THE AGL ENGINEERING CENTRE by Precision AGL Academy · Electrical Reliability

Redundancy in AGL Electrical Architecture Dual Circuits, Backup Power & Fault-Tolerant Airport Lighting

See how dual runway circuits, independent feeders, CCRs, UPS and DG backup, ALCMS and fail-safe control combine so that one failure does not take away the visual guidance a runway needs, and why two of everything is not the same as redundancy.

Overview

What Is Redundancy in AGL Electrical Architecture?

Redundancy in Airfield Ground Lighting (AGL) means designing the electrical and control system so that a single equipment or circuit failure does not unnecessarily remove the required airport visual guidance. It matters most for precision approach runways, CAT II/III operations, low-visibility procedures, runway centreline, edge and touchdown-zone lighting, stop bars, critical approach lighting and essential taxiway lighting.

A redundant AGL system is not simply a system with “two of everything.” The engineering objective is to maintain the required lighting service after a defined single failure, with the required continuity and without creating misleading or conflicting indications.

Redundancy has to be considered across the whole chain

Power source
Switchgear
UPS / DG
CCR
AGL circuit
Transformer
Connector
Light
Control & monitoring

Single system

One failure leads to loss of service.

Redundant system

One failure leaves an alternate path that continues service.

Different failures need different strategies

Utility failureCCR failureAGL circuit failureTransformer failureLight failureControl-system failureUPS failureGenerator failureCable insulation failure
ICAO guidance addresses secondary power and automatic restoration of required services after failure of the normal source. Airport-specific arrangements, including those published by AAI, differ from airport to airport, so treat them as examples, not generic rules.
Interactive

Interleaved Circuits: What Survives When a Circuit Fails?

Interleaving puts neighbouring lights on different circuits, so one circuit failure leaves a distributed pattern instead of a long dark section. Try the four cases.

Circuit A Circuit B Unlit

A simplified picture. Real circuit arrangements, light counts and the serviceability limits that apply after a circuit failure are project-specific and set by the approved design and operational procedures.

How two interleaved series circuits share one row of lights.

ONE ROW OF LIGHTS · TWO INTERLEAVED CIRCUITS 1 · BOTH CIRCUITS HEALTHY CCR A CCR B 2 · CIRCUIT A LOST GAP = 2 × PITCH every second light lit 3 · FOR COMPARISON: ONE CIRCUIT, NOT INTERLEAVED, LOST WHOLE ROW DARK WHAT TO CHECK 1 ALTERNATE LIGHTS odd on circuit A, even on B 2 TWO REGULATORS each carries about half the load 3 SWITCH TOGETHER same brightness step on both 4 MONITOR BOTH one circuit lost is an alarm SCHEMATIC · TWELVE LIGHTS SHOWN · NOT A CIRCUIT DESIGN
Figure 1. Alternate lights sit on alternate circuits, so losing one circuit thins the pattern instead of blanking it.
AGL Visual Encyclopedia: Interleaved Series Circuits Alternate-Circuit Centreline Arrangement AGL Circuit Zones
Layers

The Layers of AGL Redundancy

Source redundancy

More than one power source.

Distribution redundancy

More than one independent feeder or bus.

CCR redundancy

Multiple regulators serving independent or appropriately distributed circuits.

Circuit redundancy

Two or more AGL circuits serve a lighting system.

Spatial redundancy

Lights are distributed between circuits instead of grouping adjacent lights on one circuit.

Equipment redundancy

Critical control or monitoring equipment has standby units.

Control redundancy

Failure of a processor or network component does not necessarily mean total loss of control.

Monitoring redundancy

Faults stay detectable even when part of the monitoring architecture fails.

A five-layer way to explain it

1. Power: dual utility / UPS / DG
2. CCR: multiple regulators
3. Field circuit: dual / interleaved
4. Equipment: individual transformers and lights
5. Control: ALCMS, network, field control
Each layer addresses a different failure mechanism. Strength at one layer does not compensate for a single point of failure at another.

Redundancy is a stack of independent layers, not one duplicated box.

SIX LAYERS · EACH ONE ANSWERS: WHAT KEEPS THE LIGHTS ON IF THIS LAYER FAILS?1 SUPPLYTwo feeders, UPS and DGUtility loss or one feeder fault does not leave the vault dark2 POWER EQUIPMENTStandby CCR, changeover, bypassOne CCR failure is restored by changeover, not by a repair3 CIRCUITInterleaved or alternate circuitsOne circuit lost: the pattern thins but stays usable4 FIELDCable routes, transformers, surge protectionSeparate routes and protection limit one fault to one place5 CONTROLALCMS, dual network, field controllersControl path loss falls back to local or standby control6 OPERATIONSSpares, maintenance, procedures, serviceabilityThe people and process layer that keeps layers 1 to 5 intactSCHEMATIC · LAYERS ARE INDEPENDENT BY DESIGN · EXACT ARRANGEMENT DEPENDS ON THE APPROVED DESIGN
Figure 2. Each layer has its own measure. A weak layer cannot be hidden by a strong one above it.
AGL Visual Encyclopedia: Single Point of Failure Analysis Serviceability of Critical Lights
Principles

Electrical vs Operational Redundancy, Single Points of Failure and Common Cause

Electrical redundancy

An alternate electrical path exists.

Operational redundancy

The airport can continue the required operation after the defined failure. Two CCRs do not guarantee this: the alternate CCR may be connected to the wrong part of the circuit, or a common cable section may still be a single point of failure.

The right design question is: what lighting function remains available after the defined failure?

Potential AGL single points of failure

One incoming feederOne main switchboardOne transformerOne CCROne common circuit sectionOne field controllerOne communication switchOne ALCMS serverOne UPSOne generatorOne common cable route

Common-cause failures that defeat apparent redundancy

ExampleWhy it defeats redundancy
One flood affects both circuitsBoth circuits share the same exposure
One chamber fire damages both cablesPhysical proximity
One switchboard feeds both CCRsShared electrical source
One network switch controls both systemsShared control path
One software server controls everythingShared logic
One grounding failure affects both circuitsShared reference and protection path
Redundancy needs independence analysis, not just duplication: physical, electrical and logical independence.

Redundant vs fail-safe

Redundant means a backup exists. Fail-safe means that when something fails, the system moves to or stays in a state that minimizes safety risk. Two controllers sending conflicting commands are redundant but not necessarily safe.

Redundancy vs resilience

Redundancy is an alternate path or capacity. Resilience is the airport’s wider ability to withstand and recover from disruption: redundancy plus protection, monitoring, maintenance, procedures, spares and recovery.

Ask of every link in the chain: what if it fails alone?

SUPPLY TO LIGHT · ASK OF EVERY LINK: WHAT IF IT FAILS ALONE? 1 · NO REDUNDANCY 4 SINGLE POINTS FEEDER ALL DARK single supply BUS ALL DARK single bus CCR ALL DARK no spare CABLE ALL DARK single circuit A LIGHT ONE LIGHT not a single point 2 · REDUNDANCY ADDED, CABLE SINGLE 1 SINGLE POINT LEFT: CABLE 2 FEEDERS NO EFFECT two supplies BUS + TIE NO EFFECT two sections CCR + SPARE NO EFFECT changeover CABLE ALL DARK still single A LIGHT ONE LIGHT not a single point 3 · LOOKS REDUNDANT, FAILS TOGETHER 1 HIDDEN: SHARED TRENCH 2 FEEDERS ALL DARK one shared trench BUS + TIE NO EFFECT two sections CCR + SPARE NO EFFECT changeover 2 CIRCUITS HALF LIT interleaved A LIGHT ONE LIGHT not a single point WHAT TO CHECK 1 TRACE THE CHAIN supply to light, link by link 2 ASK WHAT IF what goes dark, and for how long 3 FIND COMMON CAUSE shared trench, control supply, UPS 4 REMOVE, REDUCE, ACCEPT and write down which, and why SCHEMATIC · ONE CHAIN SHOWN · NOT A RISK ASSESSMENT
Figure 3. Single points of failure show up link by link. Two feeders in one trench are a hidden single point.
AGL Visual Encyclopedia: Single Point of Failure Analysis Dual Feeders and Incoming Supply Redundancy
Architecture

A Redundant AGL Electrical Architecture

Representative layered arrangement. The actual topology depends on runway category, AGL configuration, operational requirements and the approved design.

Feeder A Feeder B Main switchgear and ATS / changeover UPS (fast continuity) DG (sustained backup) Critical AGL bus CCR-A CCR-B Circuit A (lights 1, 3, 5, 7) Circuit B (lights 2, 4, 6, 8) Interleaved runway lighting ALCMS / monitoring
The dashed line shows monitoring and control. If both CCRs sat behind one switchboard or both circuits shared one duct, this picture would still contain a single point of failure, which is why independence analysis comes before duplication.
Circuits

Dual-Circuit Redundancy by Lighting System

Lighting systemWhat redundancy is aiming forDesign considerations
Runway edgeContinuity after one circuit failure and distributed light availabilityCircuit independence, physical separation, light distribution, transformer allocation
Runway centrelineService maintained on the second circuit if one fails, especially on precision runwaysCircuit design and remaining light serviceability
Runway endAvoid losing the whole function on one circuitHidden common points of failure, maintenance, failure isolation
ThresholdA single circuit failure does not remove the complete threshold functionDistribution between circuits, checked against visual-aid and operational requirements
Touchdown zoneSupport low-visibility operationsFixture spacing, independent power, CCR allocation, fault detection, access
ApproachMaintain required visual guidance after defined failuresWhich portions need secondary power. It is not necessarily right to duplicate every circuit
TaxiwayLimit a failure to a small areaSeparate segments for edge, centreline, stop-bar and guard-light circuits
Stop barsDefined, controlled state on failureSupply, control, monitoring, switching, interlocking and field lighting together
A stop-bar power supply that permits contradictory commands is not safe redundancy. Stop bars can represent a protected runway holding position, so failure must lead to a defined state.

Independent vs dependent interleaved circuits

Independent interleaved

Each circuit has an electrical path sufficiently separated from the other, as set by the design. Stronger redundancy.

Dependent interleaved

Lights are interleaved physically but the circuits share some infrastructure or dependency. The consequence of losing one circuit differs.

AAI’s eAIP entries show this difference at specific airports: some runways are described with independent interleaved circuits and others with dependent ones, with different consequences when one circuit fails. These are airport-specific examples. Check the current eAIP and AIP supplements for any airport you rely on.

Alternate circuits on a centreline, and what remains after one circuit is lost.

1 · BOTH CIRCUITS ON — FULL SPACING 2 · CIRCUIT B LOST — EVERY SECOND LIGHT STAYS LIT S 2S · GUIDANCE KEPT HOW IT IS WIRED CCR A CCR B CIRCUIT A · LIGHTS 1, 3, 5 … CIRCUIT B · LIGHTS 2, 4, 6 … ISOLATING TRANSFORMER PER LIGHT ISOLATING TRANSFORMER PER LIGHT ONE CIRCUIT LOST · THE LINE STILL GUIDES · THE ALARM TELLS MAINTENANCE
Figure 4. Alternate-circuit arrangement keeps the centreline pattern recognisable after a single circuit fault.
AGL Visual Encyclopedia: Alternate-Circuit Centreline Arrangement Interleaved Series Circuits Taxiway Circuit Arrangement
Power

Source, UPS, DG and Switch-Over

ArrangementWhat it gives
Dual utility feedTwo independent utility feeders
Utility + DGGrid plus standby generator
Utility + UPS + DGUPS gives fast continuity, DG gives sustained backup
Dual DGTwo standby generators
Dual UPSTwo UPS systems serving critical loads
Independent substationsSeparate infrastructure for different critical systems
Two feeders from the same substation may not be as resilient as two genuinely independent sources. Evaluate common-cause risk.

N+1

Capacity for the load plus one extra unit. If three modules are needed, install four. A shared input panel, output panel or battery can still be a single point of failure.

Parallel redundant

Units share load, and one can fail without losing the protected function.

2N

Two complete independent systems, each able to carry the critical load. Stronger and more expensive.

Redundancy is not the same as switch-over time

Redundancy

Can the system continue after a component failure?

Switch-over time

How quickly is service restored after the source fails? Two generators can still have an unacceptable transfer time, while an online UPS gives very fast continuity with a single generator behind it.

Utility fails
UPS: immediate support
DG starts and takes over
Sustained supply
ICAO calls for automatic connection of required secondary-powered facilities after failure of the normal source, with category-dependent maximum switch-over times for visual aids (Annex 14, secondary power provisions). Study the restoration of actual AGL, not just generator start.

Generator sizing

Do not size from average AGL load. Use the critical-load schedule: which CCRs run together, starting currents, UPS charging, other essential loads, generator failure scenarios, expansion and load shedding.

UPS sizing

Reflect actual AGL load, control load, ALCMS, A-SMGCS interfaces, autonomy, battery ageing, growth and the chosen redundancy philosophy, assessed as a complete system.

Airport Lighting Power Calculation

An airport lighting power calculation starts from the critical-load schedule, not the average AGL load. Add up the loads that run together: the CCRs for each lighting circuit, control and ALCMS load, A-SMGCS interfaces and other essential loads. Then allow for CCR starting currents, UPS charging, expansion and load shedding. The same schedule feeds generator sizing, UPS sizing and the N+1 or 2N decision, so the result should be checked against each failure scenario rather than the normal case alone. For CCR and circuit-level loading, see the AGL Toolkit

Spare capacity is not redundancy. A 500 kVA generator carrying 300 kVA has spare capacity; a second generator able to carry the critical load provides redundancy.

Two incoming feeders, a bus and what each fault leaves lit.

TWO FEEDERS · TWO BUS SECTIONS · ONE NORMALLY OPEN TIE 1 · BOTH FEEDERS HEALTHY TIE OPEN · EACH BUS ON ITS OWN FEEDER FEED A FEED B TIE CCR A CCR B 2 · FEEDER A LOST TIE CLOSED · B CARRIES BOTH BUSES FEED ALOST FEED B TIE CCR A CCR B 3 · BOTH FEEDERS LOST GENERATOR CARRIES BOTH BUSES FEED ALOST FEED BLOST TIE CCR A CCR B GEN WHAT TO CHECK 1 INDEPENDENT FEEDERS different source, different route 2 FULL-LOAD RATING each feeder able to carry it all 3 INTERLOCKED TIE no unplanned paralleling of sources 4 SHARED WEAK POINTS control supply, trench, upstream SCHEMATIC · ONE LINE PER FEEDER · NOT A DESIGN
Figure 5. Dual feeders protect against loss of one supply path, provided the routes and sources are genuinely separate.
AGL Visual Encyclopedia: Dual Feeders and Incoming Supply Redundancy Standby power and UPS

Generator start, changeover and the gap the lights must ride through.

CHANGEOVER AND SWITCHOVER TIME · HOW LONG THE LIGHTS ARE INTERRUPTED the switch is the fast part — starting the generator and restarting the regulators take the time 15 s CLASS · GENERATOR ROUTE DETECT 1 GENERATOR START AND RUN-UP longest part of the break 2 TRANSFER 3 CCR RESTART 4 MARGIN keep it — do not spend it 5 LIGHTS INTERRUPTED · THIS IS THE SWITCHOVER TIME t = 0 15 s LIMIT mains fails 1 s CLASS · NEEDS A NO-BREAK ARRANGEMENT 1 s LIMIT A GENERATOR CANNOT START IN 1 s — THE LOAD MUST BE CARRIED BY A SOURCE ALREADY RUNNING second independent feed with fast transfer · no-break power system · as the design and the authority require 1 DETECT the loss or poor quality of the supply is sensed, usually with a short deliberate delay 2 GENERATOR START AND RUN-UP cranks, starts and reaches rated voltage and frequency — usually the longest part 3 TRANSFER the changeover switch moves the load to the generator 4 CCR RESTART the regulators re-establish the series-circuit current and the lights come back 5 MARGIN what is left under the limit — absorbs a slow start or an ageing switch TIMED ON TEST WITHIN THE LIMIT, PROVEN A recorded interruption on load shows the whole chain meets its class NEVER TIMED DARK LONGER THAN ALLOWED A slow start or a sticky switch is found only in a real outage SCHEMATIC · NOT TO SCALE · FOLLOW ICAO ANNEX 14 VOL. I, THE AUTHORITY AND THE EQUIPMENT MANUALS
Figure 6. Changeover time is part of the design: UPS or the CCR ride-through covers the gap until the DG takes load.
AGL Visual Encyclopedia: Generator changeover and switchover time Standby power and UPS
A row of airfield lighting vault buildings on a US Air Force base
Photo 1. A line of airfield lighting vaults. Each vault houses the regulators and switchgear for its part of the airfield, so the vault itself is part of the redundancy picture.Credit: USAF photo by Lois Walsh, 2009. Released public domain (US Government work). Original source
Field Equipment

CCRs, Cable Routes, Transformers and Surge Protection

CCR redundancy

Two CCRs on two circuits give some redundancy, but two CCRs feeding one common cable still leave a single point of failure. Link the CCR study to circuit and cable topology, and prefer an architecture where a CCR failure affects one circuit, not a whole visual-aid system.

CCR failure modes

Power-electronics fault, controller failure, input breaker failure, output fault, cooling failure, communication failure and internal protection trip.

Physical separation

Two circuits in one duct remain exposed to excavation damage, flooding and fire. Plan route separation where feasible and review crossings, duct capacity, drainage, chambers, access and construction exposure.

Transformers and connectors

Individual isolation transformers mean a transformer failure normally removes one light, a natural form of distributed redundancy. Avoid one connector or chamber affecting several lights, and keep moisture ingress from propagating across circuits.

Redundant circuits must survive the same environment. Two circuits routed together and hit by one lightning or surge event may fail together, so coordinate redundancy with lightning protection, counterpoise, bonding, surge protection and cable routing.

A duty CCR, a standby CCR and the changeover between them.

DUTY CCR · STANDBY CCR · ONE CHANGEOVER SWITCH 1 · DUTY CCR HEALTHY SWITCH ON A · STANDBY IDLE CCR A STANDBY CHANGEOVER SWITCH SERIES CIRCUIT 2 · DUTY CCR FAILED CIRCUIT DARK UNTIL CHANGEOVER CCR A STANDBY CHANGEOVER SWITCH SERIES CIRCUIT 3 · ON STANDBY A ISOLATED · STANDBY CARRIES CIRCUIT CCR A STANDBY CHANGEOVER SWITCH SERIES CIRCUIT WHAT TO CHECK 1 SAME RATING same power, same brightness steps 2 BREAK BEFORE MAKE never two CCRs on one circuit 3 DE-ENERGISE FIRST switch only with the circuit off 4 TEST THE STANDBY run it on a circuit, not just on paper SCHEMATIC · ONE CIRCUIT SHOWN · NOT A SWITCHING PROCEDURE
Figure 7. A standby CCR turns a regulator failure into a changeover instead of an outage.
AGL Visual Encyclopedia: Standby CCR and Changeover Arrangement CCR contactors and bypass CCR Failure Modes

Where surge protection sits on the CCR input, output and circuit.

SCHEMATIC · WHERE SURGE PROTECTION DEVICES SIT ON A CCR AND ITS SERIES CIRCUIT EARTHING NETWORK · ONE BONDED SYSTEM EARTHING NETWORK MAINSSUPPLY CONTROLAND COMMS CCRconstant current regulator SPD 1 · MAINS INPUT SPD 3 · CONTROL LINE SPD 2 · OUTPUT TO EARTH SERIES CIRCUIT · TRANSFORMERS AND LAMPS NEARBY STRIKE SURGE INDUCED IN THE CABLE SURGE ENERGY TO EARTH Each device is idle in normal service, conducts when a surge arrives, and passes the energy to the earthing network through short, direct leads.
Figure 8. Surge protection stops one lightning event becoming a common-cause failure across CCRs and circuits.
AGL Visual Encyclopedia: Surge Protection Devices for CCR and Circuits Lightning/Surge Interface
An airman using a multimeter to test voltage at a constant current regulator in an airfield lighting vault
Photo 2. Testing a constant current regulator in an airfield lighting vault. Regulator monitoring and test access are what make a redundant CCR arrangement usable in practice.Credit: USAF photo by Staff Sgt. Nick Wilson, 2013. Released public domain (US Government work). Original source
Control

ALCMS, Network, Field Controllers, A-SMGCS and Follow the Greens

ALCMS and network

A robust design may use a primary and standby server, two networks and two field controllers where the operational requirement justifies it. Separate lighting power from the control application so a server failure has a defined response. Field equipment may hold its last state, move to a default state or allow local or manual control. Test whichever is specified.

Field controllers

Controllers may drive CCRs, taxiway segments, stop bars, signs and Follow the Greens functions. A controller failure should give a defined safe response and an alarm.

Where AGL is integrated with A-SMGCS, the chain gets longer

Surveillance
A-SMGCS
Routing
ALCMS
Field controller
AGL
Failure analysis must cover surveillance, servers, network, route processing, ALCMS, field controllers and lighting. If control fails, a Follow the Greens system must not leave the airport displaying an unsafe or ambiguous route.

For a safety-related lighting system the question is: what should the pilot see when the system is degraded? The answer should be unambiguous. Redundancy needs interlocking, monitoring and fail-safe logic with it.

Dual network paths between the ALCMS and field controllers.

NETWORK REDUNDANCY · TWO PATHS, NO SINGLE POINT OF FAILURESERVER ALIVESERVER BSTANDBY / DUPLICATENETWORK A SWITCHPATH ANETWORK B SWITCHPATH BABLINK A LOSTSUBSTATION 1DUAL-PORT FIELD CONTROLLERSUBSTATION 2DUAL-PORT FIELD CONTROLLERCCRRUNWAY CIRCUITSCCRTAXIWAY CIRCUITSWHEN ONE PATH FAILS1 · LINK A FAILSSUBSTATION 2 LOSES PATH A2 · TRAFFIC MOVESPATH B CARRIES THE COMMANDS3 · ALARM RAISEDREPAIR BEFORE PATH B FAILSINDEPENDENT MEANSSEPARATE SWITCHESSEPARATE CABLE ROUTESSEPARATE, UPS-BACKED POWER
Figure 9. A single network path is a single point of failure for control, even when the power side is fully redundant.
AGL Visual Encyclopedia: Network Redundancy ALCMS Failure Flow ALCMS Communication Architecture
Failure Analysis

Redundancy Matrix and FMEA

Redundancy matrix: define the response to every failure

ComponentFailureDesired response
Utility feederUtility lossTransfer to secondary source
UPSUPS failureDG or alternate supply, or a defined degraded mode
DGGenerator failureAlternate DG or utility where required
CCRCCR failureAlternate circuit or CCR
Runway circuitCable or circuit failureAlternate circuit maintains required lighting
TransformerTransformer failureAffected light only
ALCMS serverServer failureStandby server
NetworkLink failureRedundant communication path
Field controllerController failureBackup controller or safe degraded mode
Stop barControl failureDefined safe response
A-SMGCS interfaceData failureDegraded or fallback operating procedure
The final response to each failure is set by operational and safety requirements.

Simplified FMEA

FailureLocal effectSystem effectMitigation
CCR-A failureCircuit A lostPartial runway lightingCCR-B / circuit redundancy
Cable-A failureCircuit A lostPartial lightingInterleaved Circuit B
Transformer failureOne light lostLocalRemaining lights
Utility failureNormal source lostPotential full outageUPS / DG
UPS failureFast backup lostChangeover degradationSecondary source
DG failureExtended backup lostLimited enduranceAlternate source
ALCMS server failureCentral control lostControl degradedStandby server / manual control
Network link failureCommunication lostPartial control lossRedundant network
Stop-bar controller failureStop-bar command affectedHolding-point riskFail-safe logic / fallback
Light fitting failureIndividual light lostLocal serviceability lossDistributed lighting

Redundancy claims are only as good as the failure modes they were checked against.

FMEA EXCERPT · ONE FAILURE AT A TIME · WHAT IS LOST, HOW IT IS SEEN, WHAT STILL WORKSFAILUREWITHOUT REDUNDANCYDETECTED BYWITH REDUNDANCYIncoming feeder lostVault darkVoltage monitor, ALCMS alarmSecond feeder or DG takes the loadCCR failureWhole circuit darkCCR monitor, output alarmStandby CCR via changeoverCable cut, circuit openWhole circuit darkOpen-circuit alarmOther circuit keeps a thinned patternNetwork switch lostField control lostHeartbeat lossSecond network path, local controlALCMS server lostRemote control lostWatchdog, failover alarmStandby server, CCR local controlILLUSTRATIVE · NOT A SITE FMEA · RATINGS AND ACTIONS COME FROM THE PROJECT RISK ASSESSMENT
Figure 10. An FMEA excerpt shows, for each failure, what is lost, how it is seen and what still works.
AGL Visual Encyclopedia: CCR Failure Modes Transformer Failure Modes Sign Failure Modes
Operations

CAT II/III, Serviceability and Maintenance

Low-visibility operations

Redundancy becomes an active operational condition. LVP readiness can depend on circuit redundancy, power redundancy, equipment serviceability and control availability together. AAI’s airport procedures link specific circuit and light failures to restrictions on CAT II/III operations; those limits are airport-specific.

Redundancy vs serviceability

Redundancy gives resilience. Serviceability criteria define how much degradation is acceptable. A failed Circuit A does not mean everything is acceptable indefinitely: Circuit B still has to meet the applicable limits.

Maintenance configurations to define

StateMeaningDefine
NormalBoth systems availableBaseline alarm status
MaintenanceOne system intentionally unavailableRestrictions, restoration procedure, communication
DegradedOne system failed unexpectedlyAlarm, operational restrictions, recovery
EmergencyMultiple failuresProcedures and escalation
A system already running on one circuit should not be exposed to an additional failure without considering operational restrictions. Redundancy also lets planned maintenance proceed without disabling the whole visual aid, subject to the approved procedure.

From serviceability measurement to a maintenance action.

RESULT TO ACTION FROM A SERVICEABILITY RESULT TO THE RIGHT RESPONSE 1 RESULT count and adjacency 2 COMPARE objective, action level 3 WHICH BAND? three responses below EXAMPLE: 60-LIGHT EDGE ROW, CAT II/III Objective 95% allows 3 out. Site action level: 2 out. The gap between 2 and 3 is the margin the flow protects. WITHIN OBJECTIVE 0 to 1 out, none adjacent LOG AND SCHEDULE repair at the next routine visit ROUTINE no notification needed ACTION LEVEL REACHED 2 to 3 out, none adjacent RAISE A WORK ORDER restore soon, spares on hand PLANNED, SOON tell control if the objective is at risk OBJECTIVE NOT MET 4+ out, adjacent pair, or critical light NOTIFY AND RESTORE NOW tell control first; restore one of a pair first URGENT control decides on operations THEN, FOR EVERY BAND RESTORE · RECHECK COUNT AND POSITIONS · LOG TIME FOUND AND RESTORED · REVIEW THE TREND The same circuit or fixture type failing again points to a cause, not a lamp. ICAO SETS THE OBJECTIVE. THE ACTION LEVEL IS THE OPERATOR'S OWN, SET ABOVE IT.
Figure 11. Redundancy hides failures. Serviceability monitoring is what exposes them before the next failure arrives.
AGL Visual Encyclopedia: Serviceability-Triggered Maintenance Flow Minimum Serviceable Lamp Percentages by Light Type Serviceability of Critical Lights
Testing

Proving Redundancy: Failure Simulation, Functional and Photometric Tests

01

Normal condition

Both redundant paths operational.

02

Circuit A failure

Disable Circuit A. Verify Circuit B stays available, the right lights stay lit, alarms appear and required operation is supported.

03

Circuit B failure

Repeat in the opposite direction.

04

Utility failure

Verify secondary power transfer and actual AGL restoration.

05

UPS and DG failure

Verify the defined backup source, alternate source or degraded mode.

06

CCR failure

Verify alternate circuit behaviour.

07

ALCMS and communication failure

Verify standby, local or manual control and the safe field state.

08

Photometry after failure

Verify the remaining lighting still meets the required optical performance.

Commissioning test matrix

TestFailure introducedExpected result
Circuit A tripCircuit A unavailableCircuit B maintains required lighting
Circuit B tripCircuit B unavailableCircuit A maintains required lighting
Utility lossNormal source unavailableSecondary source takes over
UPS failureUPS unavailableDefined backup or degraded mode
DG-1 failureGenerator unavailableAlternate source or approved restriction
CCR-A failureCCR unavailableAlternate circuit remains functional
ALCMS server failurePrimary server unavailableStandby or control fallback operates
Network link failureOne communication path lostRedundant path remains available
Field controller failureController unavailableDefined safe state
Stop-bar control failureCommand path unavailableSafe operational response
PhotometryPost-failure measurementRequired optical performance verified
Electrical design plus failure simulation plus functional testing plus photometric verification is stronger evidence than any one of them. A healthy-looking circuit on screen does not prove the remaining lights perform as required.

A redundant design is only proven when the failure has actually been simulated.

PROVING REDUNDANCY · FAIL IT ON PURPOSE, UNDER CONTROL, AND MEASURE WHAT HAPPENS1 BASELINERecord currents,insulation, photometry2 SIMULATETrip feeder, open circuit,switch CCR off3 OBSERVEChangeover time, lightslit, alarms raised4 VERIFYFunctional andphotometric checks5 RESTOREReturn to normaland record resultsBEFORE STEP 2Approved test plan, airfield and ATC coordination, safety isolation, agreed pass criteriaSCHEMATIC · TEST SEQUENCE AND PASS CRITERIA MUST FOLLOW THE APPROVED PROJECT PROCEDURE
Figure 12. Controlled failure simulation, then functional and photometric verification, then restoration.
AGL Visual Encyclopedia: Generator changeover and switchover time Cable Testing (Commissioning) Commissioning Sequence
Design Workflow

From Requirement to Redundancy Philosophy

01

Define the operational requirement

Runway category and required visual aids.

02

Identify critical loads

Which systems need continuity.

03

Map the existing architecture

Power, CCRs, circuits and controls.

04

Identify single points of failure

Where one fault causes unacceptable loss.

05

Develop the redundancy philosophy

Source, circuit and control redundancy.

06

Check physical independence

Cable routes, switchgear and common systems.

07

Design protection

Lightning, surge and grounding.

08

Design control

Fail-safe switching and monitoring.

09

Develop the failure matrix

Response to each defined failure.

10

Test

Electrical, functional, redundancy and photometric tests.

11

Document

Drawings, cause and effect, test results.

12

Maintain

Periodically prove redundancy is still available.

The sequence is Requirement → Failure Analysis → Redundancy Philosophy → Design, not “install two of everything.” Not every airport needs 2N UPS, N+1 DG, dual independent feeders or two CCRs for every lighting system.

Example gap assessment (illustrative values)

CategoryExistingRequiredGap
UtilitySingleDual / independentHigh
UPSSingleN+1Medium
Runway edgeDual interleavedIndependentLow
Cable routeCommonSeparatedHigh
ALCMSSingleRedundantHigh
Stop barSingle controllerFail-safeMedium
PhotometryPeriodicFailure verificationMedium
The gap table is an example of the format. A real assessment uses the airport’s own requirement, audit evidence and risk assessment.

Documents a redundancy project should deliver

Single-line diagramAGL circuit diagramCircuit interleaving planCCR scheduleTransformer scheduleRedundancy matrixCause and effectFMEATesting procedure
Keyword Reference

AGL Power Supply System & Airfield Lighting Circuit Protection

How the lighting is fed and how the circuits are protected.

AGL power supply system

An AGL power supply system takes power from the airport's electrical supply to the CCRs, which feed the lighting series circuits. Redundancy is added through independent feeders, backup power such as UPS or a generator, and alternate circuits, according to the category of operation. The supply design should be checked against the standard the aerodrome follows.

Airfield lighting circuit protection

Airfield lighting circuit protection guards people and equipment against faults. It includes CCR protection such as open-circuit and overcurrent protection, earth fault monitoring, and upstream breakers and fuses. Settings come from the manufacturer's data and the project design, so none are given here. Good protection also helps locate faults quickly.

Keyword Reference

AGL Circuit Fault Troubleshooting: Open Circuit & Short Circuit Faults

Finding and understanding the two most common series-circuit fault types.

AGL circuit fault troubleshooting

AGL circuit fault troubleshooting starts with the symptom and the CCR status: tripped, low current, or high current. Make the circuit safe, then split it into sections and test each with insulation and loop resistance checks until the faulty section is found. Repair, re-test and record the cause. See the testing and failure simulation sections of this guide.

AGL open circuit fault

An AGL open circuit fault is a break in the series loop, such as a damaged cable, a failed joint or a poorly mated connector. Current stops flowing, and the CCR normally trips on its open-circuit protection. Never work on a series circuit that is open but energised. Locate the break by sectionalising the loop and checking continuity.

AGL short circuit fault

An AGL short circuit fault is an unwanted path in the circuit, often a short to earth caused by water or insulation damage. It can cause overcurrent, earth-fault alarms or a CCR trip. Locate the fault by sectionalising and insulation testing, then repair or replace the damaged cable, joint or transformer.

What an open circuit does to a series circuit.

SERIES LOOP · OPEN CIRCUIT (BREAK IN THE PRIMARY PATH) ISOLATE & LOCK OUT BEFORE TOUCHING CABLE ENDS 0.0 A CCR OUTPUT TRIPPED T1 T2 T3 T4 1 OPEN-CIRCUIT PROTECTION TRIPS 2 BREAK IN THE PRIMARY LOOP 3 ALL LAMPS OUT — NO CURRENT PATH one break stops the whole series circuit 4 COMMON BREAK POINTS: CABLE · CONNECTOR · PRIMARY WINDING LAMP / SECONDARY OPEN LOOP STAYS CLOSED Only that fixture goes dark PRIMARY LOOP OPEN WHOLE CIRCUIT DARK Every fixture on the loop goes out SCHEMATIC · LOOP SHORTENED · NOT TO SCALE
Figure 13. An open circuit breaks the loop, so the whole circuit goes dark and the CCR trips on open-circuit protection.
AGL Visual Encyclopedia: Open Circuit Series Circuit Fault Path Cable Fault Location (TDR and Bridge Methods)

What a short circuit does to a series circuit.

SERIES LOOP · SHORT CIRCUIT (BYPASS BETWEEN CONDUCTORS) ISOLATE & LOCK OUT BEFORE TOUCHING CABLE ENDS LOW kV CCR LOAD REDUCED T1 T2 T3 T4 1 CCR HOLDS CURRENT 2 SHORT BETWEEN THE CONDUCTORS 3 T1–T2 STAY LIT 4 BYPASSED no current, lamps out 5 CAUSES: FAILED INSULATION · DAMAGED CABLE · WET CONNECTOR SHORT CIRCUIT SECTION GOES DARK Current flows, loop is cut short READS ON THE TESTER FAR BELOW BASELINE Open circuit reads OL instead SCHEMATIC · LOOP SHORTENED · NOT TO SCALE
Figure 14. A short removes load from the circuit, so the lights on it may stay lit while the fault is still there.
AGL Visual Encyclopedia: Short Circuit Ground-Fault Detection and Insulation Monitoring Earth Fault
Checklist

Redundancy Commissioning Checklist

Tick items as you go. Progress shows per group and is not saved.

Power

Normal source tested
Secondary source tested
UPS tested
DG tested
ATS tested
0 / 5 complete

CCR and circuits

Each CCR tested
Alternate path verified
Circuit A and B tested
Interleaving verified
Failure simulation completed
0 / 5 complete

Field and control

Transformers and connectors checked
Light serviceability verified
ALCMS tested
Field controllers tested
Network redundancy tested
0 / 5 complete

Safety, optical and documents

Stop bars and interlocks tested
Fail-safe states verified
Photometry completed
As-built drawings and circuit schedules
Redundancy matrix and test reports
Maintenance procedures
0 / 6 complete
Pitfalls

Common Mistakes in AGL Redundancy Design

Two circuits, one common failure point

Two circuits in the same vulnerable cable route may fail together.

Two CCRs on one common supply

CCR duplication is ineffective if the common supply remains a single point of failure.

UPS without generator support

The UPS gives fast continuity but may not last a prolonged outage.

Generator without fast transfer

Long-duration power that fails the required visual-aid switch-over time.

Redundant power, single control

The lighting is powered but cannot be operated correctly because control failed.

Redundant control, single field circuit

Control survives, but the physical lighting still goes dark.

No failure testing

A system that looks redundant on paper can fail in a real outage.

No common-cause analysis

Duplicate components can still depend on one cable, room, network or control power source.

Standards

Standards and Regulatory Context

Use these together. They are framework and guidance documents, not substitutes for the approved project design.

ICAO Annex 14 Vol I

Aerodromes: the principal international framework, including secondary power and switch-over provisions. Check the current edition, amendments and corrigenda.

ICAO Doc 9157 Part 5

Aerodrome Design Manual, Electrical Systems: specialist guidance for aerodrome electrical engineering.

DGCA CAR Section 4, Series B, Part I

Indian requirements referenced in AAI publications in connection with AGL serviceability and power.

AAI eAIP and AIP supplements

Airport-specific AGL, UPS, DG and low-visibility procedures, useful as practical examples.

Airport-specific figures in published procedures, such as UPS and DG ratings or circuit failure limits, belong to those installations. They are not generic sizing rules. For an Indian airport, review applicable DGCA requirements, AAI or project specifications and controlled equipment documentation. Confirm the current edition of each reference before relying on it.
Services

AGL Redundancy Engineering Services

Redundancy audit

Existing power sources, CCRs, circuits, cable routes, transformers, ALCMS and controls.

Electrical architecture

Single-line diagrams, circuit architecture, CCR allocation and critical-load classification.

Dual circuit design

Independent circuits, interleaving, segregation and fault isolation.

UPS / DG architecture

UPS, DG, ATS, secondary power and critical AGL loads.

ALCMS redundancy

Servers, networks, field controllers and control interfaces.

A-SMGCS and Follow the Greens

Fail-safe design for green centreline lighting, stop bars and control interfaces.

Failure analysis

FMEA, redundancy matrix, cause and effect and single-point-of-failure assessment.

Testing and commissioning

Circuit, power, CCR and control failure testing, plus photometric verification.

Existing-airport upgrade

Retrofit planning that improves reliability without replacing healthy infrastructure.

Final Engineering Perspective

Redundancy is not duplication for its own sake. It is a structured way to make sure one failure does not compromise the airport’s required visual guidance.

Design the system around the failure you need to survive, not around the number of components you want to duplicate.

Test Yourself

Quick Knowledge Check

Q1. What does a redundant AGL design aim to do?

A. Install two of every component
B. Maintain the required lighting service after a defined single failure, without misleading indications
C. Remove the need for maintenance
D. Replace the CCR with a UPS

Q2. Why are runway lights interleaved between two circuits?

A. To make the lights brighter
B. So one circuit failure does not create a long continuous dark section
C. To avoid using transformers
D. To reduce cable length to zero

Q3. Two circuits are interleaved but run in one duct. What is the main concern?

A. The lights will flicker
B. A common-cause event such as excavation or flooding could take out both
C. The CCR current doubles
D. Nothing, interleaving guarantees independence

Q4. How do UPS and DG differ in the typical arrangement?

A. UPS gives fast continuity, DG gives sustained backup
B. DG gives fast continuity, UPS gives sustained backup
C. They do the same job
D. Both only protect the ALCMS

Q5. What does N+1 mean?

A. Two complete independent systems
B. The required capacity plus one additional unit
C. One unit of spare floor space
D. A generator with one fuel tank
Explore More

Explore More AGL Engineering

HUB

How Airfield Lighting Works

The whole chain from power to optics.

Open How Airfield Lighting Works →
POWER

Emergency Power & Lighting Systems

UPS, DG and transfer.

Open Emergency Power & Lighting Systems →
PROTECTION

Lightning & Surge Protection

Protecting AGL circuits.

Open Lightning & Surge Protection →
ELECTRICAL

CCR Working Principle

Regulation, steps and faults.

Open CCR Working Principle →
ELECTRICAL

Airport Isolating Transformer

Ratings, types and testing.

Open Airport Isolating Transformer →
CONTROL

ALCMS

Lighting control and monitoring.

Open ALCMS →
SAFETY

Follow the Greens

Dynamic taxiway guidance.

Open Follow the Greens →
STANDARDS

ICAO Aerodrome Lighting Requirements

The ICAO AGL framework.

Open ICAO Aerodrome Lighting Requirements →
PERFORMANCE

AGL Photometry

Photometric testing and reporting.

Open AGL Photometry →
FIELD GUIDE

Runway Lighting Photometry

Runway fixture testing.

Open Runway Lighting Photometry →
SAFETY

A-SMGCS India

Surface movement guidance and control.

Open A-SMGCS India →
INDIA

AGL Photometry India

In-field photometric testing for Indian airports.

Open AGL Photometry India →
INDIA

AGL Consultant India

AGL design, specification and technical review.

Open AGL Consultant India →
INDIA

AGL Contractor India

AGL installation, testing and commissioning.

Open AGL Contractor India →
INDIA

Airport Lighting Companies India

Company types across the Indian AGL market.

Open Airport Lighting Companies India →

Design for the Failure You Need to Survive.

Audit, architecture, testing and commissioning for reliable airport lighting power and circuits.

AGL Engineering Hub

Explore the AGL Engineering Knowledge Hub

Fundamentals

Standards & Compliance

Runway, Taxiway & Approach Lighting

Surface Movement & Safety

PAPI

Electrical Engineering & Power

Photometry & Testing

Maintenance, Inspection & Safety

Resources, Documentation & Careers

FAQ

AGL Redundancy FAQ

Q1What is redundancy in AGL?

AGL redundancy is the use of alternate power, circuit, equipment or control paths so that a defined single failure does not unnecessarily remove the required airport lighting function.

Q2Why are two AGL circuits used?

Two circuits can distribute runway lights so that failure of one circuit does not remove the entire lighting system. Interleaving can preserve spatial continuity.

Q3What is an interleaved AGL circuit?

A circuit arrangement in which adjacent lights are distributed between separate circuits rather than placing all consecutive lights on one circuit.

Q4Are two interleaved circuits automatically independent?

No. The circuits may still share common feeders, cable routes, switchgear, chambers or control systems.

Q5What is the difference between independent and dependent circuits?

Independent circuits have sufficiently separate electrical infrastructure to reduce common-cause failure. Dependent circuits can share infrastructure or dependencies.

Q6What does N+1 mean?

N+1 means providing the capacity required for normal operation plus one additional unit, so that one unit can fail without removing the required capacity.

Q7What is 2N redundancy?

2N means two independent complete systems, each capable of supporting the defined critical load.

Q8Does redundancy mean all runway lights stay on after one circuit fails?

Not necessarily. The remaining serviceability is defined by the applicable operational and regulatory criteria, which differ by lighting system, runway category and airport.

Q9Why are runway lights interleaved?

Interleaving spreads lights from separate circuits across the runway so that one circuit failure does not create a large continuous dark section.

Q10Does redundancy apply to CCRs?

Yes. CCR allocation can be an important part of circuit redundancy, provided the alternate path is sufficiently independent.

Q11Does a UPS provide redundancy?

A UPS provides power continuity. It is not automatically redundant. Redundant UPS architecture may use N+1, parallel or dual-independent arrangements.

Q12Why is a diesel generator needed when there is a UPS?

The UPS can give rapid continuity, while a generator can supply the system through a much longer outage.

Q13Is dual utility power better than UPS and DG?

Not universally. Each architecture has different characteristics. The right solution depends on source independence, switch-over requirements, outage duration, reliability and project requirements.

Q14How is AGL redundancy tested?

Engineers simulate defined failures such as circuit trip, CCR failure, utility loss, UPS failure, generator failure and control or network failure, then verify the system response and remaining lighting performance.

Q15Is photometry required after redundancy testing?

Where optical performance forms part of acceptance or verification, photometry can show that the remaining lighting performs as required after a simulated failure.

Q16What is the difference between redundancy and switch-over time?

Redundancy is whether the system can continue after a component failure. Switch-over time is how quickly service is restored after the source fails. They are separate engineering parameters.

Q17How to measure AGL circuit current?

Measure the series circuit current at the CCR output, using the CCR's own display or a calibrated clamp meter or current instrument rated for the circuit. Never open a live series circuit to insert a meter. Take readings at each brightness step, compare them with the rated current for that step, and record them. Work should be done by qualified staff under your site's safety procedure.

Q18How to check an airport lighting circuit?

Start with the CCR status and any alarms on the monitoring system, then measure the circuit current at the brightness steps. With the circuit isolated, test insulation resistance and loop resistance, and inspect connectors, transformers and fixtures along the circuit. Compare results with the commissioning records and acceptance limits, and record the findings.

Q19How to restore an airfield lighting circuit?

Make the circuit safe and find the cause before re-energising, rather than resetting repeatedly into a fault. Isolate the faulty section, repair or replace the damaged cable, joint, connector or transformer, then test insulation and continuity. Re-energise at a low brightness step, check the current, and return the circuit to service in agreement with air traffic control. Where redundancy exists, switch to the alternate circuit first. Record the fault and the repair.

Field Insight

Engineering & Expert Notes

Redundancy Has to Be Proven, Not Assumed
Verification

Drawings can show duplicated paths that share a single point of failure in practice. Test failure scenarios to confirm that the airfield stays serviceable.

Look for Hidden Common Points
Single Points

Shared cable routes, ducts, control power or a common network switch can defeat apparent redundancy. Trace each path to its source.

Switch-Over Behaviour Matters as Much as the Backup Source
Power

A standby source that starts correctly may still disturb CCRs or controls during transfer. Test the transfer sequence under realistic load.

Control Redundancy Needs Its Own Check
Control

Dual servers or networks protect the operator view but only help if failover is tested and the local fallback is understood by staff.

Use a Failure Matrix to Make Reviews Concrete
FMEA

List each element, how it can fail and what happens to the lights. The table exposes unexpected consequences before an actual outage does.

Re-Test After Any Modification
Change Control

Extensions and upgrades can quietly remove a redundancy feature. Repeat the relevant failure tests after changes to the architecture.

Precision AGL Academy

Understand AGL. Build Your Engineering Skills.

You now understand how redundancy is built into AGL electrical architecture, from independent power sources and dual circuits to fail-safe control and failure testing.

Take the next step with Precision AGL Academy.

Learn AGL Step by Step

Build your knowledge through structured technical learning covering:

AGL Fundamentals Components Electrical Systems Design Photometry Testing Troubleshooting Maintenance

Move from understanding AGL redundancy to developing the practical knowledge needed to analyse, operate and maintain complete AGL systems.

Understand the system. Build the knowledge. Apply the engineering.