- Criticality ranks assets by the consequence of their failure, so maintenance strategy and spares follow risk rather than habit.
- Score consequence across safety, environment, production, quality and repair cost, then combine with failure likelihood.
- Redundancy and detectability change criticality substantially and are routinely left out of scoring.
A pump with an installed standby has a very different criticality from an identical pump with none, and a failure that announces itself is far less dangerous than one that stays hidden until it is needed. Scoring consequence without these produces a ranking that does not match reality.
A workable scoring model
- Safety. Could failure injure someone? Score highest and treat as an override rather than an average.
- Environment. Release, breach of consent, reportable incident.
- Production. Output lost per hour of downtime, multiplied by expected repair time.
- Quality. Could failure produce defective product that reaches the customer?
- Repair cost. Parts, labour and expediting.
- Likelihood. From history where available, from condition and age where not.
Adjusting for redundancy and detectability
Reduce criticality where a standby exists and is proven to start, which means it is tested. Increase criticality for hidden failures, typically protective devices such as relief valves, alarms and trips, where failure is only discovered when the protection is needed. Hidden failures are the classic underweighted category.
Running the exercise
- Get operations, maintenance and safety in one room. Criticality assessed by maintenance alone consistently underweights production impact.
- Work from the asset register, not from memory, so nothing is missed.
- Score quickly and consistently; precision matters less than applying the same scale throughout.
- Resolve the disagreements rather than averaging them, because the disagreements are where the useful information is.
What to do with the ranking
- Top band: predictive monitoring, critical spares held, defined contingency, and drilled response.
- Middle band: preventive schedule based on evidence, spares available with acceptable lead time.
- Bottom band: run to failure by deliberate decision, with the decision recorded.
- Review the ranking when process changes shift which assets matter.
Reviewing it
Criticality is not permanent. A change in product mix, a new bottleneck, the removal of a redundant unit or a new environmental consent can all move an asset between bands. An annual review that takes an hour is enough to keep the ranking honest.