A substation automation network connects equipment with very different jobs. Operators need a clear view of the station, protection devices must exchange time-sensitive information, and field equipment must deliver measurements and respond to control commands. The challenge is not simply to connect every device: the network must keep these functions coordinated when a communication link fails and preserve a consistent sense of time across the station.
A common approach divides the system into three layers: station, bay and process. Industrial Ethernet switches provide the communication paths between devices, while separate networks, redundancy mechanisms and time synchronization help support reliable operation. Understanding how those elements fit together makes it easier to evaluate a network for a new smart substation or an automation upgrade.
How the Three Layers Work Together
The station layer is where monitoring and management functions come together. It typically includes hosts, operator workstations, remote terminal equipment and a GPS/BeiDou clock. Operators use the station-level systems to monitor equipment and manage station operations. Separate station-layer A and B networks provide redundant communication paths.
The bay layer contains devices responsible for individual sections of the electrical installation and their associated protection and control functions. Typical equipment includes measurement and control devices, transformer and line protection devices, busbar protection, automatic transfer equipment, meters and disturbance recorders. Protocol converters may also be used where devices need help exchanging data.
The process layer connects communication to field measurements and equipment. Merging units collect measurement information, while intelligent electronic terminals interact with equipment such as primary switches. Electronic current transformers, transformer and bay merging units, and intelligent terminals are among the devices that can be connected through process-layer A and B networks.
This division makes the roles of the equipment easier to manage: the process layer handles field interaction, the bay layer performs protection and control functions, and the station layer brings information together for supervision. The layers still depend on one another, so their communication and timing requirements must be considered as a system.
Why Substation Networks Use Different Message Types
IEC 61850, also known in China as DL/T 860, provides a framework for communication between substation automation devices. Three communication types are particularly useful to distinguish:
- MMS (Manufacturing Message Specification) supports monitoring and control communications, especially at the station layer.
- GOOSE (Generic Object Oriented Substation Event) supports fast protection and control messages, including protection tripping coordination.
- SV (Sampled Values) carries sampled measurement data used by process-layer and protection equipment.
These message types serve different purposes. An operator workstation receiving monitoring information does not necessarily have the same timing needs as protection equipment exchanging a fast control message or a device receiving sampled values. Network design must therefore account for the traffic each layer carries, rather than treating every connection as ordinary office data.
Station-layer and process-layer networks can be built independently, with A and B networks used for redundant communication. Separating the networks helps organize traffic according to its role, while the availability of redundant paths reduces dependence on a single communication path.
Double Networks, Ring Recovery and Their Limits
A/B redundancy uses two networks to provide alternative communication paths. In a substation, both the station layer and the process layer can use this arrangement. The intended benefit is greater communication availability when a link or network component fails.
A ring is another way to provide an alternative path. If a link in a properly designed ring fails, the network can restore communication through the remaining path. Ring recovery and A/B networks address resilience in different ways, so the chosen topology and redundancy method need to match the devices and the protection requirements.
Technologies used in redundant industrial networks include RSTP, PRP and HSR. Their behavior is not interchangeable: the actual effect of a failure depends on the protocol, network topology and equipment configuration. A redundancy design should therefore be verified against the required recovery behavior rather than described as universally interruption-free.
Bandwidth and delay matter alongside redundancy. Process-layer communications must carry measurement and control data with sufficiently low delay, while the station network must support monitoring and management traffic. The appropriate capacity depends on the connected devices and the traffic they generate.
Why Accurate Time Matters
Substation devices need a consistent time reference so that monitoring, protection and recording functions can be coordinated. A GPS/BeiDou clock can supply the station’s time reference, with synchronization methods such as SNTP and IRIG-B used where appropriate.
SNTP, or Simple Network Time Protocol, distributes time over a network and can support station-layer synchronization. IRIG-B is another time synchronization method used in the station. The synchronization method and distribution path must match the precision required by each device.
The time requirements differ by layer. Station-layer synchronization can be at the millisecond level, while the process layer can require microsecond-level synchronization. A millisecond is one thousandth of a second; a microsecond is one millionth. The tighter process-layer requirement reflects the need to coordinate time-sensitive field measurements and protection functions. A shared clock alone does not establish that every connected device meets its required accuracy: the complete synchronization arrangement matters.
From Network Design to Equipment Selection
Industrial switches act as communication hubs between the devices in each layer. In Wintop’s substation automation architecture, the YT-CM6028-Z-E24G4 is an example of a high-capacity switch used for station-layer communication. The YT-CM6016-G-G16 and YT-CM6020-G-F16G4 are examples used for process-layer networking. Their roles are different because the connected devices and communication demands differ; the model names alone do not establish identical capabilities.
Selection begins with the network’s job: which devices must connect, which communications they use, whether the design calls for A/B networks or a ring, and how the station distributes time. Equipment also needs to suit the substation environment, where electromagnetic interference is a practical concern and the network is expected to operate continuously. Specific protocol support, interfaces and environmental ratings should be checked against the documentation for each model.
The same principles extend beyond substations, but the details change with the application. In an energy storage station, for example, the network brings together battery management systems, power conversion systems, storage cabinets and monitoring servers. Wintop’s storage-monitoring architecture uses layered aggregation and redundant networks to support that mix of equipment. The common task is dependable data exchange; the device types, traffic and operating requirements determine the network design.
For a substation, reliable communication comes from the relationship between architecture, redundancy, message types and time synchronization, not from a switch considered in isolation. When those elements are designed around the protection, measurement and monitoring functions they support, the network can provide a clearer foundation for both day-to-day operation and future expansion.