KPI Library / Production
Capacity Utilization
Formula: Actual output ÷ maximum practical capacity
Typical range: 80 to 85% is a healthy operating range for most discrete plants; sustained 95%+ leaves no slack for maintenance or demand spikes
Capacity utilization measures how much of a line’s or plant’s maximum practical output is actually being used, expressed as actual output divided by maximum practical capacity over the same period. “Practical” capacity matters here: theoretical capacity assumes zero downtime and zero changeovers, which no real plant runs at.
What good looks like
Most discrete manufacturing plants operate healthiest in the 80 to 85% range. That leaves room for scheduled maintenance, changeovers, and the inevitable unplanned event without blowing up the schedule. Utilization that sits at 95% or higher for a sustained stretch looks efficient on a slide but is actually a risk signal: there’s no slack to absorb a demand spike, a breakdown, or a quality excursion without missing a customer date.
The gaming risk runs the other direction from most KPIs. Because higher utilization looks good in a monthly report, it’s tempting to inflate the denominator’s practical capacity downward, or to count idle time as “capacity constrained by demand” rather than admit a line is underused. Keep the capacity figure tied to an engineering study, not a number that moves whenever the utilization result needs to look better.
Capacity Utilization in Power BI (DAX)
With an hourly production fact and a line dimension carrying rated throughput:
Actual Output =
SUM ( fact_production_hourly[units_produced] )
Maximum Practical Capacity =
SUMX (
fact_production_hourly,
fact_production_hourly[scheduled_hours] * RELATED ( dim_line[practical_rate_uph] )
)
Capacity Utilization % =
DIVIDE ( [Actual Output], [Maximum Practical Capacity] )
Filter scheduled_hours from a plant calendar in dim_date, not from run time
in the fact itself, or the measure quietly becomes availability instead of
utilization.
Common mistakes
- Confusing theoretical capacity with practical capacity. Theoretical assumes a machine that never stops; practical capacity should already account for planned maintenance and changeovers.
- Treating 100% as the goal. A plant chasing maximum utilization is choosing fragility over resilience, and it usually shows up later as missed ship dates.
- Not separating demand-constrained idle time from capacity-constrained idle time. A line sitting idle because there’s no order to run is a sales problem, not a capacity problem, and blending the two hides which lever actually needs pulling.