What is single point of failure analysis?
A link-by-link check for the one failure that takes out a whole function, including the ones that only look redundant.
A single point of failure is any one component, cable, supply or control element whose failure alone takes out a function the airfield needs. Single point of failure analysis finds them on purpose: trace the chain from supply to light, and for each link ask what is lost if it fails alone, how much, and how long it takes to restore.
Each finding is then removed with redundancy, reduced with monitoring, spares or procedures, or accepted and recorded. The hardest ones to see are shared causes: two feeders in one trench, one control supply, one UPS, one network link, or a spare that has never been run. They look redundant on a drawing and fail together in practice.
This entry closes the redundant-architecture group: interleaved circuits, dual feeders, a standby CCR and a ring cable are the fixes, and this analysis is how you decide where each is needed and whether it really works.
Every link checked, each single point removed, reduced or accepted on record.
Redundancy trusted from the drawing; shared causes found only in a failure.
| Element | If it fails alone | Finding |
|---|---|---|
| Single feeder to the substation | Every CCR loses its supply | Single point; add a second feeder |
| One CCR per circuit, no spare | Circuit dark until repaired | Single point; add a standby CCR |
| Two feeders in one trench | Both lost together | Hidden single point; separate the routes |
| Shared control supply or UPS | Transfer or remote control lost | Hidden single point; give it a second source |
| Single PLC or network link | Remote control and monitoring lost | Check what the lights do on loss of control |
| Standby never run | Fault found only when needed | Redundancy not proven; test it |
| Any open point in one series loop (cable, connector, transformer primary) | The whole circuit goes dark | Single point per circuit; limit with sound joints, testing and interleaved circuits |
| One lamp | One light affected | Not a single point while the film cutout works; maintenance item |
Frequently Asked Questions
Q1What is single point of failure analysis?
It is a structured review that finds any single component, cable, supply or control element whose failure alone would take out a function the airfield needs. You trace the chain from supply to light, ask what is lost if each link fails on its own and how long it takes to restore, then decide whether to remove the weakness, reduce its effect or accept it knowingly.
Q2How is it done?
List the functions that matter, such as a runway centreline circuit or a stop bar. Trace each one from the utility supply through the feeder, bus, CCR, cable and controls to the lights. For every link ask what goes dark if it fails alone, how much and for how long, rank the findings, and record the action: add redundancy, add monitoring or spares, or accept the risk with a reason.
Q3What are examples of hidden single points of failure?
Two feeders in one trench, a single control or DC supply for the transfer scheme, one UPS, one PLC or network link for remote control, one junction box or handhole that several circuits pass through, a standby CCR that has never been run, and one spare or one person who knows how the system is set up. They look redundant on a drawing and fail together in practice.
Q4Is a single failed light a single point of failure?
No. A failed lamp normally affects only that light, because the film cutout in the isolating transformer keeps the series loop closed, so it is a maintenance item rather than a single point of failure. The questions that matter are about the links that carry many lights: the feeder, the bus, the CCR and the cable. Any open point in a series loop, such as a broken cable, a bad connector or an open transformer primary, takes the whole circuit dark, which is why the cable is a single point.
Q5Does adding redundancy always fix the problem?
No. Redundancy can be defeated by a shared cause, by the switching or changeover device itself, or by a spare that was never tested. It also adds parts, such as switchgear and interlocks, that can fail. After adding a layer, run the analysis again on the new arrangement.
Q6When is it done, and is it required?
Do it at design, after any significant change, and as part of periodic review. Whether a formal analysis is required depends on the operator's safety management system, the authority and the standard, so confirm against your design basis. It also feeds other methods, such as failure mode and effects analysis.
See It. Understand It. Then Learn How to Engineer It.
Recognising how lights are laid out is the first step. The next step is knowing how to design, install, test, commission and maintain them correctly.
Build Practical AGL Engineering Skills
Go deeper with structured training from Precision AGL Academy, covering:
Learn how each light, sign and circuit works as part of a complete airport lighting system, not just as an isolated diagram.