Protection Relay I/O Matrices: From Functional Requirement to Terminal Plan

A controlled workflow from cause-and-effect and signal dictionary to channel allocation, terminal plans, SCL data and end-to-end verification.

A protection relay I/O list is not a terminal-number spreadsheet; it is the controlled translation of every functional requirement into an electrical or digital interface with a defined normal state, failure response, timing, ownership and test. If that chain is weak, a perfectly configured relay can be wired to the wrong source, share a hidden common, exceed contact duty or report an unsafe state as healthy.

This article gives a practical workflow from protection philosophy and cause-and-effect logic to signal dictionary, I/O matrix, relay configuration, terminal plan, IEC 61850 engineering and FAT/SAT evidence.

Executive rules

  • Define the function and safe failure state before selecting a relay input or output.
  • Give every signal one stable ID and one authoritative definition reused across all documents.
  • Record source, destination, state semantics, voltage/current, contact form, timing, isolation, supervision and test method.
  • Design I/O commons and auxiliary supplies as failure domains; do not discover them after relay hardware is ordered.
  • Check output contacts at actual DC inductive make/carry/break duty, not only a generic ampere rating.
  • Check binary-input pickup/dropout, wetting current, leakage and cable capacitance over the complete DC voltage envelope.
  • Treat hardwired and IEC 61850 GOOSE signals as engineered interfaces with equivalent traceability and failure handling.
  • Generate and cross-check terminal plans, schematics, I/O mapping, settings and SCL data from controlled records where possible.

1. Standards and information architecture

ReferenceUse
IEC 61082-1:2014General presentation rules for electrotechnical diagrams, drawings and tables
IEC 81346-1:2022System structuring and unambiguous object reference designations
IEC 81355-1:2024Classification/designation of information and documents
IEC 61850-6:2009+A1:2018+A2:2024SCL exchange of IED capability, communication configuration, substation functions and relationships
IEC 60255-27:2023Product-safety framework for protection equipment and associated auxiliaries
Relay/interface manufacturerExact terminal groups, ratings, isolation, pickup/dropout, timing, contact duty and configuration constraints

These references support documentation and products; they do not decide the project’s trip philosophy, redundancy, DC system or fail-safe state. Those must be defined by the protection and control design authority.

2. The seven-layer traceability chain

  1. Requirement: what hazard/operation must be detected or commanded?
  2. Function: protection, interlock, control, alarm, measurement or supervision logic.
  3. Signal: precise state/event/data crossing an interface.
  4. Implementation: hardwired contact/analogue circuit or IEC 61850 data/GOOSE/report.
  5. Hardware/configuration: relay card/channel, logical node/data object, output relay or gateway.
  6. Physical/communication path: terminals/cable/DC supply or publisher/network/subscriber.
  7. Verification: design review, point-to-point check, functional/negative test and evidence.

Each matrix row should navigate both directions. A terminal must lead back to a requirement; a requirement must lead forward to its test. Orphan signals and unimplemented requirements then become visible.

3. Build a controlled signal dictionary first

FieldExample / question answered
Signal IDStable, unique project identifier
Functional name“Breaker 52a — closed indication”, not “BI7”
Source objectBreaker auxiliary switch, relay function, DC MCB
Destination/useInterlock, SOE, SCADA, protection algorithm
State semanticsContact closed means breaker closed; state at de-energised device
Failure stateBroken wire/DC loss/network loss response
Electrical profileVoltage, wetting current, contact duty, polarity, isolation
Time profilePulse/maintained, debounce, SOE resolution, maximum transfer
Quality/securitySupervised, redundant, test-blocked, authenticity expectations
ReferencesRequirement, schematic, terminal, cable, setting and test case IDs

Avoid ambiguous names such as “CB status.” Use separate open and closed indications, define 00/01/10/11 handling and state whether the contact symbol represents the device de-energised, breaker open, or drawing normal-service condition.

4. Convert cause-and-effect logic into interfaces

  • For every protection operation, list local trips, remote trips, lockout, breaker-failure start, autoreclose block, disturbance trigger and alarms.
  • For every close request, list permissives, interlocks, synchronism/dead-bus condition, anti-pumping and inhibit causes.
  • For every equipment state, define position contacts, validity logic, mismatch timer and operator indication.
  • For every supervision function, define monitored path, alarm delay, latching, reset and operational response.
  • For every external interface, define ownership and the authoritative source of truth.
  • Add negative requirements: “must not close when earth switch closed” is as important as the valid close case.

5. Binary-input engineering

  • Nominal/minimum/maximum DC input voltage and polarity.
  • Guaranteed pickup/dropout thresholds and hysteresis across temperature/tolerance.
  • Input current/wetting current and whether field contacts need a minimum cleansing current.
  • Shared common terminal groups and channel-to-channel/earth isolation.
  • Maximum allowable field leakage from lamps, surge suppressors, TCS resistors or other inputs.
  • Cable capacitance/induced voltage on long runs and possible delayed dropout/false pickup.
  • Hardware/software debounce versus required event timing and contact bounce.
  • SOE timestamp location/resolution and synchronisation quality.
  • Test voltage/current and terminal access without energising adjacent channels.

Calculate the complete loop at minimum DC voltage and maximum resistance. At maximum DC, verify input power and leakage paths. A high-impedance input may remain picked up through an LED, suppression network or capacitive coupling after the field contact opens; a properly engineered bleed/interposing solution may be required.

6. Binary-output engineering

Output dutyChecks
Trip/close coilDC inductive making/carrying/breaking, pulse duration, coil current at voltage extremes, 52a/52b interruption
Interposing/lockout relayCoil inrush/hold, suppression, contact release time and failure mode
Alarm/SCADAWetting current, contact minimum load, common supply and pulse capture
High-speed outputSemiconductor leakage, withstand, polarity, fail state and isolation
Latched outputReset authority, loss-of-power state, persistence and mechanical/electrical indication

Do not select an output from “8 A” alone. Obtain the manufacturer’s DC inductive duty or L/R rating at the project voltage. Where the breaker auxiliary contact interrupts coil current, show that explicitly. Suppression protects contacts but can slow current decay and device release; verify the protection/closing timing consequence. Use an interposing relay when needed for duty, isolation or contact multiplicity, then include its own DC drop, delay, supervision and failure modes.

7. I/O commons are failure domains

Relay cards often share a common terminal, internal supply, fuse or isolation barrier among several channels. One loose common can remove apparently independent trips or indications. Before assigning channel numbers:

  • map every common and card supply from the hardware manual;
  • keep redundant protection channels on independent cards, DC feeders and terminal groups where required;
  • separate trip and noncritical indication circuits so a field short does not disable tripping;
  • respect isolation limits between different DC systems and external owners;
  • include spare-channel commons and future additions in fault analysis;
  • show internal commons on the schematic even if the relay symbol library hides them.

8. Hardwired versus IEC 61850 GOOSE

AttributeHardwiredGOOSE
IdentityTerminal/wire/cable/contactIED, logical node/data object, dataset, control block, VLAN/network path
Failure detectionTCS, line monitoring, dual contact plausibilityQuality, state/sequence, supervision, publisher/network/receiver alarms
TimingContact and input/output delaysApplication transfer plus IED/network processing
IsolationGalvanic/test switchTest/simulation flags, configuration and network controls
Common modeDC supply, cable route, common terminalSwitch, clock, configuration, multicast/VLAN, cyber controls
EvidenceContinuity/injection/functional testSCL comparison, packet/application monitoring and end-to-end test

The functional matrix should remain technology-neutral until performance, dependability, testability and independence are agreed. For IEC 61850, record the semantic data reference, quality handling, test/simulation behaviour, dataset/control-block identity, subscriber logic and fallback state. IEC 61850-6 SCL is a configuration exchange format; the approved SCD/CID/IID files and tool reports must stay consistent with the project signal dictionary.

9. From matrix to relay channel allocation

  1. Freeze mandatory functions and redundancy classes.
  2. Group signals by voltage, isolation, speed, duty, location and failure domain.
  3. Reserve dedicated trip contacts and independent breaker-failure/retrip paths as required.
  4. Allocate hardware channels using the exact card/terminal/common topology.
  5. Assign logical inversion only after contact state and fail-safe philosophy are approved.
  6. Reserve realistic spares per card/type, not only a percentage of total points.
  7. Check panel heat, auxiliary burden, terminals, cable cores and future space.
  8. Export/configure channel mappings and independently reconcile them against the matrix.

10. Terminal-plan generation

A terminal plan must preserve the interface semantics, not just number terminals sequentially. For every connection show terminal strip/number, device pin, from/to object, wire number, cable/core, voltage/system, conductor size/type, bridge/disconnect/test function, screen/earth and spare status.

  • Place field interfaces at a clear ownership boundary.
  • Use disconnect/test terminals for CT/VT/trip circuits according to their safety function.
  • Segregate different voltages, DC systems, analogue quantities and communications.
  • Avoid more conductors per terminal than certified; use distribution terminals/bridges.
  • Keep redundant channels physically separated when the reliability requirement demands it.
  • Control cross-panel links from both panels with one master cable/core allocation.
  • Make terminal-side orientation and normal link positions explicit.

11. Example matrix row

Signal IDFEED01_CB52A
FunctionFeeder breaker closed indication / open-close plausibility
SourceBreaker auxiliary switch 52a, closes when main contacts reach closed position
DestinationProtection IED binary input BI-03; SCADA derived via IED
Electrical110 Vdc wetted contact, input common group C1, project DC-1
Normal/failureBreaker open: contact open; broken wire is not uniquely distinguishable without 52b/mismatch logic
TimingSOE at IED; debounce 5 ms illustrative—validate against mechanism bounce and event need
PhysicalBreaker plug → cable core → panel terminal → IED X1:03
TestOperate breaker all positions; force 00/11 disagreement with 52b; remove DC/common

Values above are illustrative. The important point is that the same row captures state meaning, hardware constraint and negative tests.

12. Change and configuration control

  • Nominate one controlled signal database/matrix as master and define generated versus manually maintained documents.
  • Use stable IDs; never recycle a deleted ID for a different meaning.
  • Baseline relay firmware/hardware, settings, logic, SCD/CID, drawings and test cases together.
  • Run automated checks for duplicate channels/terminals, missing destinations, mismatched inversion and orphan SCL subscriptions.
  • Require impact assessment for common-terminal, DC supply, output-duty and cyber/network changes.
  • Red-line site deviations immediately and close them into as-built data before energisation.
  • Record who approved each semantic or fail-state change—not just the spreadsheet revision.

13. FAT/SAT verification matrix

  1. Static compare every I/O row to schematic, terminal, cable, hardware bill and relay configuration.
  2. Inspect actual card type, common groups, terminals, bridges, labels and segregation.
  3. Point-to-point test each hardwired input from field contact to IED state.
  4. Operate each output into representative load or approved simulator; capture voltage/current/timing for critical duties.
  5. Test contact inversion, pulse/latched behaviour, debounce and SOE timestamp.
  6. Remove each DC feeder/common and open/short field wiring to verify fail-state alarms.
  7. For GOOSE, verify publisher/subscriber references, quality, test/simulation, network loss and recovery end-to-end.
  8. Execute cause-and-effect valid and forbidden scenarios, including simultaneous operations.
  9. Check SCADA naming, quality, timestamps, alarm class and command select/execute where applicable.
  10. Archive test evidence against signal IDs and reconcile every defect/change into as-built baselines.

14. Frequent mistakes

MistakeConsequenceCorrection
Matrix starts with BI/BO numbersHardware drives functionStart with requirements/signals
“NO contact” without state definitionInversion/fail-state ambiguityDefine device state and energisation convention
Output selected by AC/general currentContact weld/failure on DC coilCheck actual DC inductive duty
Shared common hiddenOne fault defeats redundant channelsMap card/terminal failure domains
GOOSE only documented in SCDFunction/test traceability lostInclude semantic matrix and end-to-end test
Spare percentage onlyNo suitable isolated channel remainsReserve by card/type/common/location
FAT checks IED screen onlyField path/failure response unprovenTest source-to-destination and negative cases

15. Design-release checklist

  • Every requirement has implemented signals and tests?
  • Signal IDs/names/state semantics unique and stable?
  • Source/destination/owner and normal/failure states defined?
  • Input voltage, pickup/dropout, leakage and commons checked?
  • Output DC inductive duty, suppression and timing checked?
  • Redundant channels free of prohibited common failures?
  • Hardwire/GOOSE decision supported by performance and testability?
  • Terminal/cable/core allocation physically buildable and segregated?
  • Relay configuration and SCL data reconciled to matrix?
  • Valid, forbidden, loss-of-DC/network and restoration tests included?
  • As-built/configuration baseline and change authority defined?

References and further reading

Engineering note: Channel numbers, delays and ratings in examples are illustrative. Use the selected relay/interface manuals and approved project protection/control philosophy.

LearnSwitchgear

Search the engineering library