KPI Library / Production

Unplanned Downtime %

Formula: Unplanned downtime minutes ÷ total planned production minutes
Typical range: Under 5% is strong for mature equipment; 10%+ usually means maintenance is reactive instead of preventive

Unplanned downtime % measures the share of scheduled production time lost to breakdowns and stops that maintenance didn’t see coming, separate from changeovers, PMs, and breaks that were on the calendar to begin with. It is the number maintenance should own, because unlike planned downtime, every minute of it represents something that should have been caught earlier or fixed for good.

What good looks like

Mature, well-maintained equipment typically runs under 5% unplanned downtime against planned production time. Rates climbing past 10% usually mean the maintenance function has slipped into reactive mode, chasing breakdowns instead of running a preventive schedule that catches failures before they stop the line.

The most common way this metric gets gamed is through the planned/unplanned split itself. A breakdown that keeps recurring can get logged as a “planned” adjustment or a changeover extension instead of an unplanned stop, which improves the metric without improving the equipment. If unplanned downtime keeps trending down while planned downtime creeps up with no corresponding schedule change, that is the pattern to check first.

Unplanned Downtime % in Power BI (DAX)

With a downtime event fact carrying a type flag and reason code:

Unplanned Downtime Minutes =
CALCULATE (
    SUM ( fact_downtime_events[duration_minutes] ),
    fact_downtime_events[downtime_type] = "Unplanned"
)

Planned Production Minutes = SUM ( fact_production_hourly[planned_minutes] )

Unplanned Downtime % =
DIVIDE ( [Unplanned Downtime Minutes], [Planned Production Minutes] )

Mean Time Between Failures (min) =
DIVIDE (
    [Planned Production Minutes] - [Unplanned Downtime Minutes],
    DISTINCTCOUNT ( fact_downtime_events[event_id] )
)

Pair this with a Pareto of reason_code on the same page. The percentage tells a plant manager how much time was lost; the Pareto tells maintenance what to fix first.

Common mistakes

1. Loose discipline on the planned/unplanned flag. This metric only means something if operators and maintenance code events consistently; without that, it’s just a number people argue about.
2. Excluding short stops under a few minutes. Stops too brief to log manually are often the largest downtime category in aggregate, and skipping them undercounts the real loss.
3. Measuring against calendar time instead of planned production time. A multi-shift line will look artificially strong if downtime is measured against 24 hours a day instead of the hours it was actually scheduled to run.