Guide · Updated 29 October 2026

How to calculate OEE on a food production line

OEE (overall equipment effectiveness) tells you how much of the time a line was scheduled to run actually turned into good product, at the right speed. This guide walks New Zealand and Australian food and manufacturing plants through the calculation step by step, the traps that make OEE numbers meaningless, and how to turn the result into a short list of fixes.

The formula in one line

OEE = Availability × Performance × Quality

• Availability = run time ÷ planned production time. It captures stops: breakdowns, waiting, and changeovers or cleans that run past plan.
• Performance = (ideal cycle time × total units made) ÷ run time. Put simply, actual speed against the line's rated speed while it was running. It captures slow running and short stops nobody logged.
• Quality = good units ÷ total units made. It captures scrap, rework and start-up rejects.

Each factor is a percentage, and multiplying them gives one figure that shows how much of the scheduled time became saleable product.

Step by step

1. Set the planned production time. Shift length less any time the line genuinely wasn't meant to run, such as scheduled breaks with no relief crew or no orders. Be strict: if a clean or changeover is planned, keep it inside planned time so its overrun shows up.
2. Record the stop time and subtract it to get run time.
3. Use the rated speed, not the "usual" speed. The ideal rate is what the bottleneck machine is designed or validated to run on that product. Using a slowed-down "normal" rate hides the loss.
4. Count total and good units from the same point in the line, usually the end of the pack line, so the numbers match.
5. Multiply the three factors.

A worked example

Example data (not from a real plant): a yoghurt cup filling line on an 8-hour shift (480 minutes), with a 30-minute planned break and no relief crew, so planned production time is 450 minutes. The line stops for 45 minutes (a 25-minute changeover overrun and 20 minutes of jams). The rated speed is 100 cups a minute. It makes 36,000 cups, of which 34,560 pass.

• Run time: 450 − 45 = 405 minutes, so Availability = 405 ÷ 450 = 90%
• Performance: 36,000 ÷ (405 × 100) = 36,000 ÷ 40,500 = 88.9%
• Quality: 34,560 ÷ 36,000 = 96%
• OEE: 0.90 × 0.889 × 0.96 = about 76.8%

Each factor on its own looks respectable, yet roughly a quarter of the planned time didn't turn into good cups. A check that helps here: good units ÷ (planned time × rated speed) = 34,560 ÷ 45,000 = 76.8%, the same answer.

Common traps

• Moving the goalposts. Lowering the rated speed or calling overruns "planned" makes OEE look better without making more product.
• Averaging percentages across lines. Combine the minutes and units first, then calculate, or a small line can skew the site figure.
• Chasing a benchmark. Figures like "85% is world class" are widely quoted, but they depend on how each plant defines planned time and rated speed. Your own trend week to week is more useful than someone else's number.
• Stopping at the percentage. OEE says where the loss is, not what it costs. A point of quality on a high-value product can be worth more than several points of availability on a cheap one.

From OEE to action

1. Find the weakest of the three factors.
2. List the losses inside it (stop reasons, slow-running causes, reject reasons) and put a dollar value on each.
3. Fix the top one or two with the people who run the line, then write the fix into standard work.
4. Recalculate weekly with the same definitions so the gain is real and holds.

Try it on your own line

The free OEE calculator does the maths above from your shift figures in under a minute. To turn the losses into dollars and get your top three fixes on one page, use the Line Loss Finder (NZD 49). If changeovers are eating your availability, the Changeover Planner (NZD 49) builds a step-by-step plan to shorten them, and the cost of downtime guide shows how to price each stop.

This guide is general information. The example uses made-up numbers to show the method.