Unplanned downtime % = unplanned stop minutes / planned production time x 100. Count every stop inside planned time that was not on the schedule, plus the part of any planned stop that ran over. Example: 370 unplanned minutes in 4,500 planned minutes is 8.2%.
The formula, with its neighbours
Planned production time is the time the line was meant to make product. Time with no orders, holidays and shutdowns set well ahead stay out of it.
This page covers the unplanned slice only. For the total of all stops, see how to calculate downtime percentage.
Which minutes go on top of the fraction
The rule most plants can hold to is simple: was this stop on the schedule before the shift started? If not, it is unplanned.
- Count: breakdowns, jams above your stop threshold, waiting for material or a crew, quality holds.
- Count the overrun: a changeover planned at 45 minutes that takes 60 adds 15 unplanned minutes.
- Do not count: a changeover within its plan, scheduled cleaning, or excluded time.
- Do not count: stops below your threshold, which land in performance. See stop threshold.
The longer classification is on planned vs unplanned downtime, and examples are on unplanned downtime examples.
Worked calculation: one week on one line
Example: a line runs 5 days on 2 shifts. Each shift is 450 planned minutes, so planned production time is 5 x 2 x 450 = 4,500 minutes.
| Stop | Events | Minutes | Type |
|---|---|---|---|
| Breakdowns | 4 | 140 | Unplanned |
| Jams over the threshold | 9 | 65 | Unplanned |
| No material | 5 | 95 | Unplanned |
| Quality holds | 2 | 40 | Unplanned |
| Changeover overrun | 1 | 30 | Unplanned |
| Unplanned total | 21 | 370 | |
| Changeovers, to plan | 6 | 180 | Planned |
| Scheduled cleaning | 5 | 60 | Planned |
| Planned total | 11 | 240 |
Unplanned downtime % = 370 / 4,500 x 100 = 8.2%. Planned stops are 240 / 4,500 = 5.3%.
Together they are 610 / 4,500 = 13.6%, so availability is 100 - 13.6 = 86.4%. That availability is the first factor of OEE.
A KPI set instead of one number
One percentage cannot tell you whether you have many small stops or a few long ones. Report these together, from the same week.
| KPI | Calculation | Result | What it tells you |
|---|---|---|---|
| Unplanned downtime % | 370 / 4,500 | 8.2% | How much planned time you lost |
| Unplanned share | 370 / 610 | 60.7% | How much of your stop time was a surprise |
| Stops per 100 run hours | 21 / 64.8 x 100 | 32.4 | How often the line gets interrupted |
| Mean unplanned stop | 370 / 21 | 17.6 min | How long each surprise lasts |
| Equipment only % | 205 / 4,500 | 4.6% | The part that breakdowns and jams take |
The last row stops a common fight. A technical team that owns breakdowns and jams should be judged on 4.6%, and the plant on 8.2%.
Turn the percentage into output and money
Example: the line's ideal rate is 50 units a minute. At that rate, 370 minutes is up to 18,500 units of lost output in the week.
Example: at a cost of 40 euros a minute, 370 minutes is 14,800 euros for the week. The 40 is an example, so work out your own with the downtime cost calculator.
Then check OEE. With performance at 92% and quality at 98%, availability of 86.4% gives 0.864 x 0.92 x 0.98 = 77.9% OEE.
Cut unplanned downtime by half, to 185 minutes. Availability becomes 90.6%, and OEE becomes 81.6%, so the prize is 3.7 points.
How to calculate it in a spreadsheet
Keep one row per stop, with a column for minutes and a column for type. Keep planned minutes per shift on a second sheet.
=SUMIFS(Minutes,Type,"Unplanned")/PlannedMinutes gives the KPI. Add a week column to the SUMIFS when you want a trend.
For weeks and lines, add minutes first and divide after. Averaging percentages gives too much weight to short weeks.
The free downtime log template already has the stop rows. To collect the stops, see how to track downtime.
Is your number high or low?
We can point to one neutral reference. In a NIST survey of US discrete manufacturers (2016 data), unplanned downtime took 7.8% of planned production time.
In the same report, the quarter of plants that relied most on reactive maintenance lost 13.0%, and the quarter that relied least lost 4.0%. It is a survey of one US sector, so treat it as context, not a target.
Your own trend is the better test. To turn it into a goal, see how to set a downtime target, and for more sourced figures see manufacturing downtime statistics.
Errors that change the result
- Dividing by 24 hours when the line is scheduled for fewer. The KPI looks better and means nothing.
- Counting the whole overrun changeover as planned, which hides the slip.
- Comparing plants with different stop thresholds. A hand log that misses short stops reads lower.
- Reclassifying a stop as planned after it happened. The rule is what was on the schedule before the shift.
This formula is only as good as its top number. In Fabrico, stops come live from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, not from a hand log, and the operator adds the cause at the line. See how Fabrico tracks unplanned stops.
What people ask next
What is the formula for unplanned downtime?
Unplanned downtime % = unplanned stop minutes / planned production time x 100. Use the same period for both numbers. You can check the total with the availability calculator.
Is unplanned downtime the same as breakdown time?
No. Breakdowns are one part of it. Waiting for material, quality holds and overruns are also unplanned, as these examples show.
Should I divide by planned production time or by total time?
Use planned production time, because it fits availability and OEE. Total calendar time answers a different question, which belongs with TEEP. See OEE vs TEEP.
How do I link unplanned downtime to MTBF and MTTR?
For failures only, the downtime share is about MTTR / (MTBF + MTTR). The MTBF and MTTR calculator works it out from your data. Other unplanned stops, such as waiting for material, add to it.
How often should I report unplanned downtime?
Review it daily at the line and report it weekly to the plant. Use the same formula and threshold every time. A daily production meeting is a good place for it.
Fabrico is a light MES for OEE. It takes stops, speed and counts from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line.