Key takeaways
- Downtime cost has three parts: lost production value + idle labor + other (scrap, energy, expediting).
- Lost sales alone understates the real number, often by half.
- Siemens puts an hour of downtime at $36,000 in a large consumer goods plant and $2.3 million in a large automotive plant (source: Siemens Senseye, a predictive maintenance vendor's survey, 2024).
- In the worked example below, a 25% reduction is worth €86,400 a year from one line.
Ask two people what an hour of downtime costs and you'll get two very different answers. Finance sees lost sales. The plant sees a scramble. Both are right, and both are incomplete. The true cost is bigger than either, and getting it right is the difference between a maintenance budget that gets approved and one that doesn't.
The three components
The defensible cost of one stopped hour adds three things:
- Lost production value = output rate × contribution margin per unit. The profit you didn't make: not revenue, margin.
- Idle labor = operators still on the clock × their fully-loaded hourly rate. They're paid whether the line runs or not.
- Other costs = scrap produced at restart, wasted energy and utilities, and any overtime or expediting freight to catch up.
A worked example
A line making 1,000 units/hour at €1.20 margin, 4 operators at €35/hour loaded, and €100/hour of restart and energy cost:
| Component | Per hour |
|---|---|
| Lost production (1,000 × €1.20) | €1,200 |
| Idle labor (4 × €35) | €140 |
| Other (scrap, energy) | €100 |
| Total per hour | €1,440 |
| At 20 hours/month → per year | €345,600 |
Enter your own figures and get the per-hour, monthly and annual cost instantly, with a breakdown you can take to finance.
What downtime really costs, by scale
Published figures vary by orders of magnitude, which is exactly why your own number matters more than any headline average. Siemens' survey of large plants gives the two ends of its range (source: Siemens Senseye, a predictive maintenance vendor's survey, 2024):
| Context | Cost of one hour of downtime |
|---|---|
| Large fast-moving consumer goods plant | $36,000 |
| Large automotive plant | $2.3 million |
Both are estimates for the world's largest companies by a vendor of predictive maintenance software, extrapolated from 181 interviews held between April 2019 and March 2023; a single small line costs far less. Your own number comes from the downtime cost calculator.
What a neutral study found
Published figures without a vendor behind them are rare. The US National Institute of Standards and Technology surveyed US discrete manufacturers for 2016 and found that unplanned downtime took 7.8% of planned production time, and estimated that downtime from maintenance issues cost these manufacturers $18.1 billion that year (source: NIST, AMS 100-34, 2020). The plants relying most on reactive maintenance lost 13.0% of planned time to downtime, against 4.0% for the plants relying on it least. For one line, your own calculation is still the number that counts. More sourced figures are collected on the manufacturing downtime statistics page.
The part most calculations miss
Big breakdowns get logged and costed. The expensive surprise is the small stuff: the short, frequent stops that never make it into a downtime report at all. Individually they're trivial; together they can cost more than the headline breakdowns. That's the hidden factory, and it's why a plant can have "good" downtime numbers and still bleed output. You can size that loss here.
Building the business case
Once you have a credible per-hour figure, the case for reducing downtime almost writes itself. A 25% reduction on the example above is €86,400/year from one line; set that against what monitoring would cost you. The CMMS / Monitoring ROI calculator turns this into a payback figure.
The one caveat: monitoring only pays back if you act on it. You have to find the recurring causes and fix them, which means automatic detection plus a way to see what actually happened.
When a shared utility stops
A stop on a shared utility halts every line it feeds, so its cost per hour is the cost of all those lines together. The guides for the compressed air system, steam boiler, chiller, refrigeration compressor and diesel generator cover their common failures and fixes. The electrical supply works the same way: see the guides for the switchgear, transformer, motor control centre and UPS system. Environmental units can stop production too when a permit requires them to run, as the thermal oxidiser and DAF unit guides explain.
Fabrico tracks downtime live from PLCs, IoT sensors or AI cameras. Where cameras are fitted, the video of a stop sits next to its machine data with AI-suggested likely causes the team validates, and a technician opens the work order from the stop, so recurring downtime stops recurring.
Frequently asked questions
Should I use revenue or margin for lost production?
Margin. You only "lose" the profit you would have earned, not the full sale price (you didn't incur the variable cost of making the unit). Using revenue overstates the figure and gets challenged.
What about planned downtime?
Planned maintenance and changeovers are usually excluded from planned production time, so they don't count as downtime cost, but reducing changeover time (SMED) still improves availability.
What's not in this number?
Hard-to-quantify costs: missed-delivery penalties, lost customers, safety risk, brand damage. So treat the calculated figure as a floor, not a ceiling.