COMTRADE Playback and Fault Forensics
A fault record is the closest thing to a second chance at an event. Replaying it into a protection relay shows whether the relay would operate again with the present settings, and whether the recorded operation matches what the settings predict. The test set is the playback source; the quality of the conclusion depends on the record, the channel mapping and an honest account of the repeatability limits.
What playback is for
| Use | Question answered | Output |
|---|---|---|
| Post-event investigation | Did the relay see what the recorder saw, and did it operate as designed? | Event reconstruction and a written finding |
| Settings validation | Would the present settings produce the recorded operation? | Settings review with a repeatable test record |
| Scheme validation | Does a new relay or logic behave like the existing scheme on the same fault? | Comparison record for the change request |
| Training | Can an engineer see a real fault replayed into a real relay? | Classroom or laboratory exercise |
| Warranty and dispute | Was the relay or the scheme responsible for the misoperation? | Technical report for the parties involved |
The playback workflow
- Obtain the original record. Use the native COMTRADE file from the relay or the fault recorder, not a screenshot or a converted spreadsheet. Keep the original file unchanged.
- Read the channel list and the time base. Identify voltages, currents, digital channels and the sampling rate. Mark any channel that is missing, saturated or derived.
- Map channels to the test set. The test set has a fixed number of voltage, current and binary channels. A record with more channels than the test set provides must be reduced deliberately, and the reduction recorded.
- Check scaling and polarity. CT and VT ratios, star/delta compensation and polarity conventions have to match the relay settings. A reversed current channel turns a correct replay into a wrong conclusion.
- Replay and capture. Start the playback with enough pre-fault time for the relay to settle, record the relay event and the test set binary inputs on the same time base.
- Compare. Compare the relay event with the original record: pickup, trip, timing and targets. Differences are findings, not noise.
- Document. Store the record, the mapping, the settings file and the comparison in one package.
What the test set can and cannot reproduce
| Item | Capability | Limit |
|---|---|---|
| Analog voltage and current channels | Replay through the same terminals the relay uses | Limited channel count; the VAE-430 has 3 current and 4 voltage channels, the VAE-660 has 6 and 6 |
| Digital channels | Replay breaker and protection status through binary outputs | 4 outputs on the VAE-430, 8 on the VAE-660 |
| Sample rate and bandwidth | Playback within the source’s frequency range and update rate | High-sample-rate records need decimation; the result is no longer the original waveform in every detail |
| Time synchronisation | Internal time base for the test | No GPS, PTP or IRIG-B; the replay is not a time-synchronised reproduction of the station event |
| Process bus | Analog injection only | No IEC 61850 Sampled Values or GOOSE replay |
| Repeatability | Repeatable within the test set’s timing and amplitude accuracy | Small differences at the start of the record, in pre-fault data and in frequency can change the relay’s decision near a boundary |
Why two replays can disagree
A replay is not a recording. The original record was sampled by the recorder at its own rate, with its own anti-aliasing and its own time base. The test set resamples it, generates the waveform and starts it at a chosen instant. Near a threshold, a fraction of a cycle, a fraction of a hertz or a small difference in pre-fault history can move the relay from trip to no trip. When a replay is used as evidence, the report should state the record, the sampling treatment, the channel mapping and the measured repeatability. An unqualified claim that “the relay would have operated” is not supported by a single playback.
What the test set does not do
- It does not record the fault. The record comes from the relay or the fault recorder.
- It does not certify the recorder’s accuracy or the CT/PT chain.
- It does not reproduce a digital substation process bus.
- It does not establish legal chain of custody by itself. That is a procedure around the files, not a function of the instrument.
Frequently asked questions
Can the VAE-430 replay COMTRADE files?
Yes, within its channel count and frequency range. Records with more channels than the test set provides must be reduced, and the reduction has to be documented.
How many channels can be replayed?
The VAE-430 provides three current, four voltage and four binary outputs. The VAE-660 provides six current, six voltage and eight binary outputs. Digital inputs are available on both for reading the relay response.
Can playback prove that a relay misoperated?
It can show how the relay responds to the recorded waveform under the present settings. Whether that response was correct is a settings and scheme question. The report should separate the replay result from the engineering judgement.
Does it support high-sample-rate records?
Records are resampled to what the source can generate. Very high sample rates and very wide bandwidths are reduced; the replay remains useful for protection-frequency behavior, not for reproducing the original waveform bit for bit.
Can it replay IEC 61850 Sampled Values?
No. The VAE series is an analog secondary-injection source. Process-bus replay needs a dedicated digital test set.