GOOSE vs Hardwired Interlocking & Tripping
Evaluating protection signaling approaches for substation automation
GOOSE Strengths
- + Scales without exponential wiring growth
- + Logic changes require config, not re-wiring
- + Sub-4ms signaling — faster than relay contacts
- + Built-in diagnostics and health monitoring
- + PRP/HSR network redundancy available
- - Requires IEC 61850 engineering expertise
- - Cross-vendor interoperability needs careful FAT
- - Longer initial commissioning timeline
Hardwired Strengths
- + Any electrician can troubleshoot with a multimeter
- + No software configuration required
- + Completely vendor-agnostic — copper is copper
- + Decades of proven track record
- - Wiring grows combinatorially with interlock points
- - Logic changes require physical re-wiring during outage
- - No built-in diagnostics or remote monitoring
Quick Comparison
| Dimension | GOOSE | Hardwired |
|---|---|---|
| Upfront Cost | Higher — IEC 61850 engineering, SCL config, FAT | Lower — cable, terminals, labor |
| Lifecycle Cost (10yr) | Lower — changes are configuration | Higher — every change is a re-wire |
| Commissioning Time | Longer — FAT, interop testing, network validation | Faster — point-to-point continuity checks |
| Signaling Speed | Sub-4ms (Ethernet multicast) | 5–15ms (relay contact + cable propagation) |
| Scalability | Linear — add GOOSE subscribers to existing network | Combinatorial — n² wiring for n interlock points |
| Troubleshooting Skill | IEC 61850 + Wireshark + SCL knowledge | Multimeter + wiring diagram |
| Vendor Independence | Moderate — IEC 61850 is a standard, but interop varies | Full — copper doesn't care about vendor |
| Modification Ease | Config change, often no outage required | Physical re-wire, requires outage |
| Diagnostics | Built-in — GOOSE health monitoring, network stats | Manual inspection only |
| Redundancy | PRP/HSR gives zero-switchover network redundancy | Duplicate cable runs (expensive, rarely done) |
When to Choose GOOSE
GOOSE is the right choice when the project’s complexity or future flexibility justifies the upfront engineering investment.
Project characteristics that favor GOOSE:
- 50+ interlocking points across multiple bays — this is where hardwired wiring matrices become unmanageable. Each new interlock point adds connections to multiple existing devices, and the wiring grows combinatorially.
- Multi-vendor IED fleet — if you have SEL relays talking to ABB relays talking to Siemens relays, GOOSE on a shared Ethernet network is cleaner than cross-vendor wiring interfaces and interposing relays.
- Station will expand or reconfigure — if future bays, bus reconfigurations, or protection scheme changes are anticipated, GOOSE lets you implement changes in configuration rather than physical re-wiring during an outage.
- Owner requires remote diagnostics — GOOSE publishes health status continuously. You can monitor interlock state, detect communication failures, and verify signal integrity from the control room or remotely. Hardwired gives you nothing until someone walks the yard with a meter.
- Protection scheme changes are anticipated — utility standards evolve, relay firmware updates may change logic, and operational experience often reveals optimization opportunities. GOOSE makes these changes routine; hardwired makes them a project.
- Sub has demonstrated IEC 61850 capability — GOOSE done well is excellent. GOOSE done poorly is worse than hardwired. The sub must prove they can do cross-vendor FAT, SCL file management, and GOOSE dataset validation.
The strongest GOOSE case: A greenfield 345kV substation with 20+ bays, three relay vendors, and an owner who plans to add a second bus tie in 5 years. Hardwiring this would require thousands of interlock connections. GOOSE reduces it to network configuration.
When to Stay Hardwired
Hardwired is not the “old” approach — it’s the appropriate approach when the project doesn’t need GOOSE’s capabilities and would pay GOOSE’s costs without capturing GOOSE’s benefits.
Project characteristics that favor hardwired:
- Small station with under 30 interlock points — the wiring is manageable, the matrix is documentable, and the electricians already know how to test it. GOOSE engineering overhead doesn’t pay for itself at this scale.
- Single-vendor protection scheme — if every relay is from the same vendor, you don’t need GOOSE’s cross-vendor interoperability. The vendor’s native communication methods (mirror bits, direct I/O) are simpler and well-tested.
- No anticipated expansion or reconfiguration — if the station design is frozen and the protection scheme won’t change, the flexibility benefit of GOOSE is moot. You’re paying for optionality you’ll never exercise.
- O&M team lacks IEC 61850 training — a GOOSE scheme that the operations team can’t troubleshoot is a liability. Training programs exist, but if the owner hasn’t invested in them and won’t, hardwired is safer.
- Budget is fixed and front-loaded — GOOSE has lower lifecycle cost but higher upfront cost. If the budget can’t absorb the engineering and FAT phases, hardwired delivers a working system within tighter constraints.
- Regulatory environment requires hardwired backup — some jurisdictions and utilities still require hardwired trip paths as backup to GOOSE, particularly for main protection. Check the utility’s relay protection standards before assuming GOOSE-only is acceptable.
- Harsh environmental conditions — GOOSE requires Ethernet switches, SFP modules, and fiber patch panels that are often rated for narrower temperature and environmental tolerance than protection relays. Industrial-grade managed switches typically rate -40 to +75 degrees C; commercial-grade equipment may be as tight as 0 to +50 degrees C. Protection relays are designed for harsh substation environments and routinely handle wider extremes. The same applies to corrosive atmospheres (coastal salt air, petrochemical H2S, industrial SO2) and high humidity — network equipment with exposed ports, fan ventilation, and commercial-grade conformal coating degrades faster than sealed relay enclosures designed for IEC 60068 environmental classes. In non-climate-controlled relay houses, the network hardware becomes the environmental weak link — forcing either premium industrial-grade switches with conformal coating (higher upfront cost), sealed NEMA 4X enclosures for network gear, or HVAC for the relay house (higher upfront AND ongoing maintenance cost). If the installation is in temperature extremes, corrosive environments, or uncontrolled outdoor enclosures, evaluate whether GOOSE’s network infrastructure requirements will cascade into significant environmental protection costs that hardwired copper and relay contacts avoid entirely.
The strongest hardwired case: A 69kV distribution substation with 6 bays, all SEL relays, no expansion plans, and an O&M team that’s never seen IEC 61850. Forcing GOOSE here is over-engineering.
Cost Analysis
The cost comparison between GOOSE and hardwired depends heavily on project size. At small scale, hardwired is cheaper. At large scale, GOOSE is dramatically cheaper. The crossover depends on your specific labor rates and relay configuration, but typically falls around 30 to 50 interlock points.
Upfront Cost Breakdown
| Cost Category | GOOSE | Hardwired |
|---|---|---|
| Engineering | IEC 61850 system specification, SCL file development, GOOSE dataset design, network architecture | Wiring diagrams, terminal schedules, cable schedules |
| Procurement | Ethernet switches (managed, industrial), fiber/copper network cabling, PRP redundancy hardware | Control cable (often hundreds of conductors), terminal blocks, interposing relays |
| Configuration | SCL file programming per relay, GOOSE publish/subscribe mapping, network switch configuration | None — wiring IS the configuration |
| Testing | Factory Acceptance Test (cross-vendor interop), network commissioning, GOOSE signal verification | Point-to-point continuity testing, relay-to-relay signal injection |
Lifecycle Cost (10-Year Projection)
| Cost Driver | GOOSE | Hardwired |
|---|---|---|
| Adding a bay | Configure new GOOSE datasets, add network drop — often under 1 week | Pull new cables to every existing bay that needs to interlock, re-terminate — typically 2 to 4 weeks |
| Protection scheme change | Update SCL configuration, re-test affected GOOSE subscriptions | Re-wire affected circuits during outage, update drawings, re-test |
| Troubleshooting | Network analyzer captures, GOOSE subscription status checks — remote-capable | Walk-down with multimeter, trace wiring, check terminal tightness |
| Documentation maintenance | SCL files are the documentation (machine-readable) | Wiring diagrams must be manually updated — and often aren’t |
The Crossover Point
For a rough estimate, compare: GOOSE engineering and FAT cost (fixed) vs hardwired cable and labor cost (scales with interlock count). Below approximately 30 interlock points, hardwired materials and labor cost less than GOOSE engineering. Above approximately 50 interlock points, GOOSE engineering cost is dwarfed by hardwired cable runs. Between 30 and 50, the comparison is sensitive to local labor rates and vendor pricing.
Risk Comparison
GOOSE Risks
| Risk | Severity | Likelihood | Mitigation |
|---|---|---|---|
| Network failure | High — loss of all GOOSE signals | Low with PRP/HSR redundancy | Dual-network PRP architecture eliminates single point of failure |
| Cross-vendor interop failure | High — IEDs don’t exchange signals correctly | Medium without proper FAT | Require cross-vendor FAT in the sub’s scope; reject “we’ll fix it on site” |
| SCL configuration error | Medium — wrong dataset mapping | Medium | SCL validation tools, systematic FAT procedure, peer review of SCL files |
| Network congestion | Medium — delayed GOOSE messages | Low in properly designed networks | Dedicate GOOSE to its own VLAN, use priority tagging (IEEE 802.1Q) |
| Cybersecurity exposure | Medium — Ethernet is an attack surface | Low in air-gapped networks | NERC CIP compliance, network segmentation, no routable connections to GOOSE network |
| Environmental sensitivity | Medium — network hardware has tighter temperature, humidity, and corrosion ratings than relays | Medium in non-climate-controlled, coastal, or petrochemical environments | Industrial-grade switches with conformal coating, sealed NEMA 4X enclosures, HVAC — all add cost that hardwired avoids |
Hardwired Risks
| Risk | Severity | Likelihood | Mitigation |
|---|---|---|---|
| Wiring error | High — wrong interlock wired to wrong device | Medium at large scale | Point-to-point testing during commissioning, clear wire labeling |
| Loose terminal connection | High — intermittent interlock failure | Medium over time (thermal cycling, vibration) | Torque-verified terminations, periodic re-torquing, infrared scans |
| Drawing accuracy | Medium — as-built doesn’t match wiring | High after modifications | Discipline in as-built documentation, red-line drawings |
| Scalability failure | High — physically cannot accommodate growth | High if station expands beyond original design | Design cable trays and marshaling cabinets with expansion capacity |
| Troubleshooting time | Medium — extended outage while tracing faults | Medium at large scale | Organized cable routing, consistent labeling, spare conductors |
Net Risk Assessment
For small stations, hardwired has lower net risk because it’s simpler and the failure modes are understood by every protection engineer. For large stations, GOOSE with PRP has lower net risk because the network is self-monitoring and the wiring complexity of hardwired becomes its own risk vector.
Common Misconceptions
”GOOSE is unreliable because it depends on the network”
Reality: With PRP (Parallel Redundancy Protocol), GOOSE achieves zero-switchover redundancy — every frame is sent on two independent networks simultaneously. The receiving IED uses whichever copy arrives first and discards the duplicate. There is no switchover delay, no detection time, no failover mechanism to fail. A properly designed PRP+GOOSE network is more reliable than hardwired, which has no built-in redundancy unless you run duplicate cable paths (rarely done due to cost).
”Hardwired is always simpler”
Reality: Hardwired is simpler up to approximately 30 interlock points. Beyond that, the wiring matrix grows combinatorially. A 20-bay station with full bay-to-bay interlocking can require thousands of individual connections. The wiring diagrams become difficult to read, the marshaling cabinets become congested, and finding a wiring error during commissioning becomes a multi-day exercise. At this scale, a well-configured GOOSE scheme with a clear SCL file structure is actually simpler to understand, test, and maintain.
”GOOSE is too new and unproven”
Reality: IEC 61850 was first published in 2003. GOOSE messaging has been in production substations since 2004 — over 20 years. Thousands of substations worldwide use GOOSE for protection signaling. Major relay vendors (SEL, ABB, Siemens, GE) have supported GOOSE for over a decade. The question is not whether GOOSE works, but whether your project team has the expertise to implement it correctly.
”You can just add GOOSE later”
Reality: Retrofitting GOOSE into a hardwired station is significantly more expensive than designing it in from the start. You need to add network infrastructure (switches, fiber), update relay firmware, develop SCL files, and re-commission the affected protection schemes — all while the station is in service. If GOOSE is likely to be needed within 10 years, building the network infrastructure now (even if some signals start as hardwired) is far cheaper than a retrofit.
”Hardwired means no software”
Reality: Modern protective relays are software-defined regardless of how their external signals are connected. The relay logic, settings, and coordination are all programmed. What hardwired eliminates is the network communication layer and the SCL configuration layer. It does not eliminate software complexity from the protection scheme.
Red Flags in Proposals
When a sub proposes GOOSE
“We’ll handle interoperability on site.” Cross-vendor GOOSE interoperability must be proven in a Factory Acceptance Test before equipment ships. On-site discovery of interop issues causes weeks of schedule delay. If the sub doesn’t include a cross-vendor FAT in their scope, either require one or reject GOOSE.
No network redundancy (PRP or HSR) in the design. GOOSE without network redundancy is a single point of failure for your entire protection scheme. A single switch failure or fiber break would take down all interlocking and tripping. If the sub is cutting PRP/HSR to save cost, they’re creating a network that’s less reliable than the hardwired scheme it replaces.
“All relays are IEC 61850 compliant.” IEC 61850 compliance is not binary. Relays can be “compliant” but implement different editions, different GOOSE versions, or different feature subsets. Ask which edition each relay implements (Ed 1 vs Ed 2 matters), whether GOOSE is supported natively or through a gateway, and whether the vendor has published GOOSE interoperability test results with the other vendors in the project.
Single Ethernet network with a “redundant” ring topology. RSTP ring recovery takes 1 to 50ms depending on topology. During recovery, GOOSE messages are lost. PRP eliminates this entirely with parallel paths. If the sub proposes RSTP instead of PRP, they should justify why the recovery gap is acceptable for your protection scheme timing requirements.
When a sub proposes hardwired
Hardwired for 100+ interlock points. This signals either unfamiliarity with GOOSE or inflating the wiring labor estimate. At this scale, the cable plant alone becomes a project risk. Ask why they chose hardwired and what GOOSE alternative they evaluated.
No expansion capacity in cable trays and marshaling. If the station may expand, hardwired designs must include spare capacity. A design with no spare conductors, full cable trays, and no empty terminal blocks is a hidden cost that will surface at the first expansion.
“Wiring diagrams will be provided as-built.” As-built documentation for complex hardwired schemes is notoriously inaccurate. Require the sub to provide red-lined as-built drawings before substantial completion, and verify a sample during commissioning.
Technical Deep Dive
This section is for engineers reviewing the PM’s decision or validating a sub’s technical approach.
Signal Propagation Comparison
GOOSE signal path: Relay logic output, then GOOSE encoding, then Ethernet frame, then switch forwarding, then receiving relay GOOSE decoding, then relay logic input. Total time: typically 2 to 4ms end-to-end on a properly configured network.
Hardwired signal path: Relay logic output, then output contact closure, then cable propagation, then input contact wetting, then input debounce, then relay logic input. Total time: typically 5 to 15ms depending on relay contact speed and input filtering.
GOOSE is faster because Ethernet multicast is faster than electromechanical contact closure. The specific timing depends on the relay models and network configuration, but GOOSE consistently beats hardwired on raw signaling speed.
GOOSE Frame Reliability
GOOSE uses a retransmission scheme to handle packet loss without acknowledgment overhead. When a GOOSE dataset value changes:
- The publishing IED sends the new state immediately
- It retransmits at accelerating intervals: 2ms, 4ms, 8ms, 16ms, … up to a configured maximum (typically 1000ms)
- Receiving IEDs use the first valid frame and discard duplicates
- If no frames arrive within the configured timeout (stale time), the subscriber raises an alarm
This means a single dropped packet does not cause a missed interlock — the retransmission covers it. With PRP, the frame is sent on two independent networks simultaneously, making packet loss even less likely.
Hardwired Failure Modes at Scale
Hardwired reliability degrades with scale due to mechanical failure modes:
- Terminal loosening: Thermal cycling causes copper to expand and contract, gradually loosening screw terminals. In a station with 2,000+ terminals, some will loosen within 5 years.
- Cable damage: Dense cable trays with hundreds of conductors make it easy to damage adjacent cables when pulling new ones. Insulation nicks create intermittent grounds that are extremely difficult to locate.
- Drawing drift: Every modification that isn’t meticulously documented creates a gap between the drawings and the physical wiring. After 2 to 3 modifications without disciplined as-builting, the drawings become unreliable.
IEC 61850 Dataset Design for Interlocking
A well-designed GOOSE interlocking scheme organizes datasets by function, not by bay. Each dataset contains related signals that change together — for example, all breaker status signals from one bay, or all bus differential trip outputs. This minimizes unnecessary GOOSE retransmissions and keeps the SCL file structure manageable.
Key design rules:
- One GOOSE Control Block per logical function group (don’t mix unrelated signals)
- GOOSE dataset size under 64 data attributes (larger datasets increase encoding time)
- Separate fast-trip signals (under 4ms requirement) from status/alarm signals (under 100ms is fine)
- Use IEC 61850 logical node naming conventions so that the signal purpose is readable from the dataset without consulting external documentation
Need Help Choosing the Right Protocol Architecture?
We've implemented both sides of this comparison in production.
Let's figure out what fits your project.