IEC 61850 Protection Testing
IEC 61850 replaces the copper signal wiring of a substation with an Ethernet process bus and station bus. A relay receives Sampled Values instead of analog currents coming down a multicore cable, and trip, interlock and blocking signals travel as GOOSE messages between intelligent electronic devices. Protection testing does not disappear in that architecture — it moves outward: the test engineer now has to prove that samples arrive correctly, that messages are subscribed and acted on, and that the total trip path stays inside its timing budget, where the job once consisted of injecting a few amps into a terminal.
What changes on the wire
Four signal families replace the analog and hard-wired signals of a conventional bay. Each one fails in a different way, so each one is tested differently.
| Signal | Standard | Travels as | What has to be proven |
|---|---|---|---|
| Currents and voltages | IEC 61850-9-2 LE | Sampled Values on the process bus | Amplitude and phase accuracy at the relay, sample rate, quality bits, behaviour on loss of samples |
| Trips, interlocks, blocking | IEC 61850-8-1 | GOOSE messages | Subscription and dataset mapping, retransmission behaviour, transfer time, handling of the simulation flag |
| Commands, settings, events | IEC 61850-8-1 | MMS client / server | Point mapping against the SCL file, control model, event reporting and time stamps |
| Time reference | IEC 61850-9-3 | Precision Time Protocol | Sync accuracy, holdover, and whether every device on the bus shares the same time base |
The four things that have to be proven
Most IEC 61850 commissioning problems are not protection-algorithm problems. They are mapping, timing or synchronisation problems, and each of them has a distinct test.
- Signal mapping. The SCL files (ICD, SCD and CID, per IEC 61850-6) are the contract between the merging unit, the relay and the test tool. Verifying that every dataset member, control block and quality attribute agrees with the as-built configuration is the first test, and it is done before any current is injected.
- Sampled Value accuracy. A merging unit that reports 1 % low is indistinguishable from a relay that measures 1 % low until the two are separated. Accuracy is checked against a reference at the relay terminals, at the sample rate actually in use — 80 samples per cycle is the common light-edition profile, which is 4 000 samples per second at 50 Hz.
- GOOSE behaviour. A GOOSE message has to be subscribed, acted on, and repeated at the right interval. Test mode matters as much as the payload: a relay that ignores the simulation flag will trip the real breaker during a test, and one that honours it must still be proven to trip once the flag is cleared.
- Protection function and end-to-end time. The relay logic itself is still proven by injecting a fault — through conventional terminals on a hybrid bay, or through Sampled Values on a fully digital one. The end-to-end measurement then covers the whole path: sample acquisition, protection decision, GOOSE publication and the trip contact.
A test sequence for a digital or hybrid bay
- Document review. Collect the SCL files, the network design (VLANs, multicast addressing, redundancy scheme) and the protection settings, and compare them against the as-built configuration.
- Network verification. Confirm that process-bus and station-bus traffic is present, that multicast subscriptions resolve, and that redundancy (PRP or HSR per IEC 62439-3) switches over without losing a message.
- Time synchronisation. Verify PTP lock at every device before any timing result is recorded; a sync fault makes every later timing measurement meaningless.
- Sampled Value verification. Compare amplitude and phase at the relay against the reference, including quality-bit behaviour when a sample is deliberately lost.
- GOOSE verification. Check dataset content, subscription, retransmission and the simulation flag, in both the test and the live state.
- Function test. Inject the fault through whichever path the bay uses and confirm the protection decision, the phase selection and the correct trip output.
- End-to-end timing and records. Measure the total trip path and file the result against the agreed acceptance criteria, together with the SCL revision that was actually in service.
Acceptance criteria to agree before the test
Digital-bay acceptance is easier to argue when the numbers are fixed before the first injection. Four values are usually enough:
- GOOSE transfer time. IEC 61850-5 defines performance classes for message transfer; the fastest class used for protection tripping is 3 ms. Agree which class the scheme is specified to, and measure against that rather than against a vendor claim.
- Sampled Value accuracy and rate. The permitted error at the relay terminals, the samples-per-cycle profile, and what happens when samples are lost or the merging unit restarts.
- Time-sync accuracy. The maximum deviation permitted between devices, and the alarm behaviour when sync is lost.
- Record format. What the commissioning file must contain — SCL revision, test points, measured times, pass or fail per criterion — so that the record is usable at the next maintenance cycle rather than only at handover.
Where analog secondary injection still fits
Most substations being commissioned today are hybrid: a digital station bus with conventional CT and VT wiring into the relay, or a digital process bus in one bay and copper in the next. In those bays the relay is still proven through its current and voltage terminals, and the IEC 61850 part of the scope is the station-bus traffic — GOOSE between relays, MMS to the gateway, and the time base. Analog injection is not a legacy compromise in that case; it is the correct test for the interface the relay actually has.
What this set does not cover
VA-TEK relay test sets are analog secondary-injection instruments. They test a relay through its current and voltage inputs and can monitor GOOSE traffic, but they do not publish Sampled Values or perform process-bus injection. Full digital testing requires a dedicated IEC 61850 test set, which we do not currently offer, and we will not present analog injection as an equivalent substitute. If your scope includes SV injection, it should be specified separately. The guides below remain useful in that case: they define the architecture and the acceptance criteria the specialist test set will be judged against.
Read next
| Topic | Read | Use it when |
|---|---|---|
| IEC 61850 testing overview | IEC 61850 testing guide | You need the concepts — SV, GOOSE, MMS — and how each changes a test plan |
| Digital substation testing | IEC 61850 substation testing | You are testing a digital or hybrid bay with merging units and process-bus traffic |
| Secondary injection basics | Secondary injection testing guide | You are proving a relay through its conventional current and voltage inputs |
| Protection relay commissioning | Protection relay commissioning checklist | You need the test sequence, records and sign-offs required before energisation |
Not sure which article fits your project? Talk to an application engineer or request a quotation.
