David Krause1 · The base formula — and the mistake almost everyone makes
Anyone who knows a scrap rate of 3 % intuitively adds 3 % on top. That is wrong. The question is not how much of the input is lost, but how much input is needed for one good part.
at q = 3 %: 1 / 0.97 = 1.0309 blanks per good part
→ correct surcharge 3.09 %, not 3.00 %
| scrap rate | correct surcharge | naive add-on | error |
|---|---|---|---|
| 2 % | 2.04 % | 2.00 % | −0.04 pp |
| 3 % | 3.09 % | 3.00 % | −0.09 pp |
| 5 % | 5.26 % | 5.00 % | −0.26 pp |
| 10 % | 11.11 % | 10.00 % | −1.11 pp |
Titanium housing, material input €92.65 per blank, scrap rate 10 %.
| calculation path | surcharge | material cost per good part |
|---|---|---|
| naive: +10 % | 10.00 % | €101.92 |
| correct: 1/(1−q) | 11.11 % | €102.94 |
Only €1.02 per part — but at a lot size of 500 that is €510 per order that no one costed, and the effect grows more than proportionally with the rate. The correct formula is above; the rest of the article shows where the same logic applies elsewhere.
2 · When scrap is detected decides how much it costs
A part identified as scrap after roughing has caused fewer costs than one that only fails at final inspection.
| moment of detection | loss per part | per good part |
|---|---|---|
| after roughing (30 % of machining) | €104.44 | +€3.23 |
| after finishing (100 %) | €131.94 | +€4.08 |
| after final inspection (incl. admin/selling overhead) | €155.43 | +€4.81 |
Between early and late detection lies €1.58 per good part — about a 49 % difference in the scrap surcharge. This is at the same time the answer to whether an intermediate inspection after roughing pays off: it may cost up to €1.58 per good part before it stops being worthwhile.
3 · Rework is not the same as scrap
With rework the part survives — it costs additional time, but no additional material. That is why it does not belong in the scrap surcharge, but as a separate item in the manufacturing costs.
| rate | time per event | cost per event | per good part |
|---|---|---|---|
| 5 % | 8 min | €11.95 | +€0.60 |
| 10 % | 12 min | €17.92 | +€1.79 |
| 15 % | 15 min | €22.40 | +€3.36 |
Scrap of 3 % costs €4.08 per good part, rework of 10 % only €1.79. The rework rate is three times as high and stands out more in meetings — but economically scrap is the item twice as large. Anyone who prioritizes improvement projects by how conspicuous they are rather than by cost is working at the wrong end.
4 · Building it into the calculation
| item | without | with |
|---|---|---|
| material cost | €92.65 | €92.65 |
| manufacturing cost | €39.29 | €39.29 |
| + rework (10 % × €17.92) | — | +€1.79 |
| = cost of production before scrap | €131.94 | €133.73 |
| + scrap surcharge 3.09 % | — | +€4.14 |
| = cost of production | €131.94 | €137.87 |
| + admin/selling overhead 17.8 % | €23.49 | €24.54 |
| + special direct selling costs | €3.50 | €3.50 |
| = total cost | €158.93 | €165.91 |
At a profit markup of 15 %, un-costed scrap of 3 % together with 10 % rework eats up about 29 % of the planned profit. The order does not become unprofitable — but it earns considerably less than expected, and without post-calculation no one knows why.
5 · What does not belong in the surcharge
| case | treatment |
|---|---|
| start-up scrap on first orders | separate start-up surcharge, itemized in the quotation |
| scrap from a faulty drawing | customer responsibility, billed separately |
| scrap from defective material | supplier recourse |
| systematic scrap above 8–10 % | process problem — fix first, then cost |
From about 8 to 10 % scrap, costing stops being the right instrument. Pricing in such a value means financing a process problem permanently — and raising the price so far that the order is lost anyway. At that magnitude it is manufacturing that is called for, not the estimator.
