What the Equipment Downtime Calculator does
This calculator turns a list of stoppages into the four numbers a plant meeting actually argues about: total downtime, availability against planned running time, lost output in units, and what the stoppage cost. It also sorts the causes into a Pareto so the biggest one is obvious rather than the loudest one.
Availability here is run time divided by planned production time, the definition ISO 22400-2 uses. That matters, because measuring the same machine against the calendar instead gives a far lower number for reasons that have nothing to do with maintenance - the night shift nobody staffs is not downtime.
It is arithmetic on the events you enter, done in your browser. Nothing is sent anywhere.
How to use it
- Enter the planned running time in minutes: the shift or period the asset was meant to run for, with breaks and unstaffed hours already taken out.
- Add the output rate when the machine is running, the margin lost on each unit not made, and any standing cost per hour that carries on during a stoppage.
- List the stoppages as cause, minutes and (optionally) how many times it happened. Repeat a cause on several lines and they are merged in the Pareto.
- Read the Pareto: the share column shows each cause's portion of the lost time, and the cumulative column shows where the 80% line falls.
- Export the CSV for the meeting pack, or copy the summary for an email.
Reading the results
Availability is what the asset achieved against what was asked of it. Eighty per cent is a bad shift on most packaging lines and an excellent one on a mining shovel - the figure only means something against the same asset's own history.
The Pareto is the useful part. A cause that appears once for forty-five minutes and a cause that appears thirty times for three minutes look completely different in a breakdown log and are often the same total loss.
Short frequent stops are flagged separately because they rarely reach a maintenance system at all: nobody raises a work order for a ninety-second jam. Counting them usually moves the biggest loss somewhere unexpected.
The cost figure is only as good as your margin and standing-cost numbers. Where a line is not the bottleneck, lost output may cost nothing at all because the downstream process catches up - the standing cost is real either way.
Worked example: one eight-hour shift with three causes
A filler is planned to run 480 minutes. It stops for a bearing failure (45 minutes), two changeovers (30 minutes in total) and three material jams (15 minutes in total): 90 minutes of downtime from six individual stoppages.
Run time is 480 - 90 = 390 minutes, so availability is 390 / 480 = 81.25%.
The line makes 120 units an hour, so 90 minutes - 1.5 hours - is 180 units not made. At 12 of margin each that is 2,160. The standing cost of 400 an hour adds 600, giving 2,760 for the shift.
In the Pareto, the bearing failure is 45 of the 90 minutes, exactly half. Changeover is 30 minutes, taking the cumulative share to 83%, so those two causes are the 80% line. Material jams are only 15 minutes - but they happened three times, averaging five minutes each, which is the kind of loss that never gets written down and quietly repeats every shift.
Formulas and scoring rules
- Availability
availability = (planned - downtime) / plannedISO 22400-2 measures this against planned production time, not calendar time.- Lost output
lostUnits = downtime hours x units per hour- Cost
cost = lostUnits x margin per unit + downtime hours x standing cost per hour- Pareto share
share = cause minutes / total downtime minutesThe cumulative column adds them in descending order.- Mean stoppage length
mean = cause minutes / occurrencesThis is close to MTTR for that cause, if the occurrences are genuine separate events.
Planned time, calendar time, and why the number moves
The same machine can honestly be reported at 95% or at 32% availability depending on what sits in the denominator. Against planned production time - the hours it was scheduled to run - maintenance owns the number. Against calendar time, the denominator includes every unstaffed night and weekend, and the figure is really about how the plant is loaded.
Neither is wrong, but they answer different questions and they belong to different people. This page uses planned production time and says so on every result. If your board pack uses the calendar figure, TEEP on the OEE calculator is the one that matches it.
Counting what nobody logs
Most downtime systems capture events above some threshold, typically five or ten minutes, because that is when an operator has to acknowledge a stop. Everything shorter falls through: the jam cleared in ninety seconds, the sensor wiped clean, the pallet straightened.
Those losses are real and they are usually large. If a line records 8% downtime from logged events but achieves 70% of its rated output, the missing 20-odd per cent is mostly minor stops and reduced speed. Getting them into this calculator, even as an estimate from a time study, changes which cause comes top.
Limitations: what the result does not prove
- It adds up the events you enter. It cannot know about stoppages nobody recorded, and on most lines those are the majority of the count if not the majority of the minutes.
- The cost is a simple rate model. It ignores whether the line is the bottleneck, whether the output was recovered later, overtime, scrap at restart and penalties for late delivery.
- Availability here excludes the speed and quality losses that also cost output. For the full picture use the OEE calculator, where a machine that ran all shift at half speed shows up.
- A Pareto describes the period you measured. One large failure can dominate a single shift and say nothing about the pattern; a month of data is the minimum for a decision.
Privacy: where your data goes
Everything you paste, type or drop is processed in this browser tab. It is not uploaded, logged, stored or sent to analytics. Session recording and tag-manager scripts are switched off on this page.
Standards and sources
- ISO 22400-2 - Key performance indicators for manufacturing operations - checked 19 Sep 2026
- IEC 60300-3-11 - Dependability management: reliability centred maintenance
- IEC 60812 - Failure modes and effects analysis (FMEA and FMECA)
Frequently asked questions
How do I calculate equipment availability?
Divide run time by planned production time: (planned - downtime) / planned. A 480-minute shift with 90 minutes of stoppages gives 390 / 480 = 81.25%. The important part is agreeing what goes into planned time - breaks, unstaffed shifts and planned maintenance are normally excluded.
What does downtime actually cost?
Two things: the margin on the units that were not made, and the costs that carry on regardless - labour standing idle, energy, overhead. This page adds both. What it cannot know is whether the output was recovered later, which on a non-bottleneck machine it usually is.
Should planned maintenance count as downtime?
Usually it is taken out of planned production time instead, so it does not count against availability. What matters is consistency: if you exclude it, exclude it every period, and keep a separate figure for how much planned time it is consuming.
What is a Pareto analysis of downtime?
Sorting causes by lost time, largest first, and adding them up as you go. The point is the cumulative column: typically a small number of causes make up most of the loss, so fixing those two or three is worth more than a general improvement drive across all of them.
Why is my availability high but output still low?
Because availability only counts stopping. A machine running all shift at 70% of its rated speed, or making 5% scrap, shows perfect availability and poor output. That is precisely the gap OEE was invented to close.
How do I count a stoppage that spans a break?
Only the part that overlaps planned running time is downtime, because the break was never planned production. Splitting the event at the break boundary keeps both the downtime figure and the planned time honest.
What is a good availability figure?
There is no universal number. A continuous process plant may need 97%; a packaging line with frequent changeovers may plan for 85%; a mobile plant item may be excellent at 70%. The useful comparison is the same asset against its own history and against the loss you can actually remove.
How many periods of data do I need before acting?
Enough that one big failure does not dominate. A single shift tells you what happened that shift; a month usually shows the repeating pattern, which is what maintenance planning is for. Keep the raw events rather than the monthly totals, or you lose the ability to re-cut them by cause later.
Last reviewed by the A2Z.Tools team against the sources listed above.
