How Much Does Unplanned Downtime Cost? A Simple Way to Work It Out
"Downtime is expensive" is something everyone agrees with and almost nobody has put a number on. Without a number, it's hard to decide how much effort a fix deserves. The good news is that a rough figure takes about ten minutes and doesn't need a finance team.
Step 1: Count the stoppages
Look back over the last three months at unplanned stops on the machines or lines that matter most. Maintenance logs, shift handover notes or a whiteboard tally will all do. Work out an average per month. Ignore planned maintenance and changeovers.
Step 2: Find the average time to get running again
For each stop, note how long the line was down, from the moment it stopped to the moment it was producing again. Average those figures. If you don't have the data, ask two or three shift leads for an honest estimate, and use a range rather than a single number.
Step 3: Put a value on an hour
This is the part people get stuck on, so keep it simple. Use whichever of these you can estimate most reliably:
- Lost margin: what the line produces in an hour, multiplied by the margin on those products, if the lost output can't be made up later.
- Labour and overheads: the cost of people and running costs that continue while nothing is made.
- Knock-on costs: overtime to catch up, expedited freight, late-delivery penalties and scrap from restarts.
Add up what applies. A cautious figure you believe is better than an impressive one you don't.
Step 4: Put it together
The sum is: stoppages per month, times average hours per stop, times cost per hour, times twelve.
As a purely illustrative example, imagine 6 stops a month, averaging 2 hours each, at a cost of 1,500 per hour. That's 6 × 2 × 1,500 × 12 = 216,000 a year. Your numbers will differ, and that's the point of working them out.
Step 5: Look at the diagnosis time, not just the repair
Break each stop into two parts: working out what's wrong, and doing the repair. In many plants a large share of the time goes on diagnosis and hunting for information, not on the spanner work. That share is where better access to manuals, past fixes and experienced engineers' know-how can help. Even a modest cut in the average stop time is worth a lot at the scale of the example above.
Use the number
With a figure in hand, you can ask better questions. What would a 10% faster recovery be worth? Which three machines account for most of the cost? Which stops only get fixed quickly when one particular person is on shift?
If you'd like a faster way to run the sum, the calculator on our homepage does the arithmetic and lets you change the assumptions. And if the answer suggests knowledge is part of the problem, Machine Mesh is built to help engineers find the right answer faster, with the source document to back it up.
The short version
- Count unplanned stops and average the time to recover.
- Use a cautious cost per hour that includes knock-on costs.
- Multiply it out to an annual figure.
- Separate diagnosis time from repair time to see where improvement is possible.