Maintenance planning: sorting the work when everything looks urgent

In any maintenance department, the list of things to do is always longer than the capacity to do them. The question is therefore never "what needs doing", but "what do we do first, and what do we accept not doing". Useful prioritisation is not about tagging everything as urgent: it is about making the trade-off explicit and defensible, so it holds up when someone challenges it six months later.

Why everything looks urgent

Demand arrives through several channels, and each one carries its own sense of urgency: an operator reports an abnormal noise, an inspection report raises a reservation, a regulatory deadline approaches, the production manager asks for a modification. Every one of these requesters judges their own request to be a priority, and from where they stand they are right.

The maintenance department therefore receives a list of requests all labelled urgent by whoever raised them, with no common yardstick that lets you compare them against each other. In the absence of that yardstick, the trade-off defaults to three fairly unsatisfactory bases: how long the request has been sitting there, how insistent the requester is, and the last incident that made an impression.

It is that last point that distorts priorities the most. After a spectacular failure, anything resembling the equipment involved jumps to the top of the list, and stays there for months, regardless of the real risk it carries. The department ends up working on the memory of the last breakdown rather than on the condition of the plant in front of it.

The three objective criteria

Consequence

What happens if the job is not done? The answer has to be framed in concrete terms, not in abstract levels: a line stopped for x hours, loss of compliance at the next audit, an operator exposed, a rejected batch, accelerated degradation of another component downstream.

This assessment is not maintenance's alone to make. It is built with production, safety and quality, and it is done once for the equipment, not afresh for every request. That is exactly what the criticality matrix is for: a shared, agreed view of what each asset costs the site if it fails, established before anyone is under pressure.

Time before consequence

This is the dimension that criticality on its own does not give you. A highly critical item whose defect evolves slowly leaves you time; a moderately critical item whose degradation is accelerating leaves you none. Ranking by criticality alone quietly assumes every problem moves at the same speed, and that assumption is almost always wrong.

This lead time is estimated from history: measured degradation rate, the margin remaining against the acceptable threshold, and the behaviour observed on comparable equipment. This is where making proper use of inspection reports changes how the department is run in practice, rather than in theory. See estimating remaining useful life for how that figure is built from the record you already hold.

The window constraint

Some jobs can only be carried out during a turnaround, or require a part whose lead time runs to months, or a contractor whose order book fills up well in advance.

This criterion often inverts the ranking you get from the first two. A job of moderate severity, whose window has to be booked six months ahead, has to be decided before a more serious job that can be done on the run. Failing to build in this constraint produces the classic pattern: you open the turnaround and discover a list of work you should have prepared, and you push it back a year. The severity never changed; the opportunity to act on it simply passed.

A sorting method that fits in one meeting

What distorts the trade-offs

• A three-level urgency scale where everything is level 1when everything is a priority, nothing is. If half the requests are tagged urgent, the scale serves no purpose.
• Prioritising without knowing the capacitya ranking with no resource limit is a wish list, not a plan.
• Forgetting procurement lead timesthis is the single most common cause of a year-long slip on a turnaround.
• Reacting to the last incidentthe recent event takes up a disproportionate place in how risk is perceived. Historical data corrects that bias better than any amount of discussion.
• Not recording what was droppeda job set aside with no written decision reappears later as a reproach, and no one can reconstruct the reasoning behind it.
• Confusing equipment criticality with job urgencya critical item where nothing is moving can wait; a secondary item degrading fast cannot.

What data changes about the conversation

Without usable history, a trade-off is argued on convictions, and the most confidently held conviction wins. With history, the discussion changes in nature: you compare degradation rates, remaining margins, and calculated deadlines. The loudest voice in the room stops being the deciding factor.

The discipline of good data also serves to defend a budget. A request backed by a measured trend and traced back to the original reports presents very differently in a committee than a request backed by the manager's experience, even when that experience happens to be right. One can be examined and questioned; the other can only be trusted or doubted.

This is the work an integrity tool should take on: pulling in the existing inspection reports, reconstructing the history for each item of equipment and each condition monitoring location, and flagging what is degrading faster than expected. For the full approach, see risk-based maintenance, our Maintenance Intelligence page, and, on preparing windows, the page on preparing a turnaround.