Zero recovery time
PRP and HSR send every frame twice. No frames are lost during a link inoperability.
10.3 HSR and PRP
The previous chapter compared PROFINET and EtherNet/IP. Both protocols rely on MRP for ring redundancy. MRP recovers in less than 200 ms. For IEC 61850 substations and safety systems, 200 ms is too slow. Protection relays need zero packet loss. HSR and PRP deliver zero-recovery-time redundancy.
Why Zero Recovery Time Matters
Section titled “Why Zero Recovery Time Matters”When a protection relay sends a GOOSE message to trip a circuit breaker, the message arrives within 4 ms or the breaker does not trip. A 200 ms MRP failover during that window causes the message to be lost. HSR (High-availability Seamless Redundancy) and PRP (Parallel Redundancy Protocol), defined in IEC 62439-3, solve this requirement. Both protocols send every frame twice over 2 independent paths simultaneously. The receiver accepts the first copy and discards the duplicate.
PRP — Parallel Redundancy Protocol
Section titled “PRP — Parallel Redundancy Protocol”PRP connects each device to 2 independent networks (LAN A and LAN B). Every frame is sent simultaneously on both networks. The receiver uses a RCT (Redundancy Control Trailer), a 6-byte trailer containing a sequence number and LAN identifier, to detect and discard duplicates.
A DANP (Doubly Attached Node for PRP) has 2 network interfaces. 1 interface connects to each network. Standard switches work on both LAN A and LAN B. PRP requires no special switch support.
HSR — High-availability Seamless Redundancy
Section titled “HSR — High-availability Seamless Redundancy”HSR uses a ring topology. Each device has 2 ports connected to the ring. Frames travel in both directions simultaneously. The originating node removes frames that complete the full ring.
A RedBox (Redundancy Box) connects non-HSR devices to an HSR ring. The RedBox handles dual-frame transmission on behalf of non-HSR devices.
HSR vs PRP vs MRP
Section titled “HSR vs PRP vs MRP”| Feature | HSR | PRP | MRP |
|---|---|---|---|
| Recovery time | Zero | Zero | < 200 ms |
| Topology | Ring | Dual network | Ring |
| Switch requirement | HSR-capable | Standard | MRP-capable |
| Bandwidth overhead | 2x frames | 2x frames | Minimal |
| Use case | IEC 61850 substations | IEC 61850 substations | Factory automation |
Key Takeaways
Section titled “Key Takeaways”Use for substations, not factory automation
HSR/PRP are for substations and safety systems. MRP is simpler and sufficient for factory automation.
What Comes Next
Section titled “What Comes Next”HSR, PRP, and MRP solve redundancy for today’s industrial protocols. The next chapter covers TSN (Time-Sensitive Networking). TSN is the IEEE standard that adds deterministic, time-aware behavior to standard Ethernet. TSN enables multiple industrial protocols to coexist on the same network.
References
Section titled “References”- IEC 62439-3:2016 — High availability automation networks, Part 3: PRP and HSR
- Kirrmann, H., & Dzung, D. (2007). Standardizing a Redundancy Protocol for Industrial Ethernet Networks. IEEE ETFA 2007.
- Hirschmann. (2023). Application Note: HSR and PRP with Hirschmann Switches. Belden/Hirschmann.