Plant OEE Calculator

It's the traditional benchmark for discrete manufacturing (from 90×95×99.9 ≈ 85%), but context rules: well-run continuous process plants exceed 90% by nature of the process, while complex job shops with heavy changeovers may run excellently at 60%. The honest use of 85% is as a direction marker — most plants measuring honestly for the first time discover they're at 40–60%, which is the real news.

The one losing the most absolute hours — convert each factor's loss into hours/week and rank: a line at A=70/P=95/Q=99 bleeds availability; one at A=95/P=70/Q=99 has a speed/minor-stop problem (often the most underestimated, since 30-second jams never make the downtime log). Six-big-losses analysis under each factor turns the percentage into a to-do list.

In classic OEE, yes — changeover is downtime within scheduled production, which is exactly why SMED programs exist. Some sites compute TEEP (against all calendar hours) and others exclude planned events into a 'utilization' layer. Any convention works if it's consistent and the excluded losses are still visible somewhere — losses that vanish from metrics vanish from management attention.

Performance = (ideal cycle time × total count) ÷ run time. The honest prerequisite is a defensible ideal cycle time per product — use the design speed or best demonstrated sustained rate, not a negotiated 'standard' that bakes in losses. If P computes above 100%, your ideal cycle time is sandbagged; fix the standard rather than celebrating.