetry design if not caught at the design-basis stage.

6. The Value Proposition: Risk, Cost, and Bankability

Frequently Asked Questions

Executive Summary

The Owner's Engineer function pays for itself through risk avoidance. Its principal benefits include:

Early error detection. Catching conformance and coordination issues at 30% or 60% avoids change orders, rework, and schedule slippage during construction.

Contract enforcement. Independent verification that EPC deliverables meet the agreed specifications protects the owner's commercial position.

Interconnection assurance. Confirming conformance with grid-operator requirements reduces the risk of energization and commercial-operation delays.

Bankability. Lenders and investors increasingly expect independent engineering oversight; a documented OE review supports financing, diligence, and asset transactions.

Operational readiness. Commissioning support and as-built review leave the owner with a plant that performs as designed and a documentation set that supports long-term operations and augmentation.

Because design-change cost rises steeply with project maturity, the return on independent review is greatest when the OE is engaged early and retained continuously rather than brought in to troubleshoot after problems surface.

Energy-storage projects interconnecting through high-voltage substations sit at the intersection of fast-evolving power electronics and established high-voltage engineering, under schedule and capital pressure. The Owner's Engineer is the independent, multi-disciplinary function that protects the owner's interests across this complexity — verifying that the design and the constructed plant conform to contracts, interconnection requirements, and codes, and doing so early enough that issues are inexpensive to resolve. Engaged from design basis through commissioning, a capable OE reduces risk, supports bankability, and helps deliver an asset that performs as intended for its full operating life.

Energy storage has shifted from pilot-scale to portfolio-scale. Projects routinely pair tens to hundreds of megawatts of power conversion with multi-hour energy capacity, a step-up substation interconnecting at transmission voltage, and a tie-line to the point of interconnection. The engineering touches battery and power-electronics technology that evolves on annual cycles, alongside transformer, switchgear, protection, grounding, and structural work governed by decades-old standards. Few owners maintain in-house teams spanning all of these disciplines.

At the same time, the cost of a design error grows sharply as a project advances. A comment resolved at the 30% design stage may cost hours; the same issue discovered during construction can drive change orders, schedule delays, and interconnection setbacks costing orders of magnitude more. Independent technical review early and continuously is therefore among the highest-leverage risk controls available to an owner.

The Owner's Engineer is the mechanism that delivers this control. Acting solely in the owner's interest — distinct from the EPC contractor's engineer of record — the OE confirms that what is being designed and built is what the owner contracted for, what the grid operator requires, and what the codes demand.

A complete Owner's Engineer engagement typically follows the project from design review through commissioning. The phases below reflect a conventional structure; individual engagements may include or exclude optional phases based on the owner's needs and internal capabilities.

3.1 Scope-of-Work and Code Review

Before detailed review begins, the OE examines the scope-of-work documents that define technical requirements, compliance standards, and quality expectations, together with the selected equipment, auxiliary power scheme, and fire-suppression and protective systems. The objective is to surface gaps or non-conformances early and to recommend corrective action while changes are inexpensive.

3.2 Staged Design Review (30% / 60% / 90% / IFC)

The core of most engagements is milestone-based design review. The OE reviews each submittal, issues consolidated comments through a shared comment log, attends a page-turn comment-resolution meeting, and back-checks the contractor's responses. Unresolved or design-affecting items carry forward into the next milestone. A common convention is one review round per submittal phase plus one response-review round per phase.

Certain subsystems concentrate risk and reward independent scrutiny. The areas below are where OE review most often adds value on energy-storage projects interconnecting through an HV substation.

4.1 Battery and Power-Conversion Systems

Container-based battery systems paired with power-conversion stations and medium-voltage step-up transformers form the heart of the plant. Review attention centers on power and energy ratings and their relationship (C-rate and backup duration), round-trip efficiency and auxiliary (parasitic) loads, thermal management and HVAC sizing, DC architecture and protection, and the plan for capacity augmentation over the asset's life as cells fade. Because supplier drawings often mature later than balance-of-plant design, the OE tracks where design is provisional and confirms that interfaces and ratings remain consistent as vendor data is finalized.

4.2 Medium-Voltage AC Collection

Underground medium-voltage circuits aggregate the power-conversion blocks to the collector bus. Key review items include cable ampacity and derating for installation conditions, conductor and conduit sizing, voltage drop, grounding and shielding, and protection coordination across the collection feeders.

4.3 Main Power Transformer and HV Substation

The step-up substation typically centers on a multi-winding main power transformer with on-load tap changing and staged cooling, feeding a transmission-voltage bus through HV circuit breakers, disconnect switches, and instrument transformers. Review emphasis includes transformer ratings, impedance and vector group, insulation coordination and basic insulation level (BIL), short-circuit duty, bus continuous-current ratings, and revenue/interconnection metering arrangements. Point-to-point review of substation electrical drawings confirms wiring, schematics, and protection schemes batch by batch.

4.4 Protection, Grounding, and Safety

Protection and coordination — relay schemes, current- and voltage-transformer ratios and accuracy, CT saturation, and device coordination — determine how faults are detected and cleared. Grounding design (including step- and touch-potential safety) and lightning/surge protection are reviewed against the relevant standards. Arc-flash analysis informs labeling and worker safety on both AC and DC systems.

4.5 Generator Tie-Line

Where the plant interconnects through a short tie-line, the OE reviews conductor and hardware selection, structural loading, clearances, and the line model. Even a short single-span line carries interconnection and conformance requirements that must be verified.

4.6 SCADA, Controls, and Communications

Plant controls, the power-plant controller (where applicable), SCADA architecture, and the communications/fiber backbone are reviewed for conformance with the owner's technical requirements and the interconnection's telemetry and control obligations. Controls increasingly determine whether a plant can deliver its contracted grid services.

4.7 Civil, Structural, and Geotechnical

Grading and drainage, erosion and sediment control, stormwater management, access roads, foundations, and equipment structures are reviewed against geotechnical findings, hydrology, permits, and structural standards. Foundation calculation packages, concrete mix designs, and shop drawings receive particular attention because they are difficult and costly to correct after placement.

keentelengineering

Inside the PJM Rulebook: Markets, Governance & Interconnection

Inside the PJM Rulebook

1. Why the Rulebook Matters as Much as the Rules

Keentel Perspective

2. What PJM Is and What It Is Not

3. The Market Architecture: Four Mechanisms, One Balance

5. How a PJM Rule Is Actually Made

About Keentel Engineering

Part 2 — Frequently Asked Questions: Large Load Interconnection

Keentel Perspective

4. Governance: Who Actually Decides

Why This Pipeline Matters to Developers

6. Participation Is Open and Underused

7. What This Means Right Now: The Issues Moving Through the Pipeline

8. How Keentel Engineering Fits In

Frequently Asked Questions

9. Conclusion

Disclaimer :

Gigawatt-Scale Loads & Grid Interconnection Engineering Guide

Gigawatt-Scale Loads on the Bulk Power System

About This Publication

1. A New Class of Grid Citizen

2. Four Ways Large Electronic Loads Threaten Operational Security

7. What This Means for Interconnection Engineering

Work With Keentel Engineering

3. Inside the Modern Data Center Power Chain

4. AI Workloads: The Most Dynamic Load Ever Connected

Non-Affiliation Disclaimer

Part 1 Technical Blog: When the Load Becomes the Largest Machine on the Grid

Part 2 — Frequently Asked Questions: Large Load Interconnection

5. Cryptocurrency Mining: The Price-Following Gigawatt

6. Hydrogen Electrolysis: The Electrochemical Large Load

About Keentel Engineering

Disclaimer

ERCOT BESS Interconnection Guide | TNMP & AEP Texas

Interconnecting a Utility-Scale BESS in ERCOT: The Complete Guide to the TNMP and AEP Texas Process

Why the 10 MW Threshold Changes Everything ﻿

Phase 0: Preparation and Application

Case Studies ﻿

Case Study 1 — TNMP Territory: The Threshold Decision That Saved a Year

Case Study 2 — TNMP Territory: Recovering a Stalled Application

Case Study 3 — AEP Texas Territory: Parallel-Tracking Utility and ERCOT Workstreams

Case Study 4 — AEP Texas Territory: Fire Protection as a Design Driver, Not an Afterthought

Phase 1: Screening Study (approximately 90 days)

Phase 2: Full Interconnection Study (roughly 120 to 180 days)

Phase 3: Agreements, Registration, and the QSA Cycle

Phase 4: Commissioning and Energization (roughly 90 to 120 days)

Beyond the Utilities: Permitting, Codes, and Fire Protection

The Bottom Line ﻿

The AEP Texas Track: What to Expect in AEP Territory

The ERCOT GINR Process, Phase by Phase ﻿

Frequently Asked Questions

The TNMP Track: What to Expect in Texas-New Mexico Power Territory ﻿

SCADA in Field Services & Digital Substation Architecture

SCADA in Field Services and Substation SCADA Architecture

1. Why SCADA Is Now a First-Order Design Input

Keentel Perspective

3.1 Remote Terminal Units and Gateways

3.2 Intelligent Electronic Devices

3.3 Merging Units and the Process Bus

3.4 Master Stations and the HMI

8. Securing the Substation SCADA System

6. Protocols: Choosing and Integrating the Right Ones

3. The Building Blocks: RTUs, IEDs, Data Concentrators, and Merging Units

Design Rule of Thumb

5. Reliability Engineering: Fail-Safe, Fault-Tolerant, and High-Availability Design

7. SCADA in Field Services: Where Architecture Meets Reality

9. The Migration Question: Legacy, Hybrid, or Fully Digital?

4. Substation SCADA Architecture: Levels, Buses, and Topologies

Let's Talk

Disclaimer

10. How Keentel Engineering Supports SCADA and Substation Automation

2. SCADA Fundamentals: Data Acquisition to Supervisory Control

Frequently Asked Questions

Field Reality Check

IEEE Standards in Power Engineering | Keentel Engineering

IEEE Standards and How Keentel Engineering Puts Them to Work

Introduction: Standards Are Design Inputs, Not Checkboxes

Disclaimer

Why the 10 MW Threshold Changes Everything

Case Studies

The Bottom Line

The ERCOT GINR Process, Phase by Phase

The TNMP Track: What to Expect in Texas-New Mexico Power Territory

Pulling It Together: One Model, Many Standards

Frequently Asked Questions

Chapter 5 — The Compliance Clock: Phased Implementation