GOOSE Messaging: How It Works

GOOSE Messaging: How It Works

At a digital substation in the north of England, a phase-to-earth fault appears on a 132 kV feeder. Before the protection engineer’s coffee has settled in the cup, a trip signal has already left the relay, crossed the station bus, and arrived at the circuit breaker — total elapsed time, under 4 milliseconds. There is no copper pilot wire between them, no hard-wired trip contact, no conventional wiring at all. So how does a trip signal travel that fast over an Ethernet network? The answer is GOOSE — and once you understand how it works, you understand the beating heart of the modern digital substation.

What Is GOOSE in IEC 61850?

What Is GOOSE in IEC 61850?

GOOSE stands for Generic Object Oriented Substation Event. It is a Layer 2 multicast protocol designed for time-critical, peer-to-peer communication between intelligent electronic devices (IEDs) inside a substation. Unlike client-server protocols, GOOSE does not wait to be asked for data. A publisher IED simply broadcasts its message onto the network, and every subscriber that cares about that data picks it up immediately.

The protocol is defined in IEC 61850-8-1, which maps GOOSE (along with MMS) onto the Ethernet stack. A single GOOSE message can carry status indications, measurements, interlocking signals, and — most importantly — protection trip commands. It is the mechanism that lets IEC 61850 replace thousands of metres of hard-wired copper with a handful of Ethernet frames.

GOOSE vs MMS vs Sampled Values

IEC 61850 uses three main communication services, and it helps to keep them straight:

  • MMS (Manufacturing Message Specification) — client-server, routable over IP, used for supervisory control, reporting, and configuration. Fast enough for SCADA, far too slow for protection.
  • GOOSE — Layer 2 multicast, non-routable, event-driven, used for fast status and trip signals between IEDs. This is the protection workhorse.
  • Sampled Values (SV) — also Layer 2 multicast, but streams digitised current and voltage samples from merging units to protection relays at around 80 to 256 samples per cycle.

In short: MMS supervises, SV measures, and GOOSE acts.

Why GOOSE Replaced Hard-Wired Copper

Traditional substations used dedicated copper circuits for every trip, interlock, and status signal. A typical bay could consume kilometres of cable, hundreds of terminations, and weeks of point-to-point testing. GOOSE collapses all of that into a single Ethernet connection per IED. Signals become data attributes in a dataset, and adding a new interlock means editing a configuration file rather than pulling new cable. That reduction in copper also removes a long-standing failure mode: loose terminals and damaged cores.

The GOOSE Publisher and Subscriber Model

GOOSE operates on a publish-subscribe model. One IED publishes a GOOSE message to a multicast MAC address; any number of subscriber IEDs listen to that address without sending a request. The publisher does not know — or care — who is listening. This one-to-many design is what makes GOOSE so efficient for protection schemes where a single trip signal may need to reach several devices simultaneously.

In practice, UK transmission and distribution substations mix vendors freely. A GE Grid Solutions relay might publish a trip that a Siemens SIPROTEC relay subscribes to, while an ABB REF615 monitors the same dataset for interlocking. Because GOOSE is an open standard, interoperability between these vendors is achievable — provided the engineering is done correctly.

GOOSE Control Blocks and DataSets

Every GOOSE stream is anchored by a GOOSE Control Block, usually abbreviated as gCB in the SCL configuration. The gCB defines the multicast address, the dataset it publishes, the configuration revision number, and the retransmission parameters. The dataset itself is an ordered list of data attributes — for example, a position indication, a trip command, and a blocking signal — that travel together in every frame. Subscribers reference the dataset by name, so the publisher and subscriber must agree on the exact structure.

Retransmission, Heartbeat, and Time-to-Live

GOOSE does not send a message only when something changes. Instead, it retransmits continuously using a scheme designed for both speed and reliability:

  1. On a state change, the message is sent immediately, then retransmitted with exponential back-off — typically at 2 ms, 4 ms, 8 ms, and so on.
  2. Once the back-off interval reaches a configured maximum, the publisher settles into a steady heartbeat, often every 1 to 5 seconds.
  3. Each frame carries a Time-to-Live (TTL) value. If a subscriber stops receiving messages within the TTL window, it knows the publisher or the network has failed and can raise an alarm or trigger a fallback scheme.

This combination means a subscriber gets millisecond-speed notification of a change, plus continuous confirmation that the publisher is alive.

How GOOSE Messages Travel Across the Network

GOOSE frames are Ethernet Layer 2 multicast. They carry no IP header, which means they are not routed — they live entirely within a single broadcast domain, typically a substation LAN. That is deliberate: routing adds latency and complexity that protection schemes cannot tolerate. The Ethernet frame carries a destination multicast MAC address, a VLAN tag, and the GOOSE payload itself.

In UK digital substations, it is now common practice to separate the process bus (carrying Sampled Values from merging units) from the station bus (carrying GOOSE and MMS). This separation limits fault propagation and makes traffic analysis far easier during commissioning.

VLAN Priority and Traffic Separation

GOOSE frames are tagged with IEEE 802.1Q VLAN priority values typically between 4 and 7, placing them among the highest-priority traffic on the network. Switches honour these priorities, ensuring that a burst of SCADA polling or file transfer never delays a trip signal. VLANs also isolate GOOSE traffic from non-protection devices, which is essential both for performance and for security.

PRP, HSR, and Redundancy in UK Substations

Because a lost GOOSE frame could mean a missed trip, redundancy is mandatory in critical schemes. Two IEC 61850 standards dominate:

  • PRP (Parallel Redundancy Protocol) — the IED sends duplicate frames over two independent networks, and the receiver discards the second copy. Zero recovery time.
  • HSR (High-availability Seamless Redundancy) — a ring topology where each node forwards frames in both directions, again with zero switchover time.

UK network operators, including National Grid and SSEN, have adopted these redundancy schemes in their digital substation pilots and live deployments to maintain protection integrity even if one network path fails.

GOOSE Performance and Timing Requirements

IEC 61850 defines performance classes for different message types. GOOSE trip signals fall into the most demanding category — Type 1A fast messages — which must be transmitted and received within 3 milliseconds at the publisher and processed by the subscriber within a further 3 milliseconds. In well-engineered schemes, end-to-end GOOSE trip times of under 4 ms are routinely achieved, matching or beating conventional hard-wired protection.

This performance is not theoretical. National Grid has trialled digital substations across its UK transmission network, demonstrating GOOSE-based protection operating within these limits. SSEN has also deployed GOOSE in active distribution projects, using it for interlocking, blocking, and trip transfer between bays.

Type 1A Fast Messages and 4 ms Trips

The 4 ms figure is not a marketing claim — it is a design requirement. To meet it, engineers must control every element of the chain: IED processing time, switch latency, network load, and cable length. A single misconfigured switch port or an overloaded VLAN can push a trip past the limit, which is why factory acceptance testing and site commissioning are so rigorous.

Testing GOOSE with Omicron and Doble

Testing GOOSE requires tools that can both publish and subscribe to GOOSE streams while injecting fault conditions. Omicron’s CMC test sets and Doble’s protection test equipment are widely used in the UK for this purpose. Engineers verify subscription behaviour, measure end-to-end latency, simulate publisher failure to confirm TTL handling, and stress the network to ensure priority tagging holds under load.

Security, Configuration, and Common Pitfalls

GOOSE in its base form has no authentication. Any device on the same VLAN can publish a frame claiming to be a legitimate relay. This is a real risk in an era of increasing OT cyber threats, and it is why IEC 62351 — the security standard for power system communications — extends IEC 61850 with authentication and encryption mechanisms for GOOSE and other protocols.

In the UK, operators of critical national infrastructure must also comply with the NIS Regulations and follow NCSC guidance for OT security. That means network segmentation, strict VLAN management, and controlled access to engineering workstations.

IEC 62351 and GOOSE Authentication

IEC 62351-6 specifies how to secure GOOSE using digital signatures and, in more recent editions, encryption. Because authentication adds processing overhead, it must be applied carefully to avoid breaching the 4 ms timing budget. Many UK operators currently rely on network segregation and monitoring as their primary defence, with IEC 62351 authentication being phased in for the most critical inter-substation links.

SCL, CID Files, and Engineering Mistakes

Every GOOSE stream is described in an SCL file — specifically the CID (Configured IED Description) that is loaded into each device. The most common causes of GOOSE failure we see in the field are configuration errors rather than hardware faults:

  • Mismatched dataset structures between publisher and subscriber.
  • Incorrect or duplicate multicast MAC addresses.
  • VLAN ID mismatches that silently drop frames.
  • Outdated CID files after a firmware upgrade.

Careful version control and a formal change process for SCL files prevent most of these issues.

FAQ

What does GOOSE stand for in IEC 61850?

GOOSE stands for Generic Object Oriented Substation Event. It is a Layer 2 multicast protocol defined in IEC 61850-8-1 for fast peer-to-peer communication between IEDs.

How fast is a GOOSE trip signal?

GOOSE trip signals are classified as Type 1A fast messages and are designed to complete end-to-end in under 4 milliseconds. In well-engineered digital substations, this performance is routinely achieved.

Does GOOSE use IP addressing?

No. GOOSE operates directly at Ethernet Layer 2 using multicast MAC addresses. It bypasses IP routing entirely, which keeps latency low and predictable.

Is GOOSE secure?

In its base form, GOOSE has no authentication. IEC 62351 provides security extensions including digital signatures, and UK operators also rely on network segmentation and NCSC guidance to protect OT environments.

Can GOOSE work between different vendors’ relays?

Yes. GOOSE is an open standard, and interoperability between vendors such as GE Grid Solutions, Siemens, and ABB is achievable — provided the SCL configuration and datasets are aligned correctly.

GOOSE is the backbone of the modern digital substation. It delivers the millisecond performance that protection schemes demand, while eliminating the copper, terminations, and complexity of conventional wiring. But that performance is only as good as the engineering behind it — correct SCL configuration, disciplined VLAN design, rigorous testing with tools like Omicron and Doble, and a security posture aligned with IEC 62351 and UK regulatory expectations. Get those right, and GOOSE will deliver on its millisecond promise every time.

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *