Production Changeover First-Run Scrap Rate

A line switches from one product to another, and initial units may need adjustment before output meets specification, so production changeover first-run scrap rate measures material rejected as scrap during a declared startup window after each changeover. Define changeover, since a new item, recipe, die or format can trigger one while routine pauses do not necessarily count, and define the window, because the first ten units, first hour or production until validated first-good output each produce different rates.

State the rule. The Lean Enterprise Institute describes changeover as switching products or part numbers and measures changeover time from the prior last piece to the new first good piece, whereas scrap rate examines material lost in the selected startup window.

APQC describes finished-product first-pass quality yield, which tests conformance at a final quality point, while a startup scrap rate isolates one phase and one disposition, not all defects or rework. Separate changeover time from scrap, because a faster first-good result can coexist with more wasted material, so both measures should be reported.

Choose the denominator, since units started, material input by weight and equivalent output can differ, and use a unit suitable for the process. Check scrap classification, because some output is reworkable or saleable as a lower grade and should not be called final scrap without the actual disposition, and capture material cost, since a unit rate and a cost rate can move differently if the startup consumes expensive ingredients or components.

Count trials by giving quality test pieces intentionally consumed during setup an explicit inclusion rule, and treat safety holds correctly, because material awaiting test is not yet scrap and a rate built before all startup rejects are reviewed may understate eventual scrap. Keep traceability by linking rejected units to the changeover run, machine, product and lot to avoid assigning later defects to startup, and exclude unrelated loss, since a late damaged pallet should not be assigned to changeover simply because it appeared that day.

Use time stamps, because a delayed posting can still belong to the earlier startup window if its material identity is established, and watch sequential changes, defining how the window closes or restarts if another changeover occurs during an unstable run. Protect quality, since declaring a unit good early just to end the window can shift defects to customers, and remember that the first conforming piece may not prove stability, so the approved validation protocol should be used.

Control comparability, because products with different material characteristics should not be ranked on a pooled raw rate alone, and compare recipes carefully, since changes to input specification or customer quality limits can change what qualifies as scrap. Review small runs, as one rejected unit in a five-unit trial creates a large percentage, so show quantities and context, and avoid pooling zeros, because changeovers with no started units cannot enter a unit-based denominator and should be explained separately.

Investigate causes such as tool setup, cleaning, calibration and operator handoff, and track specific defects, since wrong dimensions, colour, fill weight and contamination require different responses. Inspect startup rejects by defect type and disposition, because a low scrap figure can hide repeated rework or material downgraded for use elsewhere.

Use the metric to guide setup improvements, not to excuse weaker testing or premature release.