SiteMonday, 7am. The PPM board has gone red again. Last week's grease round on the case packers didn't happen. Nor did the belt checks on the spiral. Your fitter spent Tuesday and Wednesday on the wrapper, and Thursday night on a drive that kept tripping.
So the planned jobs roll over to this week. This week has its own list. By Friday you're further behind than you were on Monday, and the wrapper's making a noise again.
Every maintenance manager knows this loop. Breakdowns eat the time you set aside to stop breakdowns. Fewer planned jobs get done, so more kit fails, so more planned jobs get dropped.
The problem is rarely that nobody believes in planned work. It gets cancelled first because it can always wait until tomorrow. Tomorrow never comes.
There's no magic ratio
You'll hear people quote a split. So much planned, so much reactive, and you're "world class". We'd ignore it. The right balance depends on your kit, your shifts and what a stop costs you.
What matters is that every machine sits where you chose to put it. Some kit should run to failure. A cheap motor on a non-critical conveyor, with a spare on the shelf and a fitter who can swap it in twenty minutes, doesn't need a monthly check. Let it fail. Change it. Move on.
Other kit can't be allowed to fail. The single palletiser at the end of every line. The compressor with no backup. The chiller that protects a cold store full of stock. Running those to failure is a gamble, and the stake is the whole site.
The trouble starts when nobody decided. The machine is on run-to-failure because the PPM kept getting skipped, and that's a very different thing from a choice. If you haven't ranked your machines yet, start with criticality ranking. It tells you where planned work earns its keep.
What the law cares about
PUWER requires work equipment to be "maintained in an efficient state, in efficient working order and in good repair" (regulation 5). The code of practice is clear that "efficient" here is about safety. In its words, it "is not concerned with productivity" (L22, paragraph 70).
So the law doesn't care how many hours of output you lose. It cares about guards, interlocks, pressure relief and anything else that keeps people safe. That gives you a useful filter. Planned work on safety-critical parts isn't optional, however busy the week is.
There's a sharp line in the same code. A fault that hits production "is normally apparent within a short time". A fault in a safety-critical system could stay hidden unless your checks are built to find it (L22, paragraph 71). Production will always tell you when the wrapper's down. Nobody will tell you the light curtain stopped working last month.
Where safety-critical parts could fail and leave a hidden risk, the code says you need "a formal system of planned preventative or condition-based maintenance" (L22, paragraph 75). Those jobs belong on the list nobody is allowed to cancel.
Breakdowns are more dangerous work
HSE makes a point most managers feel but rarely say out loud. Planned maintenance is safer maintenance. Its guidance on maintaining equipment says that "fewer breakdowns will mean less dangerous contact with machinery is required" (HSE, maintenance of work equipment).
Think about the difference. A planned bearing change happens on a day shift, with the parts in hand, the machine isolated and nobody waiting. The same bearing failing at 3am means a tired fitter, a shift manager asking how long, and a guard that's already half off. Same job. Very different risk.
So when someone says planned work is a luxury, point out that breakdown work is where people get hurt. It also costs more. If you haven't put a figure on that yet, work out what an hour of downtime costs you. Then compare it with what the skipped PPM would have cost.
Protect the planned hours
Here's our view. The balance doesn't fail because the plan is wrong. It fails because the same people do both jobs, and the breakdown always wins.
A few things help:
- Split the list. Mark the PPM jobs that can slip a week and the ones that can't. Safety-critical checks and anything on your top-ranked machines go on the "can't" list.
- Give planned work a home: a fixed slot each week, on a day shift, when the line is down or quiet. HSE's guidance suggests using downtime for maintenance, before start-up or during shutdown periods.
- Track what got dropped. If a planned job is cancelled, it gets a reason and a new date. Three rollovers on the same job is a warning sign.
- Watch the repeat faults. If the same machine fails every month, the PPM isn't working on it. Wear parts you should never run to failure is a good place to start looking.
The real cost of being short
Be honest about the numbers. A team with one fitter per shift can't do both jobs well. When that fitter is on a breakdown, nobody is doing the planned work. When they're off sick or on holiday, nobody is doing either.
So the planned work slips, and the kit gets worse. The breakdowns get longer, because the fitter is stretched and the one person who knows the drive is on leave. Each skipped job adds a little more risk of the next long stop.
Extra hands at the right moments break the loop. That might be cover for a week of holidays, so the planned list keeps moving. It might be an outside engineer for the breakdown, so your own fitter stays on the PPM. Either way, the aim is the same. Planned work should never be the thing that gives when you're short. Our guide to covering holidays without burning out the rest of the team goes into the rota side of it.
What to do on Monday
- Pull last month's PPM list and count the jobs that rolled over. Note which machines they were on.
- Mark every PPM job as "can slip" or "can't slip". Put safety-critical checks on the "can't" list.
- Pick the machines you're happy to run to failure, and check there's a spare on the shelf for each.
- Book one fixed slot a week for planned work, and decide now who covers the breakdowns during it.
* Paragraph numbers refer to L22, fourth edition (2014, amended 2018), correct as of 29 August 2026. Check HSE's L22 page and PUWER on legislation.gov.uk for the latest.