Parallel Redundancy Protocol (PRP)

A complementary zero-time recovery protocol that creates a ring topology rather than PRP's dual-star architecture. Each frame is sent in both directions around the ring, and the destination node accepts the first copy while discarding the duplicate. Unlike PRP, HSR does not require fully duplicated network infrastructure, making it more cost-effective for daisy-chained topologies. However, HSR introduces cumulative latency as frames traverse multiple hops, whereas PRP maintains consistent latency across both LANs. Both protocols are standardized in IEC 62439-3 and can be combined in hybrid PRP-HSR architectures where PRP connects redundant rings.

Defined by IEEE 1588, PTP provides sub-microsecond clock synchronization across Ethernet networks—a critical prerequisite for PRP's duplicate frame handling. When PRP nodes receive identical frames on two LANs, they must identify duplicates using sequence numbers within a bounded time window. Without precise synchronization, stale sequence numbers could cause false duplicate detection or missed duplicates. PTP's grandmaster clock architecture ensures all DANPs (Doubly Attached Nodes) share a common time reference, enabling deterministic duplicate discard logic even during network asymmetry.

The international standard that defines both PRP and HSR as Clause 4 and Clause 5 respectively. IEC 62439-3 specifies:

• Frame format: The Redundancy Control Trailer (RCT) appended to each frame
• Duplicate detection: Sequence number generation and discard algorithms
• Node types: DANP (Doubly Attached Node), RedBox (Redundancy Box), and VDAN (Virtual DAN)
• Interoperability: Requirements for mixing PRP and HSR domains The standard ensures that devices from different manufacturers can interoperate in redundant networks without proprietary extensions.

A network device that bridges PRP/HSR redundant domains with conventional single-attached Ethernet devices. A RedBox acts as a proxy, duplicating frames from a singly attached node (SAN) onto both PRP LANs and performing duplicate discard in the reverse direction. This allows legacy IEDs, SCADA systems, and engineering workstations that lack native PRP support to participate in a redundant substation network. RedBoxes are often integrated into Ethernet switches or ruggedized substation routers and must maintain the same zero-recovery performance as native DANPs.

An alternative redundancy mechanism defined by IEEE 802.1w that provides recovery times in the range of 1-2 seconds—orders of magnitude slower than PRP's zero-time failover. RSTP operates by detecting topology changes and recalculating loop-free paths, which is unacceptable for GOOSE tripping messages that require sub-millisecond delivery. However, RSTP remains widely deployed in station bus networks where brief interruptions are tolerable. PRP is specifically designed for process bus applications where any frame loss during a fault event could prevent protective relay operation.

The IEC 61850 publisher-subscriber protocol that carries time-critical protection signals—and the primary payload that PRP is designed to protect. GOOSE messages transmit trip commands, interlocking signals, and breaker status with strict latency requirements (typically <3 ms). PRP's zero-recovery characteristic ensures that even if one LAN fails during a fault event, GOOSE frames continue flowing uninterrupted on the parallel LAN. The combination of PRP and GOOSE enables hardwired replacement strategies where copper control cables are eliminated in favor of redundant Ethernet.