Multi-Vendor Relay Integration: Where Protection Fails at the Seams
Technical Brief · March 28, 2026
The Coordination Gap
A typical Tier III or Tier IV data center power system doesn’t run on a single equipment platform. The switchgear comes from one manufacturer. The generator controls come from another. The protective relays are a third. The power metering, ATS controllers, and network infrastructure each add their own vendor to the mix.
On paper, this is a standard multi-vendor procurement. In practice, it creates a coordination gap that no single equipment vendor is responsible for closing.
The protection engineer from the relay manufacturer configures protection settings for their devices. The generator controls engineer programs start/stop sequencing for theirs. The SCADA integrator connects to both — but nobody owns the seams. The interlock between the generator controller’s sync-check permissive and the relay’s breaker-close logic. The handshake between the ATS controller’s transfer sequence and the upstream feeder relay’s protection scheme. The protocol conversion between one vendor’s Modbus RTU and another’s IEC 61850 GOOSE.
These integration boundaries are where protection systems fail during commissioning.
What Eight Platforms Look Like in Practice
On complex data center programs, the vendor count adds up quickly. A 2(N+1) medium-voltage power system with full protection, automation, and monitoring can involve eight or more manufacturer platforms. For how this vendor coordination extends during facility expansion, the challenge compounds:
| Manufacturer | Typical Role | Devices |
|---|---|---|
| SEL | Protection relays, automation controllers | SEL-751, SEL-700G, RTAC-3530, SEL-2440 |
| Woodward | Generator control, synchronization, ATS | easYgen 3400, LS-6, DTSC-200 |
| CAT | Engine/generator management | EMCP 4.2, EMCP 3.3 |
| Electro Industries (EIG) | Power quality metering | Shark-250 |
| Eaton | Power metering | Power Xpert |
| Cisco | Network infrastructure | Catalyst 9500, IE-2000 |
| ABB | Medium-voltage switchgear | 2.4 kV class |
| UPS OEM | Uninterruptible power supply | Site-specific modular UPS |
Each platform has its own configuration tools, communication protocols, and engineering interfaces. SEL relays are programmed in AcSELerator QuickSet. Woodward controllers use RP-2000. CAT engine management uses a separate configuration interface. The network switches have their own CLI.
The engineering challenge isn’t configuring any one of these platforms — it’s making all eight work together as a coordinated protection and controls system.
Where Integration Breaks Down
Integration failures follow predictable patterns. Three scenarios account for most commissioning surprises on multi-vendor programs.
Protocol Mismatch
Device A publishes status via IEC 61850 GOOSE. Device B only speaks Modbus TCP. Without a protocol gateway (typically an RTAC or similar automation controller), these devices can’t exchange the interlock signals that protection coordination requires.
The failure mode isn’t dramatic — devices simply don’t communicate. The generator controller waits for a sync-check permissive that never arrives because the relay publishes it on a protocol the controller doesn’t subscribe to.
Timing Mismatch
Protection coordination depends on timing. A zone-selective interlocking (ZSI) scheme requires downstream relays to send restraint signals to upstream relays within a specific time window — typically 50-100 milliseconds including detection, messaging, and trip decision.
When devices from different vendors participate in the same protection scheme, timing assumptions that hold within a single vendor’s ecosystem may not hold across vendors. A relay with a 4.17ms processing interval at 60 Hz and a controller with a 4-7ms logic cycle create a compound delay that must be explicitly validated, not assumed.
Configuration Drift
Each vendor’s configuration tools produce independent settings files. SEL relay settings are stored in QuickSet files. Woodward parameters live in RP-2000 projects. SCADA tags are maintained in a separate database.
When a protection coordination study revises settings — new pickup values, revised time-current curves, changed interlock logic — every affected platform must be updated consistently. A change to a breaker-close permissive in the relay settings that isn’t reflected in the generator controller’s sync-check logic creates a latent failure that may not surface until witness testing.
Multi-Vendor Coordination Challenge?
One Team Across Every Platform
We program SEL, Woodward, CAT, and integrate across eight manufacturer platforms within a single coordinated P&C scope.
What Coordinated Integration Looks Like
Closing the multi-vendor coordination gap requires a single engineering scope that spans all platforms. Three practices make the difference.
Unified GOOSE Publish/Subscribe Matrix
A GOOSE matrix documents every signal exchange between every device — which IED publishes what data, which IEDs subscribe to it, and what action the subscriber takes. This matrix crosses vendor boundaries.
When a SEL-751 feeder relay publishes a fault status that a Woodward LS-6 synchronizer subscribes to for sync-check blocking, that relationship is documented in one matrix, not split across two vendor scopes. The matrix becomes the single source of truth for protection coordination across all platforms.
Cross-Platform Settings Management
Protection settings for all device types are developed from the same coordination study and managed under the same configuration control process. When an arc-flash study changes the pickup settings on a feeder relay, the downstream effects on generator controller permissives, ATS transfer logic, and load shedding priorities are traced and updated in the same engineering cycle.
This requires proficiency across multiple configuration tools — AcSELerator QuickSet for SEL, RP-2000 for Woodward, vendor-specific tools for each platform — and a settings management process that tracks dependencies across all of them.
End-to-End Commissioning Ownership
The integration boundaries between vendors are where commissioning test procedures must be most rigorous. The handshake between a Woodward easYgen 3400 generator controller and an SEL RTAC-3530 automation controller — start permissives, sync-check signals, load-share references — needs to be tested as a unit, not as two separate vendor scopes that are assumed to work together.
A structured commissioning framework that progressively builds from component testing (L3) through system testing (L4) to integrated testing (L5) catches these cross-vendor integration issues at each level — before they surface during owner witness testing.
The Schedule Impact
For prime contractors, multi-vendor integration risk translates directly to schedule risk.
When each vendor’s scope is tested independently and integration is left to the commissioning phase, the program carries a latent risk: the first time all eight platforms operate together as a coordinated system may be during witness testing. Any integration failure at that point triggers retesting — and retesting at L5 means re-mobilizing the commissioning agent and absorbing the delay.
A single P&C engineering scope across all platforms compresses this risk forward. Integration is tested during L4 functional system testing, not discovered during L5 integrated system testing. By the time the owner’s commissioning agent witnesses the test, cross-vendor coordination has already been validated.
What This Means for Subcontractor Coordination
The traditional approach — separate subcontractors for relay programming, generator controls, SCADA integration, and network infrastructure — creates the coordination gap by design. Each sub owns their vendor scope. Nobody owns the seams.
Consolidating protection and controls engineering into a single subcontractor scope eliminates the coordination gap at the contracting level. One team programs the relays, configures the generator controllers, integrates the SCADA, and commissions the complete system. The GOOSE matrix, settings files, and test procedures all come from the same engineering process.
For prime contractors managing Tier III/IV data center programs, this reduces the number of P&C subcontractors to coordinate from three or four to one — and makes that one team accountable for the integration that matters most.
See how this approach works within our full protection & controls services scope and technical methodology.
Put This Engineering Depth Behind Your Next Program
Tell us about your data center protection and controls requirements — we'll scope the work and show you how we'd approach it.