Between the instrument transformer and the relay, a digital substation replaces copper with a process bus. One thing is unchanged: the protection function still has to be proven. Currents arrive as sampled values, trips and blocks as GOOSE messages, and the wiring has gone.
In IEC 61850 testing the objective is unchanged and the interface is not. Pick-up, timing, characteristic shape and the trip path to the breaker are still verified against one dated settings revision. What changes is how quantities get in, how the relay is isolated, and what evidence the test leaves. The full sequence is in the IEC 61850 testing guide.
What IEC 61850 testing changes for the engineer
In a conventional bay the test set has a physical connection point. Test links or shorting switches isolate the CT secondary, the relay terminals are accessible, and the injected quantity is what the set produces.
In a digital bay the relay may have no analog inputs at all. A merging unit samples the CT and VT outputs and publishes them as a stream, and the relay subscribes to it. Nothing in the cabinet carries the measured current, so there is no terminal to inject into.
The test therefore has two ends: the protection function inside the relay, and the communication path the relay trusts.
Sampled values, GOOSE and test isolation
Sampled values carry the measurements; GOOSE carries the binary scheme signals: trips, blocks, intertrips, breaker-failure initiation. The engineering file issued with the installation defines which streams the relay subscribes to.
Isolation is where accidents live. On a conventional bay the hazard is an open CT circuit; on a process bus the hazard is a test that publishes. A relay left in normal mode can act on an injected quantity and trip a breaker still in service. The quieter failure is a relay that never entered test mode: it ignores the stream, and the test proves nothing.
Two mechanisms guard the boundary. Test mode isolates the relay so its outputs cannot reach the scheme. The simulation flag marks a stream as simulated, so a relay in test mode accepts it and discards the live one. Check both before injection, and confirm the relay is returned to normal mode afterwards.
Conventional and IEC 61850 test steps
| Step | Conventional substation | IEC 61850 substation |
|---|---|---|
| Quantity applied | Analog current and voltage at relay terminals | Sampled value stream matching the subscription |
| Isolation | Test links opened, trip link removed | Test mode set, simulation flag applied |
| Scheme signals | Binary inputs driven, contacts read | Publish and subscribe pairs checked |
| Time reference | Test set timer starts at injection | Common time reference across merging units |
| Evidence | Test set report and event record | Report, event record, as-configured file, port-mirrored capture |
The evidence column is what most often fails review: a capture proves what crossed the wire, not that the protection works. The record set a reviewer expects is described in building a relay test plan.
Where secondary injection still fits
The process bus changes the interface, not the protection algorithms. A relay works on the same measured quantities whether samples arrive from a merging unit or from its own input board.
For a feeder or transformer bay with analog inputs, the VAE-430 supplies three current channels at 3 × 0–40 A and four voltage channels at 4 × 0–120 V, with 8 binary inputs and 4 binary outputs that drive and time trip, blocking and reclose signals.
Where a scheme needs both ends injected at once, such as transformer or line differential, the VAE-660 provides six current channels at 6 × 0–30 A with 300 VA per phase, or 1 × 0–180 A at 1000 VA in parallel, with six voltage channels and eight binary I/O.
Neither instrument publishes sampled values. A relay that takes measurements only from the process bus has no terminal to inject into, and that side is verified with tooling that speaks it. The injection set proves the protection function and the trip path; the interface is proven separately.
Frequently asked questions
Can an IEC 61850 relay be tested with a conventional test set?
Only through the parts that are still conventional. Analog current and voltage inputs can be injected at those terminals, covering the protection elements, settings and trip logic. A relay that takes measurements only as sampled values must be tested on the process bus.
What is a simulation flag for?
It marks a stream as a test stream. A relay in test mode accepts the simulated stream and ignores the equivalent one from the live merging unit, so quantities can be applied to a relay that is still connected. Without it, the relay would reject the stream and keep acting on live data.
Is a network capture acceptable as test evidence?
As supporting evidence, yes. A port-mirrored capture shows which streams crossed the wire and when a trip message was published, which is valuable when a scheme behaves unexpectedly. It does not prove the protection function, so pair it with the event record and the as-configured file.
Where this leaves the test plan
IEC 61850 testing adds a communication path to the scope and removes a set of injection points. The instrument-transformer checks that precede injection are covered in CT and PT excitation and knee point testing. The instruments are listed under secondary test instruments, and a configuration can be discussed through contact or request a quote.
