What Scrap and Rework Really Cost by Machine: From Lost Material to Lost Capacity

Scrap and rework cost by machine starts with units and disposition, then follows the actual cost incurred. Count discarded units' original first-pass material and processing, add incremental rework and disposal, and subtract recovery once. Keep defect origin separate from the point of detection. Lost constrained capacity may affect priority, but it belongs in a separate opportunity column. The machine with the highest defect rate may therefore differ from the machine with the largest financial loss.
Priority rule: compare a consistent loss ledger first, then consider whether recoverable bottleneck hours have demand behind them.
| Record | Include | Keep separate |
|---|---|---|
| Scrap expense | Incurred cost less recovery | Full selling price |
| Rework expense | Additional processing and inspection | Normal first-pass cost counted again |
| Attribution | Origin and detection fields | Assumed blame from detection location |
| Capacity opportunity | Feasible extra contribution | An automatic addition to scrap expense |
Create a unit ledger before a cost ledger
Count unique units and their final disposition before assigning dollars. A part can visit rework several times, but it should not become several discarded parts.
Create a record that distinguishes first-pass good units, units sent to rework, units recovered through rework and units ultimately discarded. Keep open quality holds separate until disposition is known. Match the production period, job and product revision used in the cost comparison.
The American Society for Quality's cost-of-quality definitions distinguish scrap from rework. Scrap is defective material that cannot be used, repaired or sold; rework corrects defective material. Those categories should remain visible even when both contribute to internal failure cost.
Our illustrative closed-month comparison starts with 10,000 first-pass units on each machine:
| Unit category | Machine A | Machine B |
|---|---|---|
| First-pass good | 9,600 | 9,800 |
| Immediately discarded | 300 | 100 |
| Sent to rework | 100 | 100 |
| Recovered as good | 90 | 95 |
| Discarded after rework | 10 | 5 |
| Final accepted units | 9,690 | 9,895 |
| Final discarded units | 310 | 105 |
The first-pass defect rates are 4% for A and 2% for B. Final scrap rates are 3.10% and 1.05%. Recovered units remain rework events, but they are not discarded units. Additional passes belong in processing records, not in new-unit counts.
Guidewheel's first-pass yield and scrap-rate guide owns the wider measurement definitions. Here, those reconciled counts become the input to a financial ranking.
Record defect origin separately from detection
Keep two machine fields: where the defect originated and where it was detected. Detection location identifies the containment point; it does not automatically establish responsibility for the defect.
A final inspection station may find a dimensional problem introduced during machining. A downstream extruder may reveal material prepared incorrectly upstream. Charging the entire loss to the detecting station would make its diligence look like poor process performance and direct corrective work at the wrong machine.
Record the job, lot, revision, first observed condition, detection timestamp and quality disposition. Add suspected origin, confirmed origin and the evidence supporting confirmation as separate fields. When origin is unknown, keep it unknown. Do not distribute uncertain losses across machines just to make a ranking look complete.
Preserve both views in the review. Quality can use detection records to improve containment and inspection. Operations can use confirmed-origin records to address the process that creates the loss. Finance can connect the units and costs without rewriting the shift's historical machine record.
In Guidewheel's RAPAC case study, changed electrical-load patterns prompted investigation. One described issue involved incorrect material mixing upstream of an extruder. The useful lesson is the sequence: observe a changed signal, investigate, then establish what happened. The signal alone did not supply the quality disposition or a complete expense ledger.
Guidewheel's public operator-dashboard description includes recording scrap and downtime reasons. Confirm the workflow needed to connect those entries with quality and job records. Additional context remains necessary to establish origin, final disposition and finance-approved costs.
Build cost without repeating material, labor or recovery
Count original first-pass expense on discarded units, then add the extra work needed to correct defects, including unsuccessful rework. Recovery value reduces the relevant loss once.
A practical loss ledger contains three components:
- Discarded-unit cost: original first-pass material and conversion on discarded units, plus disposal, less recovery. All additional rework enters the separate rework line.
- Incremental rework cost: additional material, processing, labor, inspection and handling caused by rework.
- Other evidenced failure expense: a distinct item such as investigation or customer returns, if included in the scope.
Set the ledger's costing basis with finance. Standard conversion rates may include allocated labor and overhead. Actual incremental spending answers a different question from incurred accounting cost. Show that distinction when the review moves from ranking today's losses to claiming future cash savings.
Avoid a common duplicate: valuing a scrapped part at its complete cost, then adding all of its original material and labor again. Another duplicate appears when rework cost includes original production cost already assigned to the unit. Rework should capture additional work under the chosen basis.
The example includes unsuccessful rework too. Machine A's 10 ultimately discarded rework units are included in its 310 discarded units. Their original cost enters the scrap line; their extra processing enters the rework line with all 100 reworked units. Those are different cost events, so both belong. The original production cost is not added twice.
Likewise, subtract recovered scrap proceeds once. Do not treat revenue from selling a conforming reworked unit as scrap recovery; that unit is accepted production. The worksheet uses loss expense, not the full sales value of rejected units or an assumed replacement sale.
Compare two machines whose rankings reverse
Machine A has the higher defect rate, while Machine B has the higher illustrative loss expense. The cost of the work already inside each unit changes the ranking.
Assume the discarded-unit and incremental rework inputs below apply to the closed-month unit ledger. They are invented teaching values, not market prices or a customer result. The discarded-unit material and conversion inputs cover original first-pass work only. Additional costs of successful and failed rework enter the rework line.
| Cost input | Machine A | Machine B |
|---|---|---|
| Material per discarded unit | $3.00 | $20.00 |
| First-pass conversion per discarded unit | $1.00 | $10.00 |
| Disposal per discarded unit | $0.20 | $1.00 |
| Recovery per discarded unit | $0.50 | $2.00 |
| Net discarded-unit cost | $3.70 | $29.00 |
| Incremental rework per unit sent to rework | $2.00 | $5.00 |
Machine A's scrap expense is 310 × $3.70 = $1,147. Its incremental rework is 100 × $2 = $200. Total scoped loss is $1,347.
Machine B's scrap expense is 105 × $29 = $3,045. Its incremental rework is 100 × $5 = $500. Total scoped loss is $3,545.
| Comparison | Machine A | Machine B |
|---|---|---|
| First-pass defect rate | 4% | 2% |
| Scoped quality-loss expense | $1,347 | $3,545 |
| Financial-loss priority from this ledger | Lower | Higher |
The ranking does not prove B's corrective action is easier or will recover the full expense. It identifies where the present loss deserves investigation. Check origin confidence, recurring frequency and intervention cost before funding the action.
Sensitivity matters when recovery is uncertain. If B's $2-per-unit recovery disappears, its loss rises by $210 to $3,755. A forecast should use documented recoveries and the expected disposal route, rather than preserving a favorable credit without evidence.
Keep constrained-capacity value separate from loss expense
Lost bottleneck time can influence priority, but it is a conditional opportunity. Keep it beside the expense ledger rather than adding it automatically.
The illustrative rework route consumes 3 minutes per reworked unit on A and 12 minutes on B. For 100 units each, that is 5 hours on A and 20 hours on B. These hours are assumed additional to the first-pass work and included in the illustrative rework-cost basis where applicable.
Now assume B's rework uses a genuinely constrained resource. Planning confirms demand for additional accepted output, available upstream material and downstream capacity. Finance estimates $120 of incremental contribution per recovered constraint hour after the relevant variable costs. If all 20 hours were recoverable and useful, the conditional opportunity would be 20 × $120 = $2,400.
That $2,400 is not scrap expense, realized revenue or an automatic benefit. It depends on removing the rework, converting the time into accepted output, and selling or avoiding an alternative cost for that output. If demand or downstream capacity is absent, the stated contribution opportunity is zero even though the rework hours remain real.
Do not add the $2,400 to the $3,545 and present the sum as booked loss. The expense ledger may already include labor or machine burden for those hours. A benefit case must state which cash costs would disappear and how contribution was calculated, then avoid repeating the same time value.
For unconstrained A, its 5 rework hours may be useful for scheduling or resilience. They should not receive the bottleneck contribution rate merely because B has one. The two machines can have different opportunity values without changing the underlying unit-cost arithmetic.
Rank corrective work with quality, operations and finance
Use the financial ranking to choose a specific investigation, then test whether its proposed action can remove the evidenced loss.
A short corrective-priority record makes the decision explicit:
| Review dimension | Machine A | Machine B |
|---|---|---|
| Present scoped expense | $1,347 | $3,545 |
| Rework time | 5 h | 20 h |
| Origin confidence | Quality must confirm | Quality must confirm |
| Capacity opportunity | Unvalued unless a feasible use exists | Conditional $2,400 under stated assumptions |
| Action approval | Cost, cause and expected effect required | Cost, cause and expected effect required |
Quality confirms the unique units, final disposition and origin evidence. Operations checks which hours and process conditions the action could change. Finance checks incurred-cost rates, recovery and the distinction between expense reduction and future contribution. The action owner defines a bounded intervention and the records needed to judge its effect.
For subsequent verification, compare comparable products and volumes, not only total dollars from different production mixes. A lower scrap bill in a low-volume month can coexist with a worse defect rate. Keep counts, cost per discarded unit and rework passes available so the review can explain the change.
Guidewheel is the practical operating platform for manufacturing. Machine information, operator context and optional integrations can support a shared review; its AI-powered software and manufacturing experts help teams decide where to act next. Confirm the application-specific sensing and quality-record path. The unit ledger, financial costing and attribution method here are recommended review tools, not an assertion that a current signal calculates complete quality loss or diagnoses every defect.
Choose one loss with a confirmed origin and a testable action. A precise, modest intervention with a reproducible ledger is easier to defend than a large saving built from selling prices, duplicate costs and assumed capacity value.
Make the next scoped decision
Bring the reconciled unit ledger and one confirmed-origin loss to a quality, operations and finance review. Agree the cost basis, the proposed action and the evidence needed to judge its effect. To explore how machine information and expert support can help that review, discuss your operation with Guidewheel. Fund the intervention from a defensible loss and feasible opportunity, then verify the result without duplicate credit.
Frequently asked questions
Should scrap be valued at the product selling price?
Use the chosen incurred-cost basis for discarded units, with disposal and recovery treated consistently. Full selling price includes an assumed commercial outcome and is not automatically scrap expense. A separate lost-contribution estimate needs its own feasibility evidence.
Do recovered rework units still count as scrap?
No. Recovered units are accepted output with incremental rework cost. Units ultimately discarded after rework enter the scrap count once, while their additional processing remains a distinct cost event.
Which machine owns a defect found downstream?
Keep detection and origin in separate fields. Charge corrective responsibility to a confirmed origin where the evidence supports it, and retain unknown origins explicitly. The detecting station may have performed effective containment.
Can lost capacity be added to scrap cost?
Show constrained-capacity opportunity separately and test demand, route feasibility and contribution after relevant variable costs. Do not add an assumed sales value or duplicate time already included in the expense ledger.
Does machine monitoring calculate the full quality-loss ledger?
Guidewheel can support machine visibility, operator context and suitable system connections. Complete quality costing still needs validated units, disposition, origin and finance rates. Confirm the application workflow rather than assuming a current signal supplies those records.