Testing and Commissioning IEC 61850 Systems: A Step-by-Step Guide

During a recent commissioning job at a 132 kV grid substation in the north of England, our team watched a protection scheme refuse to trip on a simulated busbar fault. The wiring was correct, the relays were healthy, and the FAT documentation said everything had passed. The problem turned out to be a single mis-mapped GOOSE dataset — one binary output published under the wrong dataset member, invisible until we ran live publish/subscribe tests on site. That afternoon cost the project two days of programme time and a lot of awkward conversations. It is also the reason we now treat IEC 61850 commissioning as a discipline in its own right, not an extension of traditional hardwired testing.

Why IEC 61850 Testing and Commissioning Is Different

Traditional substation schemes are physically verifiable. You can bell out a copper core, inject current into a test block, and watch a trip contact close. IEC 61850 replaces much of that copper with Ethernet frames, and the signals that matter — GOOSE messages for protection and interlocking, MMS for SCADA and engineering access — exist only as configured data objects inside intelligent electronic devices (IEDs). If the configuration is wrong, the wire is effectively cut, and nothing looks broken.

This shift is now mainstream in the UK. National Grid Electricity Transmission has rolled out digital substation architecture across multiple transmission sites, and ScottishPower Energy Networks has pursued IEC 61850-based protection and control on its distribution network as part of wider smart grid investment. The direction of travel is clear, and commissioning teams need to keep pace.

The shift from copper to Ethernet-based communication

In a digital substation, current and voltage transformers feed merging units that publish Sampled Values over the process bus, while protection IEDs exchange trip and interlock signals over GOOSE on the station bus. A single fibre or copper Ethernet link can now carry hundreds of signals that previously required individual cores. That is a huge engineering gain — but it also means a VLAN misconfiguration or a duplicate GOOSE control block can take out an entire protection zone rather than one circuit.

Key challenges: interoperability, latency, and configuration errors

The three issues we encounter most often are:

• Interoperability — IEDs from different vendors interpret the standard differently, particularly around optional fields and dataset ordering.
• Latency — GOOSE transfer times must meet the performance class declared in the SCL file, typically 3 ms for type 1A applications.
• Configuration errors — wrong dataset members, mismatched APPIDs, stale CID files, and duplicate GOOSE control block references.

Pre-Commissioning Checks and Documentation

Most IEC 61850 problems are born long before energisation. We insist on a full documentation review before any IED is powered up on site, because fixing a configuration error in the office costs minutes, while fixing it during a commissioning window costs days.

Reviewing SCL, CID, and ICD files

The Substation Configuration Language (SCL) is the single source of truth for the whole scheme. We review the SSD, ICD, and SCD files for consistency, then confirm that the CID file loaded into each IED matches the issued revision. Vendor tools such as Siemens DIGSI, ABB PCM600, and GE Grid Solutions’ EnerVista each handle SCL import slightly differently, so we check the exported CID against the system configuration rather than trusting the tool’s summary view.

Network topology and VLAN validation

We validate the network design against the project’s communication specification, checking VLAN IDs, priority tagging, multicast addressing, and redundancy protocol settings (PRP or RSTP). UK transmission standards, including National Grid’s NGET documentation and guidance published by the Energy Networks Association (ENA), set expectations around network segregation and cyber security that must be reflected in the switch configuration before FAT.

FAT (Factory Acceptance Testing) best practices

Factory acceptance testing should exercise GOOSE publish/subscribe behaviour, MMS reporting, and time synchronisation — not just point-to-point wiring. We ask for a full FAT report with captured network traffic, because a screenshot of a green status indicator proves very little. Any FAT deviation should be closed out before the panels ship.

Step-by-Step Commissioning Process

Our on-site sequence follows a fixed order. Skipping ahead — particularly to functional tests — is how teams end up chasing ghosts.

Physical and wiring checks

We confirm panel build quality, fibre polarity, Ethernet port labelling, and correct CT/VT connections to merging units. Under UK safety rules, all work is controlled through the site permit-to-work system, and we never energise a bay without a signed sanction for test.

Network and IED configuration

Switch configurations are loaded and verified, then each IED is checked for correct IP addressing, subnet mask, gateway, and VLAN membership. We confirm the CID revision in every device matches the issued design.

GOOSE messaging tests: publish, subscribe, and performance

This is where the real work happens. Using OMICRON CMC 356 test sets and Wireshark captures, we verify that each publisher transmits the correct dataset, that subscribers accept it, and that retransmission behaviour matches the configured performance class. We deliberately break the link to confirm the subscriber’s loss-of-GOOSE supervision alarms correctly.

MMS and SCADA integration checks

We verify MMS reporting to the substation HMI and to the control centre, checking report control block behaviour, buffered vs unbuffered reporting, and data quality flags.

Functional and end-to-end tests

Finally, we run end-to-end functional tests: primary injection where practical, interlock proving, trip and close circuits, and SCADA point-to-point verification against the signal list.

Common Pitfalls and How to Avoid Them

Across projects with UK DNOs including UK Power Networks and SSEN, the same categories of fault keep appearing.

GOOSE configuration errors

Incorrect dataset mapping, mismatched GOOSE control block references, and stale CID files are the most common causes of protection failures. We mitigate this with automated SCL comparison tools and a mandatory peer review of every dataset.

Time synchronisation and latency issues

Mixing PTP (IEEE 1588) and NTP across a site creates timestamp inconsistencies that break disturbance recording and Sampled Value alignment. We standardise on one time source per substation and verify accuracy at each IED.

Interoperability between vendors

Where possible, we test cross-vendor GOOSE and MMS exchanges during FAT rather than discovering incompatibilities on site. Documenting vendor firmware versions in the handover pack saves enormous time later.

Tools and Test Equipment We Rely On

Our standard kit list has been refined over years of UK substation work, including National Grid’s digital substation trials.

GOOSE and SV analysis tools

Wireshark with the GOOSE and SV dissectors, plus IEDScout for live subscription monitoring and simulation.

Multifunction test sets

The OMICRON CMC 356 remains our primary injection and GOOSE simulation platform, with the Doble F6150 used for high-burden protection testing.

Network analysis and monitoring

Managed switches with port mirroring, plus permanent network recorders on critical buses to capture intermittent faults.

Post-Commissioning and Documentation

Commissioning is not finished when the bay energises. The documentation pack is what keeps the substation maintainable for the next twenty years.

As-built SCL and CID updates

Every change made on site must be reflected back into the SCL and re-issued as a controlled CID. We treat the as-built SCL as a safety-critical document.

Test reports and sign-off

Test records must satisfy the Electricity Safety, Quality and Continuity Regulations (ESQCR) and the client’s own assurance processes. We compile results, Wireshark captures, and configuration backups into a single indexed pack.

Handover to operations and maintenance

Operations teams need training on GOOSE supervision alarms, SCL revision control, and network troubleshooting — not just a folder of PDFs.

FAQ

How long does IEC 61850 commissioning typically take?

For a single 132 kV bay with digital protection, we usually allow five to eight days on site, plus pre-commissioning review time. Full substations with multiple bays and cross-bay GOOSE interlocking take considerably longer, and the schedule should reflect that.

Can we commission IEC 61850 systems without a dedicated network engineer?

Not safely. VLAN design, multicast behaviour, and redundancy protocols require network expertise. We always include a network engineer in the commissioning team, even on smaller distribution projects.

What is the difference between GOOSE and MMS in commissioning terms?

GOOSE is a fast, peer-to-peer multicast protocol used for protection and interlocking, tested with publish/subscribe and performance checks. MMS is a client-server protocol used for SCADA and engineering access, tested with report control block and point-to-point verification.

Do UK regulations specifically cover IEC 61850 substations?

ESQCR sets the overarching safety and continuity obligations, while the Energy Networks Association (ENA) and transmission operators such as National Grid Electricity Transmission publish technical standards that shape how IEC 61850 schemes are designed, tested, and documented.

Thorough testing and commissioning are what stand between a well-designed digital substation and an unreliable one. As smart grid functionality expands across the UK — from adaptive protection to wide-area monitoring — utilities must invest in skilled commissioning teams and the right test equipment. The technology is proven; the discipline of verifying it is what keeps the lights on.