Most maintenance teams can tell you how many hours a line stood still last month. Far fewer can tell you what those hours cost. That number matters, because it decides which machines deserve monitoring, how much spare stock is sensible and whether a repair that takes one more shift is acceptable.
This article shows a way to calculate it with data most plants already have.
Start with the right question
The question is not "how much revenue did we lose?". It is: how much worse off is the plant after an unplanned stop than if the same hours had been planned? That difference is what an unplanned stop costs, and it is the number to compare with the cost of preventing it.
What to count
Split the cost into two parts: costs that grow with every hour of the stop, and costs that appear once per stop.
Per hour of stopped production
- Lost contribution margin. The money the plant would have earned from the output of that hour, after the variable costs that you save when you do not produce (materials, energy). Use contribution margin, not revenue.
- Idle labour. Wages and overhead of the people who cannot work while the equipment is down. Use the loaded cost, not the base wage.
- Fixed costs that run anyway are usually left out, because they would be paid with or without the stop. Count them only if the stop pushes you into overtime or an extra shift.
Once per stop
- Restart losses. Scrap, off-spec product, material that has to be flushed or reheated, time to bring the process back to its settings.
- Repair costs. Spare parts, outside service, overtime, courier and express delivery. The premium you pay for speed is part of the cost.
- Penalties and lost orders. Late-delivery penalties, orders that went to someone else. Only count what actually happened.
- Knock-on effects. Downstream lines waiting for your output, rescheduling work, a quality check after restart.
A worked example
The numbers below are made up to show the arithmetic. Use your own.
A packaging line makes 120 units per hour. Each unit leaves a contribution margin of €18. Eight operators work on the line at a loaded cost of €28 per hour each.
| Item | Calculation | Result |
|---|---|---|
| Lost margin per hour | 120 × €18 | €2,160 |
| Idle labour per hour | 8 × €28 | €224 |
| Cost per hour | €2,384 |
A stop lasts 6 hours. Restarting wastes about €300 of material, and the repair needs €1,200 of overtime and an express-delivered part.
| Item | Result |
|---|---|
| 6 hours × €2,384 | €14,304 |
| Restart losses | €300 |
| Repair, overtime, express delivery | €1,200 |
| Cost of this stop | €15,804 |
If the same repair had been planned for a weekend, most of the €14,304 and the express premium would not have been paid. That gap is the real price of the surprise.
Mistakes that distort the figure
- Using revenue instead of margin. This makes every stop look more expensive than it was.
- Ignoring that lost output can sometimes be made up. If the line runs an extra shift on Saturday, the cost is that extra shift, not the lost margin. Count one or the other, not both.
- Averaging over the whole plant. A stop on the bottleneck machine costs far more per hour than a stop on a machine with a spare. Calculate per line, or per critical machine.
- Counting only hours and forgetting events. Ten five-minute stops can cost more than one hour-long stop, because each one has restart losses.
- Forgetting the hidden hours. The time between the failure and the moment someone notices, and the time waiting for a part, are part of the stop.
Where to get the data
- Production and ERP data: output per hour, margin per product.
- Shift logs and the maintenance system (CMMS): stop times, causes, repair costs. Use at least twelve months, so seasonal effects and rare long stops are included.
- Finance: loaded labour cost, penalties.
- Maintenance staff: they know which machines really stop the line and which only slow it down.
If a number is missing, make an honest estimate and write it down as an estimate. A rough figure with stated assumptions is more useful than none.
What to do with the number
Once you have a cost per hour for each critical machine, three decisions become easier:
- Where to monitor first. Start with the machines where an hour of stop is expensive and the equipment gives measurable warning signs, such as vibration or temperature.
- How much to spend on prevention. If a failure costs €15,000 and happens twice a year, a monitoring setup that catches even one of them already has a clear reference.
- How much spare stock to hold. The longer the supplier needs to deliver, the more expensive waiting is.
You can try this reasoning with your own figures in the cost calculator on our site: enter the hours of unplanned stops per year and the cost of an hour, and choose the reduction you assume. The result is only as good as your inputs, and the reduction is your assumption, not a promise; a pilot on your own machines is what measures the real figure.