How to Calculate the Real Cost of Unplanned Downtime at Your Plant
The number on the maintenance report is the cheapest line on the bill. This is how to price an hour of downtime on your own floor, one cost at a time.
The slide with hours and no dollars
Most monthly operations reviews have a downtime slide. It shows hours, maybe split by line or by reason code. When it shows dollars at all, the figure is usually idle labor plus the repair: people standing around, a technician's time, a part. Those costs are real. On most stoppages they're also the smallest ones.
Siemens tracks this across the world's largest manufacturers. Its 2024 True Cost of Downtime report puts a lost hour at $36,000 in consumer goods and $2.3 million at a large automotive plant, and estimates that unplanned downtime costs the 500 biggest companies about 11% of their revenue. Those are averages across very large operations. Your number is somewhere else, and the only way to find it is to build it from your own books.
First, decide which hour you lost
Not every lost hour costs the same. Eli Goldratt's Theory of Constraints makes the point bluntly: an hour lost at the bottleneck is an hour of output lost for the whole plant, and an hour lost at a machine with spare capacity may cost little beyond the repair, because the work can be made up before anyone downstream notices.
So before any arithmetic, answer two questions about the stoppage. Was the machine that stopped your constraint, or feeding it directly? And was it scheduled to run? A non-constraint machine that goes down during a planned idle window costs you a repair. The press that paces the whole plant going down on a Thursday afternoon with a Friday ship date costs you everything below.
The six lines of the bill
Siemens lists the hidden costs as lost revenue, wages for staff who can't work, the people fixing the problem, emergency parts, and contractual penalties. Add what it costs to recover, and you have a worksheet you can fill in from systems you already run.
| Cost | How to calculate it | Where the number lives |
|---|---|---|
| Lost margin | Units not produced and never made up, times contribution margin per unit (price minus variable cost). Use margin, not revenue. | ERP standard costs, pricing |
| Idle labor | Crew hours standing idle, times the loaded labor rate | Payroll, timekeeping |
| Repair | Maintenance hours (at call-out rates if after hours), emergency parts and shipping, outside contractors | CMMS work order, purchasing |
| Recovery | Overtime to catch up, restart scrap and first-off parts, rework | Payroll, quality system scrap log |
| Expediting | Premium freight to hit or salvage the ship date | Shipping invoices |
| Customer | Late-delivery penalties and chargebacks | Contracts, accounts receivable |
Lost margin and recovery usually trade places. If you make the parts up on Saturday, the margin isn't lost; you paid for the hour in overtime instead. If the plant already runs Saturdays and there's no spare shift to catch up, the margin is gone. Count one or the other for the same units, never both.
A worked example
The numbers below are illustrative, for a hypothetical plant. Swap in your own.
A stamping line that paces the plant goes down for four hours on a Thursday with a bad bearing. It runs 400 parts an hour with a six-person crew. The parts are made up on Saturday, and part of a Friday shipment goes out on Monday instead.
| Cost line | Calculation | Cost |
|---|---|---|
| Idle labor | 6 operators x 4 hours x $45 loaded rate | $1,080 |
| Repair | 2 technicians x 4 hours x $60, plus an emergency bearing shipped overnight at $1,800 | $2,280 |
| Recovery overtime | 1,600 parts at 400 an hour is 4 hours; 6 operators x 4 hours x $45 x 1.5 | $1,620 |
| Restart scrap | 120 first-off parts x $9 material | $1,080 |
| Expediting | Premium freight for the late portion of the shipment | $2,500 |
| Customer | 2% late-delivery penalty on an $80,000 order | $1,600 |
| Total | Four hours of downtime | $10,160 |
The maintenance record for this stoppage shows $2,280. The full cost is $10,160, or $2,540 an hour, more than four times what the work order says. Lost margin is zero here only because Saturday was available. In a plant already running weekends, those 1,600 parts at $6 of contribution margin each would add another $9,600 of margin that never comes back.
The costs that never get an invoice
Some of the bill never shows up in a system. The buyer who quietly moves a fifth of next year's volume to a second supplier after the third late shipment. The two jobs bumped to make room for the recovery run, now late themselves. The hours your planner and supervisor spent on the phone instead of planning and prepping for next week. The rushed restart that caused a skipped step.
You can't price these precisely, and trying to will cost you credibility with finance. List them next to the number instead, so everyone can see the whole cost.
Siemens' data shows why this matters more each year. Large plants lost an average of 27 hours a month to unplanned downtime in 2023, down from 39 in 2019. Over the same period, the cost of an hour roughly doubled in automotive and more than quadrupled in heavy industry, rising far faster than inflation. Stoppages are rarer, and each one costs more. The report points to penalty clauses that flow down the supply chain as one reason, with suppliers near the bottom of the chain facing the stiffest penalties for missed deliveries.
What to do with the number
Once you have a per-hour cost for each line, pull the last twelve months of unplanned downtime from your CMMS or MES and multiply it out. Rank the lines by dollars, not hours. The order often changes, because four hours on the constraint outweighs twenty on a machine with slack.
That ranked list tells you where an AI pilot should start, which is the subject of Article 07. The per-hour figure also becomes the denominator of any business case you'll ever write for the floor: a system that prevents 40 hours a year on that stamping line is worth about $100,000 a year before anyone argues about the rest. If half of your downtime reasons come back as "Other," the first fix is the data itself, covered in Article 02.
References
- Goldratt, E. M., & Cox, J. (1984). The goal: A process of ongoing improvement. North River Press.
- Siemens. (2024). The true cost of downtime 2024. Senseye Predictive Maintenance. https://assets.new.siemens.com/siemens/assets/api/uuid:1b43afb5-2d07-47f7-9eb7-893fe7d0bc59/TCOD-2024_original.pdf
Thought Series
The AI in Manufacturing Playbook- 01The Difference Between Monitoring and Orchestration
- 02What a Unified Namespace Does (and Doesn't Do)
- 03How to Calculate the Real Cost of Unplanned Downtime at Your Plant
-
04Coming Soon
-
05Coming Soon
-
06Coming Soon
-
07Coming Soon
Your operation is ready for orchestration.
If you're still relying on monitoring alone, you're running behind on recoverable time, margin, and customer trust. We work with manufacturers to close that gap.
Let's Talk