What Is IEC 61850?
IEC 61850 is the international standard for communication in electrical substations and power grid automation. Published by the International Electrotechnical Commission (IEC), it defines a unified data model, communication services, and protocols that allow intelligent electronic devices (IEDs) from different manufacturers to interoperate within a digital substation.
Before IEC 61850, substation communication relied on proprietary protocols and point-to-point copper wiring. IEC 61850 replaces these with a standardized Ethernet-based architecture, enabling faster protection operation, simplified configuration, and reduced wiring costs. It defines three primary communication services: GOOSE, MMS, and Sampled Values (SV) — each serving a distinct role in the substation communication hierarchy.
Station Bus vs. Process Bus
IEC 61850 defines two logical communication buses within a digital substation:
| Feature | Station Bus | Process Bus |
|---|---|---|
| Location | Within the substation control room / relay room | Between the control room and primary switchyard equipment (transformers, circuit breakers) |
| Connected Devices | IEDs (relays, controllers), HMI, SCADA gateway | Merging Units (MUs), Current/Voltage transformers (CT/VT), circuit breaker interfaces |
| Primary Protocol | GOOSE, MMS | Sampled Values (SV), GOOSE |
| Main Purpose | Relay-to-relay protection coordination, engineering/monitoring access | Digital transmission of analog measurements (current, voltage) from primary equipment to IEDs |
| Latency Requirement | <4 ms (GOOSE), <1 second (MMS) | <1 ms (SV must be synchronized within microseconds) |
In a conventional substation, analog current and voltage signals travel via copper wiring from CT/VT to protection relays. In a digital substation, Merging Units (MUs) digitize these signals and transmit them as IEC 61850-9-2 Sampled Values over the Process Bus Ethernet network to IEDs — replacing kilometers of copper wiring with optical fiber.
What Is GOOSE?
Generic Object-Oriented Substation Event (GOOSE) is defined in IEC 61850-8-1. It is a high-speed, publisher-subscriber Ethernet multicast protocol designed for time-critical protection and control messaging in substations.
How GOOSE works:
- An IED (the publisher) broadcasts a GOOSE message directly at the Ethernet layer (Layer 2 multicast) — bypassing TCP/IP entirely for minimum latency.
- All subscribed IEDs receive the message and act on it immediately.
- GOOSE messages use a retransmission scheme: after an event, the message is sent immediately, then retransmitted at increasing intervals (T1 < T2 < T3...) until the next state change or heartbeat — ensuring delivery even under brief Ethernet congestion.
Key GOOSE characteristics:
- Protocol: EtherType 0x88B8 (direct Ethernet multicast, no IP/TCP/UDP)
- Latency requirement: Class P1 ≤ 3 ms, Class P2 ≤ 4 ms (from event to all subscribers)
- Use cases: Trip signals, interlocking between circuit breakers, busbar protection, auto-reclosing sequences, distributed protection schemes
- QoS: GOOSE messages carry 802.1p priority = 4 (minimum) to 6 for trip signals — must be prioritized by all network switches
What Is MMS?
Manufacturing Message Specification (MMS), defined in IEC 61850-8-1 using the MMS protocol (ISO 9506), is a client-server protocol for non-time-critical communication between IEDs and engineering tools, SCADA systems, and human-machine interfaces (HMIs).
How MMS works:
- MMS operates over TCP/IP (Ethernet → IP → TCP → MMS) — the standard internet stack.
- An MMS server (typically the IED) exposes a data model of its logical nodes, data objects, and attributes — defined per IEC 61850-7-x.
- MMS clients (SCADA, engineering workstations, HMI) connect via TCP to read/write values, configure settings, or retrieve event logs.
Key MMS characteristics:
- Protocol: TCP/IP (typically port 102 for ACSE/ISO 8327)
- Latency requirement: 100 ms to 1 second (non-time-critical reporting and control)
- Use cases: Remote configuration of relay settings, SCADA data acquisition, fault record retrieval, system monitoring, event logging
- Reporting: Buffered Reporting (BRCB) for event delivery with guaranteed no-data-loss, and Unbuffered Reporting (URCB) for near-real-time streaming
What Are Sampled Values (SV)?
Sampled Values (SV), defined in IEC 61850-9-2, transmit digitized current and voltage waveforms from Merging Units (MUs) to protection IEDs over the Process Bus. SV replaces the traditional analog copper CT/VT wiring with optical fiber and Ethernet.
How SV works:
- A Merging Unit samples analog current and voltage signals at 80 samples/cycle (4000 samples/second at 50 Hz) or higher.
- These samples are packaged as SV messages (EtherType 0x88BA) and sent as Ethernet multicast at precise intervals synchronized by IEEE 1588 PTP.
- Protection IEDs subscribe to the SV stream and reconstruct the waveforms to perform differential protection, overcurrent detection, or power quality measurement.
Key SV characteristics:
- Protocol: EtherType 0x88BA (direct Ethernet multicast, no IP)
- Latency requirement: <1 ms, with microsecond-level synchronization via IEEE 1588 PTP
- Synchronization: All MUs must be synchronized to a common PTP grandmaster to ensure SV samples are coherent across IEDs
- Use cases: Transformer differential protection, busbar differential protection, distance protection, power quality monitoring
GOOSE vs. MMS vs. Sampled Values (SV)
| Feature | GOOSE | MMS | Sampled Values (SV) |
|---|---|---|---|
| IEC 61850 Standard | 8-1 | 8-1 | 9-2 |
| EtherType / Protocol | 0x88B8 (Ethernet multicast) | TCP/IP (port 102) | 0x88BA (Ethernet multicast) |
| Communication Model | Publisher-Subscriber | Client-Server | Publisher-Subscriber |
| Latency Class | P1: ≤3 ms / P2: ≤4 ms | 100 ms – 1 second | <1 ms |
| Primary Use | Protection trip signals, interlocking | Configuration, monitoring, SCADA | Digital CT/VT data to protection IEDs |
| Bus | Station bus | Station bus | Process bus |
| Requires IP | No | Yes | No |
| Requires PTP Sync | No (event-driven) | No | Yes (IEEE 1588) |
| QoS Priority (802.1p) | 4–6 (trip messages = 6) | 2–4 (normal data) | 4–6 (critical measurement) |
Network Design Requirements: VLAN, QoS, PTP and Redundancy
An IEC 61850 Ethernet network must be carefully designed to meet the latency and reliability requirements of GOOSE, MMS, and SV traffic:
- VLAN segmentation: Separate VLANs for GOOSE/SV (process bus), MMS (station bus control), engineering access, and SCADA data — preventing broadcast storms from affecting protection-grade traffic. IEC 61850 recommends at minimum separating GOOSE/SV from MMS traffic.
- QoS (802.1p): All IEC 61850-3 compliant switches must implement IEEE 802.1p priority queuing. GOOSE trip messages use PCP = 6; SV messages use PCP = 4 minimum; MMS and engineering traffic use lower priorities (PCP 2–4).
- IEEE 1588 PTP (Precision Time Protocol): Required for SV synchronization on the Process Bus. The network must support transparent clocks (TC) or boundary clocks (BC) in each switch to maintain microsecond timing accuracy across all Merging Units.
- Redundancy: IEC 62439-3 defines High-availability Seamless Redundancy (HSR) and Parallel Redundancy Protocol (PRP) for IEC 61850 networks. HSR provides zero-switchover-time redundancy for ring topologies (Process Bus); PRP provides parallel path redundancy for star topologies (Station Bus). No traffic interruption occurs on link failure.
- Latency budget: For GOOSE P1 (≤3 ms end-to-end), network switches must contribute no more than 1 ms per hop. IEC 61850-3 certified switches are designed to meet this requirement.
How HSR / PRP Supports IEC 61850 Networks
High-availability Seamless Redundancy (HSR) and Parallel Redundancy Protocol (PRP), both defined in IEC 62439-3, provide zero-recovery-time network redundancy — essential for protection-grade IEC 61850 applications where even 20ms of packet loss during a fault could delay a trip signal.
- PRP (IEC 62439-3 Clause 4): Devices send identical frames simultaneously over two independent parallel networks (LAN A and LAN B). The receiver accepts the first copy and discards the duplicate. If one network fails, the other delivers without any delay. Ideal for Station Bus applications (MMS, GOOSE) where star topology is preferred.
- HSR (IEC 62439-3 Clause 5): Devices are connected in a ring; each frame is sent in both directions simultaneously. If a link fails, the frame arriving from the other direction still delivers without any interruption. Ideal for Process Bus ring applications (SV, GOOSE) where daisy-chaining of devices is common.
Both protocols require support at the switch level: IEC 61850-3 certified industrial switches from ORing support PRP/HSR forwarding (RedBox / DANP) to integrate non-PRP/HSR devices into the redundant network.
ORing IEC 61850-3 Switch Applications
ORing's IEC 61850-3 and IEEE 1613-certified industrial Ethernet switches are designed for the demands of digital substation communication:
- IEC 61850-3 / IEEE 1613 compliance: Withstand the EMC, temperature, humidity, and vibration requirements of substation environments — including immunity to electrical fast transients, magnetic fields, and surge voltages common near high-voltage equipment.
- GOOSE / SV-aware forwarding: Hardware-accelerated forwarding of Ethernet multicast traffic at sub-millisecond latency, with 802.1p QoS queue enforcement for GOOSE and SV priority.
- IEEE 1588 PTP transparent clock: Maintains microsecond timing accuracy for SV synchronization across the Process Bus without introducing clock error accumulation.
- HSR / PRP support: RedBox and DANP functionality to connect non-HSR/PRP legacy devices into the redundant IEC 61850 network.
- Layer 3 routing: For large substations with multiple IEC 61850 segments, ORing L3 managed switches provide inter-VLAN routing with OSPF support — enabling MMS connectivity across geographically distributed protection zones.
Frequently Asked Questions
Q: What is the difference between GOOSE and MMS in IEC 61850?
A: GOOSE is a high-speed Layer 2 multicast protocol for time-critical protection events (trip signals, interlocking) with latency requirements under 4 ms. MMS operates over TCP/IP for non-time-critical configuration, monitoring, and SCADA data exchange, with acceptable latency of 100 ms to 1 second. Both serve distinct roles and coexist on the IEC 61850 station bus.
Q: Why does IEC 61850 use Layer 2 multicast for GOOSE and SV instead of TCP/IP?
A: TCP/IP adds protocol overhead and non-deterministic latency from TCP handshaking and retransmission mechanisms — unacceptable for protection-grade messages that must arrive within 3 ms. GOOSE and SV bypass IP entirely, using direct Ethernet multicast (EtherType 0x88B8 and 0x88BA) to achieve deterministic, wire-speed delivery. This requires all network switches to implement proper multicast filtering (IGMP snooping or static multicast filtering) to contain GOOSE/SV traffic to intended subscribers.
Q: What is a Merging Unit and why is it used?
A: A Merging Unit (MU) is a device that digitizes analog current and voltage signals from conventional CT/VT instrument transformers and transmits the digitized samples as IEC 61850-9-2 Sampled Values over the Process Bus. MUs enable the replacement of thick copper wiring between the switchyard and relay room with optical fiber and Ethernet, reducing cost, installation complexity, and potential for ground loops.
Q: Does every switch in an IEC 61850 network need to be certified?
A: For protection-grade applications (GOOSE, SV), all switches in the data path should be IEC 61850-3 / IEEE 1613 compliant to guarantee EMC immunity in the substation environment and to support QoS and latency requirements. Non-certified commercial switches risk malfunction from electrical transients common near high-voltage equipment and may not meet the latency budget for GOOSE P1/P2 classes.
Q: What is PRP and how is it different from ring redundancy protocols like O-Ring or MRP?
A: PRP (Parallel Redundancy Protocol) connects each device to two completely independent networks simultaneously and sends duplicate frames on both. There is zero recovery time because the device is always connected to both networks. O-Ring and MRP are ring recovery protocols that detect a ring break and reroute traffic — they provide sub-20ms or sub-200ms recovery, but there is a brief period of packet loss during switchover. For IEC 61850 protection applications where even 20ms of loss is unacceptable, PRP or HSR is required.
