In short. A failure mode and effects analysis (FMEA) walks through each component, asks how it can fail and what happens to the loads, then ranks the risks. Failure scenarios extend this to sequences of events. Together they show where redundancy is thin and tell you what to test.
What an FMEA is
An FMEA is a structured table. For every element on the single-line diagram it records the failure modes, the local and system effects, how the failure would be detected, and what protects against it. It is a design tool, best done early and repeated when the design changes.
Typical columns
| Column | Content |
|---|---|
| Item | Component or function |
| Failure mode | For example fails to start, fails to open, spurious trip, overheats |
| Cause | Mechanical, electrical, control, human, environmental |
| Effect | On the item, the subsystem and the loads |
| Detection | Alarm, monitoring, test, none |
| Existing protection | Redundancy, interlock, procedure |
| Severity, occurrence, detection scores | Used to rank risk |
| Action | Design change, test, maintenance task |
Scoring risk
A common method scores severity, occurrence and detectability from 1 to 10 and multiplies them to give a risk priority number. Treat the number as a ranking aid, not a measurement. A low-likelihood failure with catastrophic consequence deserves attention regardless of its product. A failure that is hard to detect deserves extra weight.
Failure modes of interest in power systems
- Fail to start, fail to transfer, fail to open or close.
- Spurious operation: trip or transfer when none is needed.
- Degraded output: low voltage, frequency drift, overload.
- Loss of control or communications.
- Loss of DC or auxiliary supply.
- Failure to detect a fault, or incorrect setting.
From FMEA to scenarios
Scenarios combine events: utility fails and the first generator does not start; a breaker fails to close during a transfer; a fault occurs while one path is under maintenance. Write each as a sequence with expected system response and acceptable outcome. They become test cases. See failure simulation and recovery testing.
Worked illustration
For an ATS, one failure mode is “fails to transfer on generator availability”. Effects: the load stays dead after the generator starts. Causes: stuck contactor, control relay fault, lost control supply. Detection today: none, until demand. Severity is high, occurrence is low, detection is poor, so the ranking is high. Actions: add a transfer-failure alarm, a manual operating handle and procedure, a periodic transfer test, and a second control supply. The FMEA led to specific, testable actions rather than general worry.
Making the FMEA useful
- Involve operations and maintenance, who know how things fail.
- Include controls, support systems and people, not only power equipment.
- Track actions to closure and update the study after changes.
- Link each high-ranked item to a test.
Scenario thinking beats component thinking
A failure scenario describes a sequence: an initiating event, the system response and the outcome. Walking through scenarios finds problems that component lists miss.
| Scenario | Question |
|---|---|
| Utility outage lasting 10 seconds, 10 minutes, 10 hours, 3 days | Does each source and fuel stock cover it? |
| Generator fails to start | What happens when UPS autonomy ends? |
| Generator trips under load | Does standby start and pick up? |
| Breaker fails to open on fault | Does backup protection clear it, and how much is lost? |
| Fire in switchgear room | Which paths survive? |
| Loss of control power or communications | What is the safe state? |
| Wrong operator action | Do interlocks stop it? |
| Flood or extreme heat | Which equipment is affected? |
Scoring and action
For each scenario score severity, likelihood and detectability, rank them and assign actions. Record the evidence that the action works.
Local scenarios
For Hyderabad and Secunderabad, add monsoon flooding of basements, summer ambient above design, dust loading of filters and lightning strikes on outdoor equipment.
Output
A scenario register with owners and review dates, feeding the test plan. See failure simulation and recovery testing and electrical FMEA.
Common mistakes
- Treating the FMEA as a document to file rather than a tool to act on.
- Considering only single failures and missing combinations during maintenance.
- Ignoring detection, so hidden failures stay hidden.
What to send us
Send the single-line diagram, control philosophy and operating procedures. Send your application details. See also electrical FMEA.
Questions engineers ask
When should an FMEA be done?
During design, so findings can change the design, and again after major changes or incidents.
Is a risk priority number reliable?
It is a ranking aid based on judgement. Use it to prioritise discussion, not to claim precision.
What is the difference between FMEA and a fault tree?
An FMEA starts from component failures and traces effects. A fault tree starts from an unwanted outcome and traces possible causes.
How many scenarios should I test?
All high-ranked ones, and a sample of others, as the site allows.
This guide gives general principles. A site-specific FMEA needs your drawings and operating experience.