E•ENGINEERING
DESK TOOLS
All guides
Maintenance5 min readUpdated 5 October 2026

Production downtime cost: contribution versus revenue

Estimate event costs without counting recovered production or avoidable costs twice.

What an event cost estimate is for

A downtime cost figure supports decisions: whether a repair, spare part or inspection task is justified, and which stops deserve root-cause work. The production downtime cost calculator estimates the cost of one stop. It is an event estimate, not a forecast of annual downtime.

The central choice is the basis for lost output. Revenue overstates the loss, because the materials and other avoidable variable costs of units that were not made were never spent. The calculator therefore asks for contribution margin: selling price less avoidable variable cost per unit.

The estimate is deliberately simple: it takes eight inputs and returns one total. That simplicity is a strength when stops need to be valued quickly, and a limit when a stop has unusual features such as contractual penalties or lost customers. Those items are outside the calculation; if they apply, keep them in a separate note so that the contribution-based figure remains comparable between events.

The formulas

  • Unrecovered units = downtime hours × production rate × (1 − recovered % ÷ 100)
  • Lost contribution = unrecovered units × contribution margin per unit
  • Additional costs = repair labour + parts + scrap and restart + additional recovery cost
  • Total impact = lost contribution + additional costs
  • Cost per downtime hour = total impact ÷ downtime hours

The recovered percentage reduces lost contribution only. Making up output later usually costs money, such as overtime, so that goes in the separate recovery field rather than being netted off.

Worked example 1: a short stop with half recovered

A two-hour stop on a line running 100 units per hour, contribution of 3 per unit, 50% recovered later, repair labour 50, parts 100, scrap and restart 20, recovery cost 30.

  1. Initial lost output = 2 × 100 = 200 units.
  2. Unrecovered = 200 × (1 − 0.50) = 100 units.
  3. Lost contribution = 100 × 3 = 300.
  4. Additional costs = 50 + 100 + 20 + 30 = 200.
  5. Total impact = 300 + 200 = 500, or 250 per downtime hour.

The calculator gives these same figures.

Worked example 2: a longer stop on a higher-value line

A 6.5-hour stop on a line at 240 units per hour, contribution of 4.75 per unit, 25% recovered, labour 420, parts 1,350, scrap and restart 600, recovery cost 380.

ItemCalculationResult
Initial output lost6.5 × 2401,560 units
Unrecovered units1,560 × 0.751,170 units
Lost contribution1,170 × 4.755,557.50
Additional costs420 + 1,350 + 600 + 3802,750.00
Total impact5,557.50 + 2,7508,307.50

Cost per downtime hour is 8,307.50 ÷ 6.5 = 1,278.08. The calculator returns the same values.

Why revenue and recovery change the answer

Using example 2 and changing one input at a time shows how sensitive the estimate is.

VariationLost contributionTotal impact
Contribution 4.75 (as above)5,557.508,307.50
A selling price of 12 entered as the margin14,040.0016,790.00
100% recovered, same extra costs0.002,750.00

The middle row is only an illustration of the error: entering a price of 12 as the margin more than doubles the impact. The last row shows that full recovery does not make the event free, because the additional costs remain.

Checks before you trust the result

  • Downtime hours match the period when the chosen line or machine was actually unable to produce, from a logged start and restoration time.
  • The production rate is the rate for the stopped period, not a brochure speed. Compare it with OEE data; see OEE data checks.
  • Contribution margin comes from the finance team or a documented source, and excludes costs that were not avoided.
  • Recovered production is real: buffers, spare capacity or a later shift made up the units, and the output was sold rather than stocked unsold.
  • No cost is counted in both the margin and the extra cost fields.
  • Regular wages and fixed overheads are not entered unless the stop created an additional expense.
  • All amounts are in one currency.

A final reasonableness test is to divide the total by the downtime hours and compare the result with earlier events on the same line. A cost per hour that is far from the usual range usually points to a data entry error, such as a unit mix-up or a margin entered as a price, rather than a genuinely unusual event.

Common mistakes and FAQ

  • Multiplying lost units by selling price.
  • Ignoring recovery, or counting recovered units as lost while also paying overtime.
  • Treating machine downtime as line downtime where a buffer absorbs the stop.
  • Using an event estimate to forecast annual cost without checking how often similar stops occur.
  • Changing definitions between periods, so costs cannot be compared.

Should planned stops be included?

Only if the stop displaced production that would otherwise have been sold. Keep planned and unplanned stops in separate categories.

Can the cost per hour be used for another stop?

Only as a rough guide. It mixes fixed repair costs with a time-driven contribution loss, so a different duration changes the ratio.

Where does availability fit?

Availability and MTBF from the MTBF, MTTR and availability calculator show how often stops occur; this calculator values one event.

Choosing the scope of a stop

The same physical failure can carry different costs depending on where the boundary is drawn. If a machine stop is absorbed by a buffer or an alternative route, line output may not fall at all, and the lost contribution is close to zero while repair costs remain. If the machine is the bottleneck, every hour lost on it is an hour lost for the whole line. A practical approach is to identify the bottleneck for the product being made and apply the production rate of that step, noting any buffer that delayed the effect.

Partial stops are another case. A machine running at reduced speed after a repair has not stopped, so entering the full duration as downtime overstates the loss. Either enter the equivalent full-stop hours or record the slow running separately as a performance loss. The same idea applies to start-up scrap after a restart: it belongs in the scrap and restart field, not in the downtime hours.

Finally, state whether the figure covers one event or a period. Summing event costs over a year gives a total, but it only equals a forecast if the mix of stops is expected to repeat.

Recording the data

Record start and restoration times, line or asset, failure type, repair action, labour hours, parts cost, scrap, units recovered and who approved the margin used. Consistent records let the event cost be repeated for each stop and summed later. The free breakdown and downtime log template fits: it holds downtime, MTTR and cost per failure. Read MTBF, MTTR and downtime records for how to define the start and end of a stop.

Try it with the calculators

Free Excel templateBreakdown & downtime log: blank records, dropdowns and a live dashboard.
Download .xlsx

Browse all Excel templates and dashboard previews

References