ErgoTechWiki

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
====== Application note: finding intermittent faults from Sampled Values ======

===== The challenge =====

A feeder trips every few weeks, but the cause is unclear.

The protection relay's disturbance record captures the trip and the short interval around it. That is usually enough to confirm that the relay operated as configured, but it may provide little insight into **what caused the event in the first place**.

Even more difficult are disturbances that never cause a trip. Operators may report flickering lights, a drive shutting down, or a contactor chattering, but by the time an engineer investigates, the event is over. The evidence may have existed for only a few milliseconds and then disappeared.

To diagnose these problems effectively, the system needs to do more than record the moment of a trip. It needs to continuously watch the waveform, retain the period before the event, capture relevant disturbances automatically, and make the resulting data easy to analyze.

===== The approach =====

The basic approach is straightforward:

  *  continuously monitor high-speed current and voltage waveforms,
  *  maintain a rolling pre-event buffer,
  *  trigger a capture from a relay trip or another abnormal condition,
  *  continue recording for a short period after the event,
  *  save the result in a standard disturbance-record format, and
  *  automatically analyze each capture to identify the events most worth reviewing.

This allows engineers to see not just **that** something happened, but what the system was doing immediately before, during, and after the event.

The trigger does not need to come only from protection equipment. It can also come from analytics, equipment status, a power-quality threshold, or another condition of interest. This is especially useful for disturbances that affect plant operation without ever causing a protection trip.

===== Why the pre-trigger data matters =====

For intermittent faults, the most valuable evidence is often found in the cycles leading up to the event.

A recorder that starts when the trigger arrives may capture the relay action perfectly while missing the behavior that caused it.

A rolling waveform buffer solves that problem. When a trigger occurs, the system retains both:

  * the waveform leading up to the event, and
  * the system response afterward.

This makes it possible to investigate questions such as:

  * Was the current already becoming unstable?
  * Did one phase begin behaving differently?
  * Was there evidence of arcing or switching before the trip?
  * Did the disturbance begin before the protection system reacted?
  * Did the system recover normally afterward?

===== Using Sampled Values as the waveform source =====

IEC 61850 Sampled Values provide the high-speed current and voltage data needed for this kind of analysis.

Rather than treating the data as a stream of isolated samples, the waveform can be grouped into short time windows or electrical cycles and processed as a time-series record.

That makes the same captured waveform useful for several purposes:

  * disturbance recording,
  * RMS and power-quality monitoring,
  * spectral analysis,
  * automated anomaly detection, and
  * later offline review.

The high-resolution data is retained where it is useful, without requiring every downstream application to process thousands of individual samples per second.

===== Using GOOSE or analytics as the trigger =====

A relay trip delivered over GOOSE is a natural trigger for disturbance recording, but it should not be the only one.

Many of the most useful captures are events that never result in a trip.

Additional triggers can therefore be based on:

  * waveform distortion,
  * RMS excursions,
  * rate-of-change,
  * equipment state,
  * operator input, or
  * application-specific rules.

This makes the recorder useful for both protection events and less obvious operational disturbances.

===== COMTRADE as the common record format =====

Captured waveforms can be stored as IEEE C37.111 COMTRADE records.

This is important because COMTRADE is already widely used in protection and power-system engineering. The same capture can be:

  * opened in existing engineering tools,
  * archived with other disturbance records,
  * shared with a relay or equipment vendor,
  * analyzed by another application, or
  * replayed later for testing.

A standard record format also separates the live data source from the analysis. The same analytics can be developed against historical COMTRADE files and then applied to live Sampled Values when the system is deployed.

===== Reducing the waveform to useful evidence =====

A single event may contain thousands of waveform samples, and a site may accumulate hundreds of events over time.

Opening every capture manually is not practical.

The more useful approach is to first reduce each waveform to a set of meaningful measurements that can be sorted, compared, and screened automatically.

One useful example is **spectral impurity**: the proportion of each electrical cycle's energy that lies outside the expected fundamental-frequency band.

For a 50 Hz system:

<code>
  impurity = 1 - (energy in the 45-55 Hz band / total energy in the cycle)
</code>

A clean waveform remains close to zero even if its amplitude changes.

Arcing, switching transients, converter problems, contact operation, and other distorted events move energy away from the fundamental and produce a larger value.

Two summary values can then describe a large part of each capture:

  * the **median impurity**, representing the normal background condition, and
  * the **worst cycle**, including the affected channel and time.

A low median with a short, large excursion suggests a transient disturbance. A consistently elevated median suggests a sustained waveform-quality problem.

This turns a large waveform into a small amount of evidence that can be searched, ranked, and compared.

===== AI-assisted interpretation =====

Once the waveform has been reduced to meaningful features, AI can help interpret the results.

The important point is that the AI does not need to inspect thousands of raw samples directly.

Instead, analytics first identify the significant measurements, trends, frequencies, and event characteristics. The AI then reasons over that much smaller and more meaningful representation.

This has several benefits:

  * less data must be presented to the model,
  * the important evidence is already identified,
  * results are easier for an engineer to verify,
  * smaller or local models become more practical, and
  * the same diagnostic process can be applied consistently across large numbers of captures.

The AI can describe what the event is consistent with and identify the evidence behind that interpretation. The engineer remains responsible for the final diagnosis.

===== Worked example =====

The screen below shows one disturbance record from an RTE digital fault recorder. The original waveform is sampled at 6400 Hz with a nominal frequency of 50 Hz.

{{ :pasted:microgridweb.png?600 |Waveform anomaly screening: raw waveform, per-cycle spectral purity, and the diagnosis}}

The display contains three views.

* **Raw waveform** — six channels representing three currents and three voltages. A disturbance begins just before 17:27:41, current collapses, and approximately one second later the system recovers. The waveform clearly shows that an event occurred, but does not by itself explain its character.

* **Per-cycle spectral purity** — during normal operation the waveform contains very little energy outside the 50 Hz fundamental. During the event the impurity rises sharply for several cycles, showing that much of the signal energy has moved away from the fundamental.

* **Diagnosis** — the steady-state waveform is clean, followed by a short, highly distorted event. Energy is distributed across a much wider frequency range rather than remaining concentrated at 50 Hz. That is evidence consistent with arcing, contact operation, switching, or another transient disturbance rather than a simple clean fundamental-frequency fault.

The analysis also identifies what should be checked next. RMS and peak current, for example, can help distinguish between remaining possibilities.

That combination of **automated screening and explicit supporting evidence** is more useful than simply assigning a fault label.

===== Important limitations =====

No single analytic identifies every electrical fault, so practical monitoring combines several measurements.

* **Spectral impurity does not detect every fault.** A high-amplitude fault that remains almost entirely at the fundamental frequency may still have very low spectral impurity. Amplitude, RMS, peak, and other tests should therefore be used alongside spectral analysis.

* **A failed measurement channel may appear artificially clean.** A channel stuck at zero has no out-of-band energy. A liveness or RMS check should therefore be used before classifying a low impurity value as healthy operation.

* **This is not a protection system.** The relay remains responsible for protecting the plant. The recorder and analytics provide additional evidence to explain what happened.

* **Accurate timing matters.** If records from multiple devices or applications are to be compared, appropriate time synchronization is required.

===== Developing without a live substation =====

The analysis does not need to begin with live IEC 61850 equipment.

Existing COMTRADE records can be used to develop and validate the disturbance analysis, event screening, visualization, and AI-assisted interpretation.

Those records may come from:

  * protection relays,
  * digital fault recorders,
  * laboratory tests,
  * archived plant events, or
  * public disturbance data sets.

Once the analysis is working as expected, a live Sampled Values source can be connected to the same processing chain.

Waveforms can also be replayed in a laboratory environment, allowing the complete workflow to be tested repeatedly without waiting for the next real intermittent event.

===== Implementing the solution with MIStudio =====

MIStudio provides the building blocks needed to assemble this workflow in a single application.

For live IEC 61850 systems, MIStudio can subscribe to Sampled Values for waveform acquisition and to GOOSE for relay or equipment events.

A rolling time-series buffer retains the pre-trigger waveform. When a trip, anomaly, or other condition occurs, the disturbance recorder combines that history with the post-trigger data and creates a complete event record.

MIStudio's COMTRADE support can then write the capture to disk or read existing records back for analysis and testing.

The analytics layer can extract per-cycle or per-window measurements such as:

  * RMS,
  * minimum and maximum,
  * spectral energy,
  * frequency-band ratios,
  * rate of change,
  * peaks, and
  * other waveform features.

Those analytics can be used directly for rules and anomaly detection, or passed to an AI model for higher-level interpretation.

The resulting application can therefore combine:

  * IEC 61850 connectivity,
  * continuous high-speed waveform acquisition,
  * pre- and post-event disturbance recording,
  * COMTRADE interoperability,
  * waveform analytics,
  * automated event screening,
  * visualization, and
  * AI-assisted diagnosis.

Because MIStudio treats the waveform analysis separately from the live protocol source, the same application can be developed using historical COMTRADE files, validated in the laboratory, and then connected to live IEC 61850 equipment without redesigning the analytics.

That provides a practical path from **historical fault analysis to continuous field monitoring and automated diagnosis**.