Preventive vs. Corrective Maintenance

There is a version of this argument that treats preventive maintenance as always right and corrective maintenance as always a failure of planning. It is not that simple. Servicing every machine on the same fixed calendar wastes parts and labour on equipment that was working perfectly well. Waiting for everything to break costs far more, and usually at the worst possible moment. The useful question is not which approach is better, but which machines deserve which treatment.

Start from what failure actually costs

Before deciding how to maintain a machine, work out what happens when it stops. A conveyor that halts the whole production line and a spare pump sitting in parallel with two others are not the same problem, even if they cost the same to repair. The first justifies a scheduled servicing plan, spare parts held in stock and condition monitoring. The second can reasonably be run until it fails.

This sounds obvious written down, and it is routinely ignored in practice, because maintenance plans tend to be built around what is easy to schedule rather than around what is expensive to lose.

What a preventive plan is actually made of

A preventive plan that works is built on operating hours and duty cycle, not on the calendar. A pump that runs four hours a day and one that runs continuously will not wear at the same rate, and servicing them both every six months means one is being over-serviced while the other is already past due.

• Servicing intervals derived from operating hours, not from the month on the calendar
• Inspection points defined in advance, so different technicians check the same things
• Wear limits written down, so "still acceptable" is a measurement and not an opinion
• Spare parts identified before they are needed, with lead times known

Where corrective maintenance is the correct choice

Corrective maintenance is the right answer for equipment that is cheap to repair, easy to reach, and whose failure does not stop anything else. Redundant equipment falls into this category. So does anything approaching the end of its service life, where the sensible plan is to run it out and replace it rather than keep investing in it.

The mistake is not choosing corrective maintenance. The mistake is choosing it by default, for everything, and calling the result a strategy.

A maintenance plan that nobody can execute during a normal working week is not a plan. It is a wish list that will be abandoned by the third month.

The middle ground: condition-based

Between the two sits condition-based maintenance: intervene when a measurement says the machine needs it, not when the calendar says so. Vibration readings on rotating equipment, thermographic inspection on electrical connections, oil analysis on gearboxes. It requires more instrumentation and more discipline in recording readings, but it catches the failures that a calendar misses and skips the interventions a calendar would have forced.

It is not appropriate everywhere. Instrumenting a machine that costs less than the instrumentation is not engineering, it is theatre. But for the critical equipment identified in the first step, it is usually where the return is.

A practical starting point

If you have no plan at all, do not try to build the complete system in one go. List the equipment whose failure stops production. For each item, write down what fails, how you would know it was about to fail, and what the spare part costs and takes to arrive. That single list will tell you more about where to spend maintenance money than any generic servicing schedule.