PJM Now Wants a PSCAD Model. Here Is What That Actually Involves

A Step-by-Step Guide to the PJM EMT Model Development Guidelines for Inverter-Based Resources — the Thirty-Five Item Checklist, the Eighteen Test Cases, the PSS/E Benchmark, and Where Submissions Fail

1. Executive Summary

PJM issued Revision 0 of its EMT Model Development Guidelines for Inverter-Based Resources on 5 March 2026. It sets out what an electromagnetic transient model of a solar, storage, wind or hybrid plant must contain, how it must be tested, how it must be benchmarked against the positive-sequence model, and what a complete submission looks like.



The document is short on prose and long on specifics: a thirty-five item model requirements checklist that the model supplier must sign, eighteen model quality test cases with defined operating points and acceptance criteria, a cross-platform benchmark against the PSS/E model, and a deliverables list that includes every case and file used to produce the results so that PJM can rerun them.

Two things about it deserve to be understood before anyone starts work. The first is that PJM states its purpose plainly: to surface deficiencies and conflicts between the PSS/E and EMT models before the data package is submitted and studies begin, so that study delays are minimised. This is a front-loading exercise. The work moves earlier in the schedule; it does not disappear.

The second is that most of the requirements cannot be satisfied by the developer. Real hardware control code, detailed inner control loops, full switching representation, actual protection code, and a plant controller that matches the equipment as configured all come from the inverter manufacturer and the plant controller supplier. If the right model and the right rights to use it were not secured at procurement, no amount of engineering effort later will produce a compliant submission.

This paper works through the scope, the timeline, what the model must contain, the checklist, all eighteen tests, the benchmark requirements, the deliverables, a step-by-step process, and the places where submissions fail.

2. Why PJM Issued This

The reasoning in the introduction is worth restating because it explains the shape of everything that follows.

Inverter-based resources connect asynchronously through power electronics, and most of their control systems depend on a stable terminal voltage magnitude and angle. In a weak system the terminal voltage becomes sensitive to variations in current injection, which can provoke adverse controller reactions and threaten stability. Resources in close electrical proximity interact with each other, compounding it. PJM cites sub-cycle overvoltage and sub-synchronous oscillation as the undesirable dynamics it is seeing, and weak-grid stability as the most significant challenge to higher penetration.

Conventional positive-sequence phasor-domain models were never built to represent that. They are intended for the frequency range typical of power system stability work, and PJM notes that range as roughly a tenth of a hertz to ten hertz. Switching transients, torsional interaction, harmonics and other electromagnetic phenomena sit outside it. That is not a criticism of those models; it is a statement of what they were designed to do.

So both are now required. The positive-sequence model in PSS/E under PJM’s existing dynamic model guidelines, and the electromagnetic transient model in PSCAD under this one. PJM points to the interconnection performance standard for inverter-based resources, the distributed resource interconnection standard, the NERC reliability guideline on EMT modelling, the NERC Level 3 Alert issued in May 2025, and the model verification and validation requirements that oblige transmission planners and planning coordinators to develop dynamic model requirements for both model types.

In the planning horizon, PJM states that EMT studies are performed for three scenarios drawn from the NERC guideline: low system strength, poor stability performance in positive-sequence studies, and system topology or conditions carrying stability risk.

3. Who and What It Applies To

The scope is broad and worth reading carefully if you assumed your project was outside it.



• All inverter-based resources: solar photovoltaic generation, battery energy storage systems, Type III and Type IV wind plants, and co-located and hybrid plants such as photovoltaic plus storage or wind plus storage.
• Models must be in a format compatible with the PSCAD and EMTDC simulation environment — project, case and library file types.
• The plant is modelled up to the point of interconnection.
• Where a single plant is split into multiple generating resources requiring separate submissions, PJM states it is acceptable and desirable to submit the model for the entire plant with each individual submission.

One point deserves emphasis for hybrid developers. Adding storage to an existing wind or solar facility is named explicitly as a qualifying change that triggers a new submission — a full set of revised deliverables plus a report explaining what changed. A hybrid retrofit is therefore not a variation on an existing model package; it is a new one.

4. When a Model Is Due

The screening distinction matters commercially. In the transition cycle, only projects in low system strength areas are notified. From Cycle 1 onward, every bulk-power-system-connected project submits regardless of where it connects. A developer whose previous project was not screened in should not assume the next one escapes.



One more item is easy to miss. PJM states that any difference between the model used for study and the actual installed settings in the field should be treated as a qualified change under the applicable facility change standard and reviewed before commercial operation, to determine whether the existing studies remain valid. Commissioning settings drift is therefore not a housekeeping matter — it is a study validity question.

5. What the Model Must Represent

For most bulk system reliability studies PJM expects an aggregate or partially equivalent representation, with multiple inverters represented as one larger unit and the collector system as an equivalent line or cable.



Where a plant uses more than one inverter type or design, a separate aggregate representation is required for each. A plant with two different inverter models needs two distinct aggregated units.

The model must include:

• The aggregated generation or energy storage model.
• The aggregated generator transformer.
• An equivalent collector branch.
• Main step-up transformers.
• The generator tie line.
• Supplemental devices where applicable — static compensators, synchronous condensers, capacitor and reactive banks, harmonic filter banks.
• Station service load.

The external system is represented by an AC source with a Thevenin impedance. PJM determines, based on project size and location, whether a study of interactions with nearby equipment or an evaluation of performance under low short-circuit conditions is required as part of the interconnection study.

6. The Thirty-Five Item Checklist

The checklist is the part of this guideline that most directly determines whether a submission is acceptable, and it must be completed and signed by the supplier of the model, not the developer. Every item must be reviewed, and where the model does not meet a requirement, an explanation is required in the comments column.

It divides into four groups. The following are the requirements with the most consequence for procurement and for study quality.

6.1 Model and Project Documentation

• An identification mechanism — control revision codes, settings files, or a combination — that allows the specific settings and equipment configuration used in a study to be identified and checked at commissioning. This is what makes the settings-drift question in Section 4 answerable.
• The vendor name and specific model version clearly observable in the case file, with documentation and file names that do not conflict with it, and setup and running instructions.
• A sample implementation test case configured to the site-specific equipment configuration up to the point of interconnection, using a single-machine infinite bus representation at an appropriate short-circuit ratio.

6.2 Model Accuracy

• Full detailed inner control loops of the power electronics, as implemented in the installed equipment. PJM states explicitly that the approximate representation used in transient stability modelling is not sufficient, and recommends real-code models embedding the actual hardware code. Models assembled from standard library blocks or manually translated from block diagrams require validation against actual hardware performance, and against field measurement at the as-built stage.
• A full switching-device representation, or a voltage source interface that mimics switching such as a firing-pulse-based model. A three-phase sinusoidal source representation is not acceptable.
• All pertinent control features — external voltage controllers, plant controllers, customised phase-locked loop systems, ride-through controllers, sub-synchronous control interaction damping. Actual hardware code is required for control and protection features.
• Plant control represented in sufficient detail for short-term performance: measurement methods, communication time delays, transitions into and out of ride-through modes, and settable parameters. Generic plant controller representations are not sufficient unless the final controller is designed to exactly match the generic model — a sentence worth reading twice before accepting a vendor package.
• Sub-synchronous oscillation mitigation and protection, with the ability to enable and disable it.
• Transformer magnetising curves, dynamic reactive devices, machine slip for doubly-fed wind, and pertinent electrical and mechanical features including filters and specialised transformers.
• All installed protection systems in detail for balanced and unbalanced faults — overvoltage and undervoltage on individual phase and RMS, frequency, DC bus voltage, converter overcurrent and other inverter-specific protections. Actual hardware code is required.
• Behaviour accurately reflected across the full real and reactive power output range.
• For as-built submissions, settings matched to the site equipment. Default parameters are not acceptable unless they match what is installed.

6.3 Model Usability

• Control and hardware options accessible to the user, with diagnostic flags that clearly identify why a model trips during simulation.
• Operation across solution time steps from five to twenty microseconds, with ten microseconds recommended, and no dependence on a specific time step.
• Self-initialisation to ordered output without intervention, tolerating terminal condition deviations while other devices initialise.
• Acceptance of external reference values for real power or frequency and reactive power or voltage, both at initialisation and dynamically mid-simulation.
• Protection functions that can be disabled — PJM notes this gives study engineers valuable diagnostic information when equipment trips inadvertently.
• Scalable active power capacity, and dispatch capability below nameplate. These are distinct requirements serving different purposes.

6.4 Model Efficiency

• No dependence on a specific compiler, development environment or PSCAD version, with recommended minimum versions stated.
• Initialisation to ordered conditions as quickly as possible, with under five seconds given as the example.
• Support for multiple instances of the same definition in one case, and for the timed snapshot and multiple run features.
• No reliance on global variables, no use of multiple layers including disabled ones, and replication permitted through copy and copy-transfer.
• Contact information for modelling support engineers.
  

7. The Eighteen Test Cases

Unless a test specifies otherwise, the model is set up with a Thevenin equivalent at a short-circuit ratio of three and an impedance ratio of five. Storage projects run every test at full charge and full discharge. Hybrid plants run all five operating modes — generation alone, storage charging alone, storage discharging alone, generation with storage charging, and generation with storage discharging — with any inapplicable mode noted and justified.

8. The Acceptance Criteria That Fail Models

Three criteria appear repeatedly across the test set and account for most failures.

8.1 Damping Ratio of 0.3 or Greater

PJM sets an acceptable damping ratio of three tenths, referenced to the interconnection performance standard for inverter-based resources. It applies to the fault tests, the overvoltage test, the phase angle tests and the rate-of-change test. This is a numerical criterion, not a visual judgment, and it means the report must actually compute damping from the response rather than asserting that the plot looks stable.

8.2 Momentary Cessation

The treatment is nuanced and worth getting right. For the three fault tests, if the plant enters momentary cessation it must resume current injection in no fewer than five cycles following voltage recovery, and if momentary cessation cannot be eliminated because of equipment limitations, PJM must be notified when the report is submitted. For the ride-through profile tests, the overvoltage test, the phase angle tests and the rate-of-change test, the criterion is that momentary cessation is not observed — with the same notification route where equipment limitations prevent it.

The practical consequence is that momentary cessation behaviour is an equipment characteristic that has to be known before the tests are run. Discovering it during testing, late in the interconnection process, leaves a developer with a disclosure to make and limited options.

8.3 The System Strength Test

Test 13 is the one that most directly reflects why the guideline exists. The model is driven through progressively weaker system conditions — short-circuit ratio five, then three, then one and a half, then one point two — with each transition made under a three-cycle bolted fault. Acceptance requires a positively damped response with no unit tripping at a ratio of three and above.



Note what that acceptance criterion does and does not say. It requires stable behaviour down to a ratio of three. It does not require stability at one and a half or one point two — but the model is run there anyway, so PJM sees what happens. A plant that becomes unstable just below the acceptance threshold is visible in the results, and that visibility is the point.

9. Benchmarking Against PSS/E

The benchmark requirement is the mechanism by which PJM achieves its stated aim of catching conflicts between the two models before studies begin.

For as-designed submissions, the PSCAD and PSS/E signals must be overlaid and properly aligned for each signal plot of each required test. The stated objective is that the two models agree, so that the PSCAD model can be used confidently to address complex issues arising from the positive-sequence study. The required benchmark set is the six functional tests, the three-phase fault, the low-voltage ride-through profile and the high-voltage ride-through profile.

For as-built submissions the requirement inverts and expands. Model quality tests are benchmarked against the latest PSS/E user-defined model and library model results, and rather than listing what must be benchmarked, PJM lists what is excluded: the single line-to-ground fault, the line-to-line fault, and the two phase angle tests. Everything else is benchmarked — a substantially larger set than at as-designed stage.

The exclusions are logical. Those four cases involve unbalanced conditions or angle steps that a positive-sequence model cannot represent meaningfully, so a comparison would not be informative. It is a useful reminder of where the two model classes genuinely diverge.

10. As-Designed and As-Built

The two stages ask different questions and the difference shapes the whole programme.

As-designed covers the cycle process and necessary studies — the model of the plant as it is intended to be built. Validation of a model assembled from library blocks is against actual hardware performance.

As-built is the model of the plant as it exists. The requirements tighten in three ways. Settings must match the site equipment, with default parameters explicitly unacceptable unless they match what is installed.

Validation of a non-real-code model is against actual field measurement data rather than hardware performance. And a Parameter Verification Report is required, comparing as-built site parameters — both plant controller and inverter — against the model parameters.

Beyond that sits the tuned model obligation: no later than one year after commercial operation, a tuned set of models aligned with the field tests specified in the model verification and validation standards, reflecting the actual dynamic response of the resource.

The through-line is that the model is expected to converge on the plant over time, and each stage narrows the permitted gap between them.

11. Deliverables and Submission Mechanics

A complete submission consists of four things.

1. The PSCAD model. At as-built stage and beyond, the PSS/E user-defined model and library model are also submitted as part of the benchmark if not previously provided.
2. The completed and signed PSCAD Model Requirements Checklist.
3. The Model Quality Test and PSCAD-PSS/E Model Benchmark Reports.
4. All cases and files used for every model quality test and benchmark report, so that PJM can rerun any of them for verification.



Files are placed in a folder, zipped, and uploaded through PJM’s data submission platform under the stability model or dynamic file location appropriate to the queue process. The upload limit is two hundred and fifty megabytes; larger packages go through secure transfer arranged with the project manager.

That size limit is worth planning around. A full test programme across five hybrid operating modes, at charge and discharge, with all cases and output files retained, generates a large package. Deciding late how to deliver it is an avoidable delay.

On plotting, the guideline is specific: RMS voltage in per unit, active power in megawatts, reactive power in megavars, and frequency in hertz; measured at both the point of connection or inverter terminals and the point of interconnection; and presented at two zoom levels — the full simulation period, and a focused view of the one to two seconds around the disturbance. Reports that omit the transient zoom or a measurement point create avoidable review cycles.

12. The Step-by-Step Process

The following is how the work sequences on a real project, from procurement through to the post-commercial-operation obligation.

STEP 1  Establish whether and when a model is required

Determine which submission pathway applies transition cycle with screening, Cycle 1 or beyond where every bulk-power-system-connected project submits, a necessary or surplus study, or an as-built verification. That determines the deadline and the deliverable set. Do this before anything else, because it fixes the schedule everything else has to meet.

STEP 2  Secure the model at procurement

Write the model deliverables into the equipment purchase specification: a PSCAD model meeting the thirty-five item checklist, real-code content where available, version identification, documentation and running instructions, contact details for modelling support, and the right to use and share the model with PJM. Tie delivery dates to project milestones. This is the single highest-leverage step in the whole process and it happens before any engineering.

STEP 3  Assemble the plant model

Build the aggregate representation up to the point of interconnection: aggregated generation or storage, aggregated generator transformer, equivalent collector branch, main step-up transformers, generator tie line, supplemental reactive devices, and station service load. Where the plant uses more than one inverter type, build a separate aggregate for each. Represent the external system as an AC source behind a Thevenin impedance.

STEP 4  Complete the checklist honestly

Work through all thirty-five items with the model supplier and record compliance. Where an item is not met, write the explanation in the comments column rather than leaving it blank. A checklist with candid gaps and explanations is a better submission than one that claims full compliance and fails a test and the explanations are what allow PJM to assess whether the gap matters for the studies it intends to run.

STEP 5  Confirm the PSS/E model is the current one

The benchmark compares the PSCAD model against the latest submitted PSS/E model. Establish which version that is and whether it reflects the current project configuration before running anything. Benchmarking against a superseded positive-sequence model wastes the whole test programme.

STEP 6  Set up the test environment

Configure the Thevenin source at the default short-circuit ratio of three and impedance ratio of five, with the tests that specify otherwise set up separately. Build the operating point matrix: for storage, full charge and full discharge; for hybrids, all five operating modes. Establish the plotting configuration four quantities, two measurement points, two zoom levels once, so every test produces consistent output.

STEP 7  Run the six functional tests

Flat start, active power reference step down, frequency step up, frequency step down with headroom, and voltage step up and down. These are quick and they surface fundamental problems early: a model that will not initialise cleanly or hold steady state has a defect that must be resolved before any ride-through work is worth doing.

STEP 8  Run the twelve ride-through and performance tests

The three fault cases, the low and high voltage ride-through profiles and the overvoltage step, the system strength sequence, the two phase angle changes, the rate-of-change sequence, and the high and low frequency profiles. Compute damping ratios rather than judging plots by eye, and record momentary cessation behaviour explicitly wherever it occurs.

STEP 9  Benchmark against PSS/E

Overlay and align the PSCAD and PSS/E signals for the required benchmark set — the six functional tests, the three-phase fault, and the two ride-through profiles at as-designed stage, or the substantially larger as-built set. Where the two disagree, resolve the discrepancy rather than presenting it, because an unexplained divergence is the finding PJM is looking for.

STEP 10  Assemble and submit the package

Model, signed checklist, quality test and benchmark reports, and every case and file used to produce the results. Check the package against the size limit and arrange secure transfer in advance if needed. Upload to the correct location for the queue process.

STEP 11  Manage change through the project

Any qualifying change — device updates, adding storage to a wind or solar plant, operating mode conversion — triggers a full resubmission with a report explaining what changed. Treat model configuration as a controlled item so that a settings change during construction is captured rather than discovered.

STEP 12  Close out at as-built and beyond

Produce the as-built model with settings matched to installed equipment, the expanded benchmark set, and the Parameter Verification Report comparing site parameters against model parameters. Then plan for the tuned model obligation within one year of commercial operation, aligned to the field tests the model verification standards specify.

13. Where Submissions Fail

14. What This Connects To

This guideline does not sit alone, and a project team that treats it as an isolated deliverable will duplicate work.

• The positive-sequence model requirement under PJM’s existing dynamic model guidelines. The two are a pair, and the benchmark forces them to agree.
• The model verification and validation obligations in the applicable reliability standards, which drive the tuned model within one year of commercial operation and which require transmission planners and planning coordinators to establish requirements for both model types.
• The interconnection performance standard for inverter-based resources, which is the source of the damping ratio criterion and much of the underlying performance expectation.
• The reliability standard governing frequency and voltage ride-through for inverter-based resources, which takes effect on 1 October 2026 and which the ride-through test profiles relate directly to.
• The disturbance monitoring obligations that supply the recorded data by which real performance is later demonstrated.
• The facility change requirements, under which a difference between the studied model and the installed settings is a qualified change.

The efficient way to run this is as one modelling and compliance workstream with several outputs, rather than as a series of unrelated submissions to different parties.

15. Keentel PJM Modeling Services

Keentel Engineering supports developers, owners and their EPC partners through this process, from procurement specification to the post-commercial-operation obligation.

15.1 Model Development and Assembly

• Model deliverable and usage-rights requirements written into equipment purchase specifications, with delivery dates tied to project milestones.
• Assembly of the aggregate plant representation to the point of interconnection — generation or storage, transformers, equivalent collector, tie line, supplemental reactive devices and station service — including separate aggregates where multiple inverter types are present.
• Review of vendor-supplied models against the thirty-five item checklist, with gaps identified and explanations prepared before submission rather than discovered in review.
• Positive-sequence model development and reconciliation, so that the PSS/E and PSCAD models describe the same plant before benchmarking begins.

15.2 Testing, Benchmarking, and Reporting

• Execution of the full eighteen-case model quality test programme, across the operating point matrix required for storage and hybrid plants.
• Damping ratio computation, momentary cessation characterisation, and diagnosis of trips and instabilities against the model’s own diagnostic flags.
• PSCAD to PSS/E benchmarking with aligned overlays, and resolution of divergence between the two models.
• Model Quality Test and Benchmark Report preparation to the plotting specification, and package assembly against the deliverables list including all cases and files.

15.3 Studies and Interconnection Support

• Weak-grid and control interaction studies, sub-synchronous oscillation screening, and determination of the minimum system strength at which the plant is stable.
• Ride-through capability assessment at the point of interconnection, supporting both the interconnection obligation and the applicable reliability standard.
• Short-circuit, protective coordination, arc-flash, load flow, reactive capability, harmonic and stability studies, with inverter-based resources represented as current-limited sources.
• Interconnection application support and study-phase technical packages, and coordination with PJM and the transmission owner.

15.4 As-Built, Compliance, and Change Management

• As-built model development with settings matched to installed equipment, the expanded benchmark set, and Parameter Verification Report preparation.
• Model tuning against field test data for the post-commercial-operation obligation, and model verification and validation support under the applicable standards.
• Model and settings configuration control, so that changes during construction and commissioning are captured and the resubmission obligation is met deliberately.
• Owner’s engineer services, review of EPC and vendor model packages, and QA/QC of third-party study and model submissions.

Keentel Engineering holds a Florida Certificate of Authorization and maintains offices in Tampa, Austin, Sacramento, and Baltimore, supporting projects across the interconnections.

17. Frequently Asked Questions

16. References and Further Reading

The primary source for this paper is the PJM guideline itself. PJM documents are revised, and this paper describes Revision 0; check for a later revision before relying on any requirement here.

PJM Documents

• PJM EMT Model Development Guidelines for Inverter-Based Resources, Revision 0, System Planning Division, 5 March 2026, together with the standalone PSCAD Model Requirements Checklist — PJM Interconnection
  
• PJM Dynamic Model Development Guidelines, covering the positive-sequence model requirements in PSS/E format — PJM Interconnection
  
• PJM interconnection process materials, including the cycle process, decision points and the data submission platform — PJM Interconnection
  

NERC Guidance and Standards

• NERC Reliability Guideline, Electromagnetic Transient Modeling for BPS-Connected Inverter-Based Resources — Recommended Model Requirements and Verification Practices, March 2023  —  North American Electric Reliability Corporation
  
• NERC Level 3 Alert on inverter-based resource performance and modelling, issued May 2025  —  No

NERC Reliability Standards — including the MOD series covering model data, verification and validation, the PRC series covering protection settings, ride-through, disturbance monitoring and post-event analysis, and the FAC series covering facility changes  —  North American Electric Reliability Corporation

Industry Standards and Guidance

• IEEE Std 2800-2022, Standard for Interconnection and Interoperability of Inverter-Based Resources Interconnecting with Associated Transmission Electric Power Systems — the source of the damping ratio criterion referenced in the test acceptance criteria  —  IEEE Standards Association
  
• IEEE Std 1547-2018, Standard for Interconnection and Interoperability of Distributed Energy Resources with Associated Electric Power Systems Interfaces  —  IEEE Standards Association
  
• WECC Modeling and Validation Subcommittee, Inverter-Based Resources Power Plant Modeling and Validation Guideline  —  Western Electricity Coordinating Council
  

Notice and Disclaimer

This document is original technical content prepared by Keentel Engineering LLC for general professional information. It is not legal, regulatory, or project-specific engineering advice, and it does not constitute a registration determination, an applicability determination, a compliance determination, or a design for any facility.

Facts about events, dates, standards, thresholds and programme status are taken from the three NERC publications identified in the References section, as published. This programme is moving quickly under a regulatory deadline: registry criteria, standard content, comment periods and effective dates are subject to change, and several matters described here were expressly stated by NERC as not yet final. Verify the current position with NERC, the applicable Regional Entity and the interconnecting utility before making any decision.

Analysis, inference, engineering commentary and the recommended actions in this document are Keentel Engineering’s own assessment and should not be attributed to NERC, any Regional Entity, any regulator, or any party named or referenced.

Descriptions of facility protection and transfer scheme behaviour are general engineering discussion; any change to a protection or control scheme must be made by qualified persons with the equipment manufacturer and against the actual tolerance of the equipment being protected.

Keentel Engineering LLC is an independent engineering consultancy. Reference to any regulator, reliability organisation, regional entity, standard, industry organisation, utility, or facility in this document does not imply affiliation with, endorsement by, or sponsorship from any such organisation.