PRC-028-1 Disturbance Monitoring: Requirements, Recording & IBR Compliance

Part A — PRC-028-1 at a Glance

When a 230 kV line fault 200 miles away causes a solar plant to drop 150 MW, three engineering questions follow at once:

• What did the plant actually do, and when?
• Which inverters, feeders or controls caused the loss?
• Can the owner prove it — with synchronized, high-resolution data — within days, not months?

For most of the last decade, the honest answer at many inverter-based resource (IBR) sites was "we don't know." Event investigators repeatedly found that plants had no fault recordings, low-resolution SCADA data, inverter logs with unsynchronized clocks, and fault codes that were overwritten before anyone asked for them.

NERC Reliability Standard PRC-028-1, Disturbance Monitoring and Reporting Requirements for Inverter-Based Resources, closes that gap. It requires Generator Owners (GOs) of applicable IBRs to record three kinds of data, synchronize it to UTC, keep it retrievable, provide it quickly on request, and restore failed recording capability on a fixed clock.

The standard on one page

Key dates

Figure 1. Keentel Engineering's illustration of PRC-028-1 data at a 230 kV point of interconnection during a cleared single-line-to-ground fault. Top: fault recording at 64 samples per cycle with the 2-cycle pre-trigger and 2.0 s minimum record marked. Middle: dynamic disturbance recording of real and reactive power at 60 records per second. Bottom: a 20 Hz plant control oscillation sampled at 60 records per second (resolved) and at 30 records per second (misread as 10 Hz).

Five lessons

• PRC-028-1 is an engineering standard, not a paperwork standard. Compliance is proven by recorder settings, channel lists, trigger logic, clock architecture and retrieved files — not by a procedure alone.
• Existing PRC-002 equipment is a starting point, not the answer. PRC-028 adds collector feeder FR, a 2.0 s record, 60 records/s DDR and IBR unit-level SER that most legacy installations do not provide.
• Fifteen days is the real deadline. The 20-day figure is how long data must remain retrievable; the response deadline is 15 calendar days from the request.
• Time synchronization decides whether data is usable. A 1 ms error is 21.6° of phase at 60 Hz. Unsynchronized inverter logs are the most common reason event timelines cannot be reconstructed.
• Design once for the whole Order No. 901 roadmap. The same data feeds ride-through evaluation under PRC-029-1, event analysis under PRC-030-1 and model validation. A minimum-compliance design often fails the analyses that follow.

Engineering review of the original draft

Keentel Engineering reviewed the earlier version of this guide and the common industry statements it contained. The corrections below are carried into this edition.

Part B — PRC-028-1 Disturbance Monitoring for Inverter-Based Resources: An Engineering Guide From Requirement to Recorder Settings

1. Why IBR disturbance monitoring became mandatory

1.1 What event analysis kept finding

Between 2016 and 2023, NERC and the Regional Entities published a series of disturbance reports on large losses of solar and wind output during normally cleared transmission faults. The recurring causes were well known by the end of the series:

• Inverter phase-locked loop (PLL) loss of synchronism during phase jumps.
• Instantaneous AC overvoltage and DC overvoltage trips.
• Momentary cessation, with slow or incomplete recovery.
• Plant controller interactions that held output down after the fault.
• Feeder and collector protection that tripped on transient conditions.
• Frequency measurement errors that tripped inverters on false over- or underfrequency.

The reports also kept repeating one complaint: many plants could not explain what had happened because the data did not exist. SCADA at 2–4 s resolution cannot see a 50 ms event. Inverter logs had clocks drifting by seconds or minutes. Fault codes were overwritten. Several sites had no fault recorder at all because PRC-002 had not required one at their location.

1.2 From FERC Order No. 901 to PRC-028-1

FERC Order No. 901 (October 2023) directed NERC to develop IBR standards in four areas: disturbance monitoring, ride-through performance, analysis and mitigation of performance issues, and data sharing and model validation. NERC filed the first milestone in late 2024. FERC approved PRC-028-1 together with the IBR glossary definitions, PRC-002-5 and PRC-030-1 on February 20, 2025. PRC-028-1 became effective on April 1, 2025, with phased compliance dates running to January 1, 2030.

PRC-002-5 now excludes IBRs, which are covered by PRC-028-1. A transmission bus that meets PRC-002 criteria stays under PRC-002; the IBR plant at the other end of the line falls under PRC-028.

1.3 How the IBR standards depend on each other

A recording system designed only to tick the PRC-028 boxes often turns out to be inadequate for the PRC-030 analysis that follows the first significant event.

2. Applicability: which plants and which owners

2.1 Functional entity

PRC-028-1 applies to Generator Owners. Earlier drafts also listed Transmission Owners; the approved standard does not. Where a Transmission Owner owns the interconnection substation breakers, the GO must still obtain the required data, which usually means a written data-sharing agreement.

2.2 Facilities

The standard covers:



• BES IBRs — generally, IBR plants of more than 75 MVA aggregate connected through a system designed primarily to deliver that capacity to a common point at 100 kV or above.
• Non-BES IBRs — IBRs that have, or contribute to, an aggregate nameplate capacity of at least 20 MVA, connected through a system designed primarily for delivering that capacity to a common point of connection at 60 kV or above. Owners of these resources were to register as GOs and Generator Operators (GOPs) by May 2026 under NERC's revised Rules of Procedure.

2.3 Worked example: applicability screening

Plant F and G show why aggregation matters: neither qualifies alone, but together they exceed 20 MVA at a 69 kV common point. Screening must be repeated whenever capacity, configuration or the point of interconnection changes — for example, when storage is added to an existing solar site.

3. R1 — Sequence of events recording

3.1 Plant-level SER

The GO must have SER data for the position of circuit breakers (and, per the Technical Rationale, other interrupting devices such as circuit switchers) associated with:

• Each main power transformer.
• The collector bus and each collector feeder.
• Shunt static reactive devices (capacitor banks, reactors, harmonic filter banks) and shunt dynamic reactive devices (STATCOMs, SVCs, synchronous condensers).
• The AC-DC and DC-AC converters of a dedicated VSC-HVDC connection, where one exists.

SER is the backbone of the event timeline. In an analysis, the first question is usually "what opened, and in what order?" A breaker status change recorded with a 1 ms time stamp answers it directly.

3.2 IBR unit-level SER

R1 also requires, at the IBR unit level:

• All fault codes.
• All fault alarms.
• High- and low-voltage and frequency ride-through mode status.

These are recorded when triggered by ride-through operation or by tripping of the unit. The Technical Rationale states that entry into momentary cessation should also trigger recording.

For units placed in commercial operation after the effective date, unit-level SER is mandatory. For units already in service, it is required only if the unit is capable of recording it. Lack of capability must be supported by evidence: equipment specifications, a manufacturer's letter, or documentation of the attempt to obtain one. If the manufacturer no longer exists and capability cannot be determined, the Technical Rationale indicates that the data is not required, provided the owner keeps records of its efforts.

3.3 The firmware gray area

A common situation: the inverter hardware can log ride-through status, but only after a firmware update. Is the unit "capable"? The standard does not say. Keentel's recommended position is to document the manufacturer's statement, the cost and risk of the update (including any effect on PRC-029 ride-through performance and on the validated dynamic model), and the decision. If the update is applied, re-run the model validation and PRC-019 coordination screens, because firmware changes frequently alter controller behavior.

3.4 Worked example: SER point count

For a 200 MW solar + BESS plant with one MPT, a 34.5 kV collector bus, eight collector feeders (six PV, two BESS), one STATCOM and one 34.5 kV harmonic filter bank:

Recording both "a" and "b" contacts allows the recorder to detect contact discrepancy (both open or both closed), a useful check on auxiliary-switch failure. At the unit level, the plant (120 MW PV and 80 MW BESS) has 36 PV inverters and 24 BESS power conversion systems of 3.6 MVA each — 60 units, each with its own fault-code stream. Unit-level SER is usually collected by the plant SCADA or the inverter manufacturer's data logger rather than by the plant recorder.

3.5 SER file format

SER data must be provided in the comma-separated value (CSV) format defined in Attachment 1 of the standard. Many recorders and SCADA historians export SER in their own formats; a conversion script that has been tested against Attachment 1 should be in place before the first request arrives.

4. R2 and R3 — Fault recording

4.1 Where fault recording is required

R2 requires FR data sufficient to determine specified electrical quantities at three locations:

Collector feeder FR is the most significant departure from PRC-002 practice. Feeder-level data lets the analyst see whether a loss of output came from all feeders (a plant-wide controller or ride-through issue) or from one feeder (a feeder protection operation or an inverter-specific problem).

4.2 "Measured or determined"

The standard requires quantities to be recorded or determinable. Residual current, for example, does not need its own CT if all three phase currents are recorded:

Three-phase real power can be computed from the sampled waveforms:

and reactive power from the fundamental phasors:

The trap is triggering. R3 requires a residual overcurrent trigger. If residual current is only calculated after the fact, the device cannot trigger on it. Confirm that the recorder or relay has a real-time calculated residual element that can initiate a record.



4.3 R3 performance criteria

The 2.0 s minimum matters because IBR behavior unfolds after the fault. A fault at a 230 kV substation clears in 3–5 cycles, but the inverters' current injection, PLL resynchronization, momentary cessation recovery and plant controller response continue for hundreds of milliseconds to several seconds. A 0.5 s record — common in relays set up only for protection analysis — shows the fault and nothing of the response that PRC-029 and PRC-030 care about.

4.4 Worked example: sample rate, Nyquist and harmonics

At 64 samples per cycle and 60 Hz:

The Nyquist frequency — the highest frequency that can be represented without aliasing — is:

That is the 32nd harmonic of 60 Hz. In practice the recorder's anti-aliasing filter rolls off below the Nyquist frequency, so reliable content extends to roughly the 20th to 25th harmonic. That is adequate for fault analysis, DC offset, CT saturation, transformer inrush and lower-order harmonic resonance. It is not adequate for inverter switching ripple (typically several kilohertz) or for high-frequency control interactions, which need a dedicated power quality or high-speed recorder.

4.5 Worked example: residual overcurrent trigger at a collector feeder

A 34.5 kV collector feeder uses 1200:5 CTs. The collector system is effectively grounded through a zig-zag grounding transformer at the collector bus. Field measurements show a standing residual current of up to 30 A primary from unbalanced cable charging and load.

Setting the trigger at four times the standing residual:

A bolted single-line-to-ground fault on this feeder produces several thousand amperes of residual current, and a high-resistance fault of a few hundred amperes still triggers reliably. The trigger is far below the feeder's ground protection pickup, so the recorder captures events the relay does not trip on — which is the point.

If the collector system is ungrounded or high-resistance grounded, residual current during a ground fault may be only a few amperes. The "if applicable" qualifier in Part 3.2.3.1 was written for this case, but the standard does not define it. Document the grounding design, the reason a residual trigger is not meaningful, and the alternative used — for example a zero-sequence voltage (3V0) trigger from a broken-delta VT.

4.6 Worked example: voltage and frequency triggers

Triggers must be sensitive enough to capture disturbances that matter but not so sensitive that they fill the recorder with nuisance records.

These are starting points only. Keentel tunes triggers against the plant's measured voltage and frequency distribution so that the expected number of records per month is known and storage is sized accordingly.

4.7 Relays or dedicated recorders?

The standard is equipment-neutral. Modern feeder relays often support 64 samples per cycle and synchronized time stamps, and the Technical Rationale allows multiple records for the same trigger point, which helps relays that cannot store 2.0 s in one record. The trade-offs:

Many owners use a hybrid: feeder relays for feeder-level FR and a multifunction DFR at the MPT for FR, DDR and SER.

4.8 Worked example: FR storage and overwrite risk

A binary COMTRADE record stores, for each sample, a 4-byte sample number, a 4-byte time stamp, 2 bytes per analog channel and 2 bytes per 16 digital channels. For an MPT or feeder record with 7 analog channels (three voltages, three phase currents and neutral current) and 16 digital channels:

For a 2.5 s record at 3,840 samples/s:

Ten FR measurement points (one MPT, eight feeders, one STATCOM) during a storm with 40 triggers per day per device generate:

Storage is cheap at the DFR. The problem is relays. A feeder relay with a 30 s high-rate event buffer holds about 12 records of 2.5 s. On a storm day with 40 triggers, the first 28 records are overwritten before anyone retrieves them — and the most important record is often the first one. Automatic retrieval to a station server after every trigger is the only reliable fix.

5. R4 and R5 — Dynamic disturbance recording

5.1 What DDR captures

FR captures seconds of high-rate waveform around a trigger. DDR records continuously at a lower rate and shows the slower evolution of a disturbance: power swings, frequency excursions, voltage recovery, plant controller response and oscillations. R4 requires DDR data for each MPT sufficient to determine:

• One phase-to-neutral or positive-sequence voltage on the high side (phase-to-phase is acceptable per the Technical Rationale).
• The corresponding phase current or positive-sequence current.
• Real and reactive power on a three-phase basis.
• Frequency (one measurement per site is sufficient).

5.2 R5 rates

R5 requires an input sampling rate of at least 960 samples per second (16 samples per cycle) and an output recording rate of at least 60 records per second.

5.3 Worked example: why 60 records per second

The output stream is a sequence of phasors or RMS values. Its own Nyquist frequency is:

At 30 records per second (the PRC-002 minimum), fN is 15 Hz. At 60 records per second, fN is 30 Hz.



IBR plants can exhibit control interactions and sub-synchronous oscillations in the 5–25 Hz range, particularly in weak grids. Consider a 20 Hz oscillation in plant real power. At 60 records per second it is visible directly. At 30 records per second, without adequate filtering, it aliases to:

An analyst would see a 10 Hz oscillation that does not exist, and could tune the wrong controller. With the filtering used in phasor measurement units, the 20 Hz content is attenuated rather than aliased — but then it is simply invisible. Either way, 30 records per second cannot resolve the behavior. This is why existing PMUs configured at 30 frames per second must be reconfigured, and some older devices replaced.

5.4 Worked example: DDR storage

For one MPT, with an 8-byte time stamp and six 4-byte values (V, I, P, Q, f and phase angle):

At 60 records per second for one day:

Twenty days need about 3.3 GB in binary, or about 8 GB as CSV. Keentel recommends a ring buffer of at least 60 days at the station plus automatic archiving of any period that contains a trigger, so that an event on Day 19 is never lost on Day 21.

5.5 Measurement accuracy is not specified

PRC-028-1 sets rates but not measurement accuracy. Frequency and phasor accuracy depend on the estimation algorithm, the window length and the instrument transformers. For model validation, Keentel recommends devices that meet the measurement classes of IEEE/IEC 60255-118-1 (synchrophasor measurement), and documents CT and VT accuracy classes and burdens for each DDR channel.

6. R6 — Time synchronization

6.1 The requirement

All SER, FR and DDR data must be synchronized to Coordinated Universal Time (UTC), with or without a local time offset. Clock accuracy must be within:

• ±1 ms of UTC for plant-level recording devices.
• ±100 ms of UTC for IBR units.

The accuracy applies to the synchronized clock, not to every time stamp, which also includes measurement and processing delays.

6.2 Do not confuse PRC-028 with IEEE 2800

IEEE Std 2800-2022, the interconnection and interoperability standard for transmission-connected IBRs, specifies ±1 µs at plant level and ±100 µs at unit level for new plants that adopt it through interconnection requirements. These figures are 1,000 times tighter than PRC-028-1. A plant built to IEEE 2800 will meet PRC-028 timing; the reverse is not true. Check which applies to each plant before specifying clocks.

6.3 Worked example: what a timing error does to the data

At 60 Hz, a time error Δt produces a phase angle error:

For a fault cleared in 4 cycles (67 ms), a unit trip time-stamped with ±100 ms uncertainty cannot be placed definitively before or after fault clearing. That distinction often decides whether a trip was a ride-through failure (PRC-029) or a protection response. Keentel therefore recommends distributing NTP or PTP to inverters and plant controllers with a target accuracy of a few milliseconds — well inside the ±100 ms limit.

6.4 Time synchronization architecture

Two often-missed details:

• Loss of synchronization must be detected and logged. A GPS antenna failure that goes unnoticed for months makes every record from that period noncompliant and may count as a recording failure under R8.
• Leap seconds and time zones. Record in UTC (with any local offset declared in the file), and confirm that daylight-saving changes do not shift time stamps on inverters configured with local time.

6.5 Worked example: verifying inverter clock accuracy

Keentel's field method is to compare inverter event logs against the DFR for a known plant-level event — for example, a collector feeder breaker opening during a planned outage. Every inverter on the feeder records loss of grid voltage. If the DFR records the breaker opening at 14:02:07.412 UTC and an inverter logs the loss of voltage at 14:02:07.389, the inverter recorded the effect 23 ms before its cause, so its clock runs at least 23 ms ahead of UTC (more, once the inverter's few milliseconds of detection delay are added). Repeating this across all units gives a measured clock error distribution, which is strong R6 evidence.

7. R7 — Data retention and provision

7.1 The requirement

On request from the Transmission Planner, Planning Coordinator, Transmission Operator, Balancing Authority, Reliability Coordinator, Regional Entity or NERC, the GO must provide SER, FR and DDR data as follows:

7.2 Two clocks, not one

Watch for this. A worked example in the PRC-028-1 supporting documents describes data requested on Day 6 as due within 20 calendar days of the request. That example conflates the 20-day retrievability window (Part 7.1) with the 15-day response deadline (Part 7.2). The standard governs. Build internal procedures around 15 days and document any extension granted by the requestor.

7.3 Worked example: the request timeline

A disturbance occurs on March 3. The Reliability Coordinator requests data on March 8.

• Data recorded on March 3 must remain retrievable through March 22 (20 calendar days, inclusive of March 3).
• The response is due by March 23 (15 calendar days after March 8).

Now suppose a second request for the same March 3 data arrives on March 20. The data is still within its retrievability window, and the response is due April 4. But if the GO waits to retrieve it, the ring buffer may overwrite it on March 23. The practical rule:

• On receipt of any request, immediately preserve (copy and lock) all data in the requested window.
• Keep raw data longer than 20 days — Keentel recommends at least 60 days — so that retrieval delays never put data at risk.
• After any significant event, hold the data until analysis under PRC-030-1 is complete.

7.4 What a compliant data package looks like

7.5 Why COMNAME matters

IEEE C37.232 file names encode the start date and time, time zone offset, station, device and other identifiers. An analyst receiving thousands of files from dozens of owners after a wide-area event sorts and aligns them by name. Files named "event001.cfg" slow the analysis and draw attention. Configure naming at the retrieval server, and test the full chain — retrieve, convert, rename, package, transmit — before the first request arrives.

8. R8 — Restoring failed recording capability

8.1 The requirement

Within 90 calendar days of discovering a failure of SER, FR or DDR recording capability, the GO must either restore the capability or submit a Corrective Action Plan to the Regional Entity and implement it. Maintenance or testing outages longer than 90 days count as failures; an outage of the monitored Element (for example, a feeder out of service) does not. Supporting materials indicate that CAPs are submitted through the Periodic Data Submittals module of the ERO's Align platform.

8.2 What counts as a failure

• Recorder or relay hardware failure.
• Loss of communications that prevents storage or retrieval.
• Loss of time synchronization that makes data noncompliant.
• Configuration errors — wrong trigger, wrong channel mapping, wrong sample rate — found after the fact.
• Storage full, or ring buffer overwriting data before 20 days.

8.3 Worked example: the 90-day clock

A GPS antenna failure is discovered on March 10. Counting 90 calendar days:

Capability must be restored, or a CAP submitted, by June 8. Long-lead items — a replacement DFR, an outage to install CTs — rarely fit inside 90 days, so the CAP route should be decided within the first two to three weeks.

8.4 Discovery is the key fact

The clock starts at discovery. An owner without monitoring may "discover" a six-month-old failure during an audit, and the auditor will ask why it was not found sooner. Keentel recommends:

• Daily automated health checks of every recorder: communications, clock status, storage, last record time.
• A weekly test trigger (or review of the most recent natural trigger) on each FR device.
• Alarms to the operations center for clock loss and recorder failure.
• A failure log recording discovery date, cause, restoration date or CAP submission.

9. Compliance dates and phase-in strategy

9.1 Reading the Implementation Plan

The phased dates are summarized in Part A. Three details deserve attention:

• "Within three calendar years." Some read this as April 1, 2028; the RSAW lists January 1, 2029. Until your Regional Entity confirms otherwise, plan to April 1, 2028.
• The 50% basis is undefined. The plan does not say whether 50% is measured by number of facilities, nameplate capacity or another measure. Document your basis.
• The single-IBR provision. A GO with only one BES IBR must reach 100% within three calendar years — not by 2030. This provision is easy to miss because the RSAW compliance-date summary omits it.

9.2 Worked example: meeting the 50% milestone

A GO owns six existing BES IBRs:

Fifty percent by count is three plants. Fifty percent by capacity is 540 MW. Choosing the three smallest plants (P4, P5, P6) gives 330 MW — compliant by count, not by capacity. Choosing P1, P4 and P5 gives three plants and 550 MW — compliant on both bases. Keentel recommends a plan that satisfies both, which removes the interpretation risk at little cost, and prioritizing plants by reliability risk: size, grid strength, event history and inverter vintage.

9.3 New and non-BES resources

A new BES IBR must comply at the later of July 1, 2026 or its commercial operation date. Because the non-BES new-resource date is measured from the standard's effective date rather than from registration, a non-BES IBR that reaches commercial operation shortly after May 15, 2026 must comply at commercial operation. Monitoring must be in the construction scope, not added afterward.

9.4 Extensions

The Implementation Plan allows GOs with IBRs in service before the effective date to request an extension when circumstances beyond their control — supply chain delays, inability to obtain outages — prevent installation. The request must identify the IBRs, give an installation plan and timetable, explain the circumstances, and be submitted no later than three months before the compliance date. For an April 1, 2028 milestone, that means by January 1, 2028.

9.5 The PRC-002 gap

PRC-002-5 removed IBRs immediately, while existing IBRs have until 2028–2030 to comply with PRC-028. Monitoring equipment at existing IBRs may therefore not be strictly required by either standard for a period. Do not neglect or decommission it. It is likely to form part of the PRC-028 solution, and event analysts will expect its data to exist.

10. Designing a PRC-028 monitoring system

10.1 The example plant

10.2 Worked example: channel and device count

Plus 60 IBR units' fault codes, alarms and ride-through status through the SCADA or the manufacturer's data logger.



At 64 samples per cycle, the aggregate FR sampling load across the plant is:

That is spread across nine devices, well within modern hardware — but it should be checked for any single DFR that takes all feeder channels directly.

10.3 Architecture

• One multifunction DFR at the collector substation: MPT FR and DDR, STATCOM FR, all plant SER.
• Feeder relays configured for 64 samples per cycle, 2.5 s records, residual and voltage triggers, with automatic retrieval.
• GPS clock with IRIG-B to the DFR and relays, PTP or NTP to the plant controller, SCADA and inverters.
• Station data server collecting all records, converting to COMTRADE with COMNAME names, keeping a 60-day ring buffer and permanent archive of triggered events.
• Health monitoring feeding alarms to the operations center.
• Cyber security — the recorder network is part of the plant's BES Cyber Systems scope; low-impact controls under CIP-003, including vendor electronic remote access controls, apply.

10.4 Instrument transformers

The data is only as good as the CTs and VTs. Check:

• CT saturation. For the MPT high-side CTs, verify that the CT will not saturate for the maximum asymmetrical fault current and burden; saturated currents in a fault record lead to wrong conclusions about inverter fault current contribution.
• VT location. High-side voltage must represent the MPT high side, not the line side of an open breaker.
• Polarity and phasing. Verify with a primary injection or load check at commissioning; reversed CT polarity produces negative power in the DDR and an apparent "export" during a fault.

10.5 Commissioning test plan

11. Evidence and audit readiness



11.1 Evidence by requirement

11.2 Read the RSAW — but rely on the standard

The RSAW tells you how the auditor will test you. It is not the requirement. In Keentel's review, the PRC-028-1 RSAW contains at least three points that differ from the standard: the 20-day response example, an evidence request for real power at shunt dynamic reactive devices (R2 requires only reactive power there), and an auditor note on time synchronization that cites only the 1 ms figure, without the 100 ms tolerance for IBR units. Where the RSAW and the standard differ, rely on the standard and be ready to cite the requirement part.

11.3 Requirements that may never be triggered

R7 and R8 apply only when a request or a failure occurs. If none occurred during the audit period, the evidence is an attestation supported by the processes that would have caught one: the request intake log, the health-monitoring system and the failure log. An auditor will ask, "How would you know?" The answer must be a system, not a person.

11.4 A 90-day readiness plan

12. Common misconceptions

13. Key takeaways

• PRC-028-1 is effective now; the R8 date for BES IBRs (January 1, 2026) has passed, and the first R1–R7 milestones are approaching.
• The minimums — 64 samples per cycle, 2.0 s records, 60 records/s DDR, ±1 ms plant clocks — are achievable with modern equipment, but rarely without settings changes.
• Collector feeder FR and unit-level SER are the most common gaps.
• Fifteen days is the response deadline; design retrieval to deliver in days.
• Monitoring health is a compliance function: discovery starts the R8 clock.
• Good PRC-028 data is the foundation of PRC-029, PRC-030 and model validation.

14. How Keentel Engineering helps

Keentel Engineering supports Generator Owners, developers and asset managers with NERC compliance built on engineering evidence, for IBR plants interconnected from 69 kV to 500 kV.

Why clients choose Keentel Engineering

• Engineering-first compliance: settings files, drawings and test records — the evidence auditors trust.
• IBR depth: protection, controls, ride-through, EMT and positive-sequence modeling under one roof.
• Practical field experience: commissioning and event analysis at solar, wind and storage plants.
• Clear deliverables: gap registers, plans and narratives that owners can act on.
• Licensed professional engineering: a Florida-registered firm with offices in Tampa, Austin, Sacramento and Baltimore, supporting projects nationwide.

Facing a PRC-028-1 compliance date or a data request? Talk to Keentel Engineering at (813) 389-7871, contact@keentelengineering.com, or book a 15-minute scoping call at calendly.com/keentel-engineering/15min.

Part C — Case Studies

The following are illustrative composite scenarios based on conditions typical of US IBR compliance work. They show how Keentel Engineering approaches PRC-028-1; site details are generalized and figures are rounded.

Case Study 1 — Portfolio gap assessment and phased implementation

Situation. A Generator Owner with seven existing BES solar and solar-plus-storage plants, interconnected at 115 kV to 345 kV across two Regional Entities, needed a PRC-028-1 plan. Each plant had a DFR at the point of interconnection installed under earlier PRC-002 practice. The owner assumed most requirements were already met.

Analysis. Keentel performed a requirement-by-requirement gap assessment at every plant. Findings:

• The DFRs met 64 samples per cycle, but records were set to 1.0 s with 5 cycles of pre-trigger.
• DDR was configured at 30 records per second at four plants.
• No plant had collector feeder FR beyond protection event reports of 15–30 cycles.
• Feeder relays could record at 64 samples per cycle but stored only a few records before overwriting.
• Inverter clocks at three plants were synchronized to the SCADA server, which itself drifted by up to 4 s.
• No plant had a tested data request procedure.

Engineered solution. Keentel separated settings-only fixes from hardware work. Record length, DDR rate and triggers were changed by settings at all seven plants. Feeder relays were reconfigured and automatic retrieval added to a station server. GPS clocks were added at two plants, and NTP was redistributed from the substation clock to the SCADA server and inverters. The 50% milestone was planned on both count and capacity, with the two highest-risk plants (largest, in the weakest part of the grid) first.

Outcome. About 70% of the gaps closed through settings and configuration in the first six months. The hardware program fit the owner's capital cycle, and the measured inverter clock error fell from several seconds to under 10 ms. A data request drill delivered a complete COMTRADE package in four days.

Case Study 2 — Event analysis that the data could not support

Situation. A 250 MW wind and storage plant at 345 kV lost about 90 MW during a nearby line fault that cleared in 4 cycles. The Reliability Coordinator requested data and an explanation. The owner's records consisted of a 0.5 s fault record at the POI, SCADA data at 2 s and inverter logs with unknown clock offsets.

Analysis. Keentel reconstructed the event. The fault record showed fault clearing, but ended before output fell. SCADA showed the loss but not its timing. Turbine and PCS logs showed ride-through mode entries and several trips, but time stamps differed by up to 40 s between turbine groups. Keentel aligned the logs using a common feeder breaker event, which narrowed the trips to a 200 ms window after fault clearing — enough to identify a converter DC overvoltage protection setting on one turbine type, but not enough to rule out a plant controller contribution.

Engineered solution. Keentel redesigned the plant's monitoring to PRC-028-1: 3 s FR at the MPT and feeders, 60 records/s DDR, PTP to the DFR and NTP to turbines and PCS, unit-level SER with ride-through status. The turbine protection setting was corrected through the manufacturer, and the change was screened for PRC-019 coordination and model updates.

Outcome. Eight months later a similar fault occurred. The plant rode through with a brief, recovering dip, and the full event timeline — breaker operations, unit ride-through mode entries and plant P, Q and frequency — was assembled in two days.

Case Study 3 — A newly registering non-BES IBR owner

Situation. An independent owner of three plants — 40 MW solar at 69 kV, 30 MW solar at 69 kV and 25 MW storage at 115 kV — was newly within NERC's non-BES IBR registration scope. The owner had no compliance program and no disturbance monitoring beyond revenue metering and SCADA. A fourth plant, a 60 MW solar-plus-storage project, was due to reach commercial operation shortly after May 15, 2026.

Analysis. Keentel confirmed that all three existing plants met the ≥ 20 MVA, ≥ 60 kV criteria. Their PRC-028 compliance date for R1–R7 was January 1, 2030, but R8 applied from April 1, 2027. The new plant was the urgent issue: as a non-BES IBR entering service after May 15, 2026, it had to comply at commercial operation.

Engineered solution. Keentel added a DFR, a GPS clock and feeder relay recording settings to the new plant's construction scope, with commissioning tests for every PRC-028 requirement. For the existing plants, Keentel prepared a phased design and budget, set up health monitoring ahead of the R8 date, and built a compliance program covering PRC-028 alongside CIP-003, PRC-005, PRC-019, PRC-024/PRC-029, PRC-030, MOD and VAR obligations.



Outcome. The new plant entered service with compliant monitoring and evidence in place. The owner completed registration with a functioning program rather than building one after the fact.

Part D — Technical FAQ: PRC-028-1 Disturbance Monitoring for IBRs

References and Further Reading

Links were current at publication (October 2026).

• NERC, PRC-028-1 — Disturbance Monitoring and Reporting Requirements for Inverter-Based Resources, with Implementation Plan, Technical Rationale and RSAW. North American Electric Reliability Corporation, nerc.com.
• NERC, PRC-002-5 — Disturbance Monitoring and Reporting Requirements. NERC.
• NERC, PRC-029-1 — Frequency and Voltage Ride-through Requirements for Inverter-Based Resources; PRC-030-1 — Unexpected Inverter-Based Resource Event Mitigation. NERC.
• FERC, Order No. 901, Reliability Standards to Address Inverter-Based Resources, October 2023. Federal Energy Regulatory Commission.
• NERC, Glossary of Terms Used in NERC Reliability Standards (IBR and IBR Unit definitions). NERC.
• NERC, Rules of Procedure and Compliance Registry materials for Category 2 non-BES IBR Generator Owners and Generator Operators. NERC.
• NERC and Regional Entity disturbance reports on solar and wind resource loss events, 2016–2023. NERC.
• IEEE Std 2800-2022, Standard for Interconnection and Interoperability of Inverter-Based Resources Interconnecting with Associated Transmission Electric Power Systems. IEEE.
• IEEE Std C37.111, Common Format for Transient Data Exchange (COMTRADE). IEEE.
• IEEE Std C37.232, Recommended Practice for Naming Time Sequence Data Files (COMNAME). IEEE.
• IEEE/IEC 60255-118-1, Synchrophasor for Power Systems — Measurements. IEEE/IEC.
• IEEE Std C37.238, Profile for Use of IEEE 1588 Precision Time Protocol in Power System Applications. IEEE.

Disclaimer

This article is general technical information for educational purposes. It is not legal or compliance advice for any specific entity and does not create a professional relationship. Requirement summaries are paraphrases for orientation; the approved NERC Reliability Standards, Implementation Plans and the guidance of your Compliance Enforcement Authority govern. Dates reflect the public record as understood at publication and should be confirmed before reliance. Worked examples and settings are illustrative and must be verified by project-specific engineering, sealed by a licensed professional engineer where required.

Case studies are illustrative composite scenarios based on typical project conditions, not records of specific client projects. Figures are rounded and indicative.

NERC, FERC, IEEE, IEC and other names are the property of their respective owners. Keentel Engineering is not affiliated with or endorsed by these organizations or any equipment vendor.

© Keentel Engineering. All rights reserved.