The types of maintenance work, and why the labels matter

Two plants both report ten thousand hours of maintenance last month. One spent most of those hours catching problems before they bit; the other spent them fighting fires. The headline number is identical, but the stories could not be more different, and that difference only shows up once you label the work.

Classifying maintenance by what was actually done lets you see the health of your preventive and predictive programmes, the state of your work management, and how proactive your plant really is. It starts with a simple rule: every hour of maintenance labour falls into one of four buckets, and the four add up to 100%.

The four buckets

• Improvement and modification work: labour on plant improvements and expansions, whether or not the cost is capitalised.
• Structured work identification: the planned routines, predictive or condition-based maintenance, which judges an asset's condition before failure, and preventive maintenance, which services on a fixed interval regardless of condition.
• Corrective maintenance: work that restores function, in two flavours, the corrective work your routines caught before failure, and the corrective work done after failure.
• Other work: any maintenance labour that does not fit the above.

Two more lenses laid over the top

Here is where care pays off. Two further pairs of labels get laid over those buckets, and they are not the same thing as each other. Mixing them up is one of the most common mistakes in work management.

Proactive versus reactive is about intent. Proactive work is done to avoid failures or catch the defects that lead to them: predictive and preventive maintenance, plus the corrective work your routines find early. Reactive work is what breaks into the weekly schedule, the urgent, emergent corrective work done after a failure.

Planned versus unplanned is about preparation. Planned work has been through a formal job-planning process, with labour, materials, tools and safety all identified and packaged before the work began. Unplanned work has not. Planned plus unplanned should total 100%.

The trap to avoid

Planned is not a synonym for proactive, and this is where smart teams slip. A breakdown repair can be fully planned, every part staged and the job package ready, yet it is still reactive, because the failure already happened. Equally, proactive work is not automatically planned; the aim is for it to be, but the two labels answer different questions. One asks whether the work was prepared properly; the other asks whether it was done to prevent or catch a failure. Keep those two questions firmly apart and the whole family of work-management metrics suddenly reads like plain English.

Why the labels decide how much you spend

These labels are not bookkeeping for its own sake; the mix they reveal drives the cost and reliability of the whole plant. Reactive work is genuinely expensive: a repair done in the heat of a breakdown typically costs around three times the same repair done on a plan, once the overtime, the premium-freight parts and the collateral damage are counted. And the evidence that the mix matters is striking: studies of real plants have found that those which hold reactive work to a low share run at a fraction of the maintenance cost of those that do not. The widely used target that falls out of this is to keep proactive work above roughly eighty per cent of the total.

There is a deeper reason the predictive and condition-based bucket exists at all, and it overturns most people's intuition. Decades of failure studies have shown that only around a tenth of failure modes are genuinely age-related, the kind a scheduled overhaul can prevent; the overwhelming majority strike at random or early in life. That single fact explains why simply piling on more time-based preventive maintenance is not the route out of a reactive plant, and can even make matters worse, since opening healthy machines on a calendar addresses few of the failures that actually occur while risking new ones. The way forward is condition-based work that watches for the random failures a calendar will always miss.

A word on run-to-failure

One label deserves rescuing from its bad reputation. Choosing to run a cheap, low-consequence item until it fails is not the same as being trapped in reaction. Done deliberately, costed and documented, run-to-failure is a perfectly legitimate maintenance tactic, the right answer for the light fitting you replace when it blows rather than inspecting it every week. What the work-type metrics should worry you about is not that kind of chosen, planned failure, but the unchosen, unprepared kind, the breakdowns that arrive without warning on equipment you needed. Keep that distinction clear and a high reactive figure can be read for what it really is: mostly a problem, occasionally a sound decision in disguise.

Sort every maintenance hour first by what it was, then by intent, proactive or reactive, and by preparation, planned or unplanned. Keep those last two questions firmly apart, remember that most failures were never going to respond to a calendar, and you will never again mistake a well-organised firefight, or a sensible run-to-failure, for a plant that has its work under control.