Asset criticality assessment: ranking equipment so maintenance effort lands where it matters

The list is long and the crew is not

Most plants have far more tagged assets than they can maintain proactively, and treating them equally means everything gets a superficial routine while the machines that stop production get no more attention than the ones that do not. Criticality is the rationing mechanism. It works because it forces an argument that is otherwise had implicitly and badly: which failures does this business genuinely care about, and what is it willing to run until it breaks? A run-to-failure decision, taken deliberately and documented, is a legitimate strategy and quite different from neglect.

Criteria that actually discriminate between assets

Useful criteria cover consequence and recoverability. On consequence: injury or environmental release potential, effect on output if it stops, effect on product quality, and regulatory or contractual exposure. On recoverability: whether redundancy exists, how long a repair takes, whether the spare is held or has a long procurement lead, and whether failure gives warning or arrives without notice. Scoring these separately matters — a machine with modest output impact but a very long spare lead time can outrank a more visible asset with a shelf-stocked replacement, and that outcome is invisible if you score only on production impact.

Running the session and who has to be in the room

Do it as a facilitated workshop, not as an engineer's desk exercise. Production supplies what actually happens when each asset stops, maintenance supplies failure history and repair reality, quality supplies the product consequences, and safety supplies the consequence ceiling. Work asset by asset against a written scoring guide with plant-specific examples, because otherwise scores drift as the day goes on and the last assets assessed are systematically rated differently from the first. Record the reasoning, not just the score; the reasoning is what makes the ranking defensible when somebody disputes their machine's position.

The ranking has to change what people do

An assessment that produces a coloured spreadsheet and no operational change is common and pointless. Highest-criticality assets should attract a defined maintenance strategy, condition monitoring where a measurable degradation path exists, assured spare availability, priority in the response order and a documented recovery plan. Mid-range assets get routine preventive work. The lowest tier gets inspection and run-to-failure. Each tier should have stated consequences for spares stocking and response, so that when a fault arrives at midnight the priority order is already decided and nobody is negotiating it with a production manager.

Keeping the ranking honest as the plant changes

Criticality moves. The constraint shifts when the product mix changes, a redundant pump loses its redundancy when the standby is cannibalised, a spare that used to sit on the shelf becomes a long-lead item when the builder discontinues it, and a new customer contract makes a previously unimportant line critical. Set review triggers rather than only an annual date: new equipment, layout change, a significant mix change, a discontinued part, or any failure whose consequences exceeded what the assessment predicted. That last trigger is the most informative one and the one most often ignored.

Frequently asked questions

How detailed should the asset register be before assessing criticality?

Detailed enough that each entry is something you would maintain or replace as a unit, and no finer. Registers broken down to individual valves and bearings become unmanageable and the assessment never finishes. Assess at machine or system level first, then decompose only the highest-criticality items into their components, since that is where knowing the weak sub-assembly changes what you stock and what you monitor. Starting too granular is the most common reason these exercises stall.

Does criticality scoring need a formal reliability methodology?

Not to begin with. A structured workshop with agreed criteria and honest scoring delivers most of the practical benefit and can be completed for a whole site in a manageable period. Formal reliability analysis is worth applying afterwards to the small number of assets that came out at the top, where understanding specific failure modes justifies the effort. Applying a heavy methodology to an entire plant usually means it is applied to nothing, because the resource runs out well before the register does.

Should safety consequences be scored alongside production impact?

Score them separately and let a serious safety or environmental consequence override the production score outright. Averaging them lets a machine with a genuine injury potential but low output impact fall down the list, which is indefensible. Many sites use a rule that any asset with a credible severe-harm failure mode automatically enters the top tier regardless of its other scores, and that the resulting maintenance requirement is not subject to the usual budget argument.