Scrap Rate in Manufacturing: Calculate the Real Loss

Home » Resources » Scrap Rate in Manufacturing: Calculate the Real Loss

Estimated reading time: 10 minutes

The batch is finished, but the order is not. Inspection has rejected parts that cannot be recovered, the material has already been consumed, and the customer still expects the full quantity on Friday.

Scrap rate in manufacturing is the percentage of a defined production quantity that is permanently rejected rather than accepted for its intended use. For a completed group of comparable units, calculate it as:

Scrap rate (%) = Units finally scrapped ÷ Total units in the same production group × 100

The denominator includes both accepted and scrapped units. Parts successfully recovered through rework are accepted output, not final scrap.

A trustworthy percentage helps you find a recurring loss. Responding to that loss also means correcting the material picture and finding room for any replacement work. SkyPlanner supports scrap reporting through faulty-piece records and accounts for their material consumption when material tracking is configured. Its AI scheduling then helps you fit approved replacement operations into the same plan as the orders already promised.

Calculate scrap rate from final outcomes, not every failed inspection

Imagine a batch of 1,000 machined parts. You follow the original parts through inspection and any permitted rework until every part has a final outcome:

Outcome for the original batch	Parts	Treatment in final scrap rate
Accepted on the first attempt	920	Accepted output
Accepted after rework	30	Accepted output, with extra work recorded separately
Finally scrapped	50	Scrap numerator
Total original units	1,000	Denominator

The final scrap rate is 50 ÷ 1,000 × 100 = 5%. There are 950 accepted parts, so final yield is 95%.

First pass yield answers a different question. Only the 920 parts accepted without correction passed first time, giving 92% FPY. The remaining 8% was not all scrap: 30 parts were recovered. Rework consumes time, but it does not necessarily consume another complete blank or destroy the original part.

Likewise, dividing the 50 rejects by the 950 accepted parts gives approximately 5.26%, not the 5% share of original production lost. A scrap-to-good-output ratio can be useful for a separate costing purpose, but label it as such. Do not compare it with a scrap-to-total-production percentage.

Once replacements are made, keep their production identifiable. They are new units, not extra inspection attempts on the original batch. If you report a combined result that includes them, include both their accepted output and any additional scrap in that same population.

Set counting rules your team can apply consistently

A clear measurement definition prevents the number from changing merely because one shift records defects differently from another. Agree the product group, process boundary, reporting basis and final-disposition rule before interpreting a trend.

Match the rejects to the production that created them

Start with a traceable batch or order quantity whose final outcomes can be reconciled. Dividing today’s scrap decisions by today’s new starts can mix different groups when inspection or rework takes several days.

For regular period reporting, define how late decisions are attributed and how open quantities are shown. If 20 parts are still awaiting a disposition decision, they are neither confirmed good output nor final scrap. Show them as unresolved and finalize the result when the group is closed. Do not quietly remove them to make the percentage look better.

Count each original unit once at the chosen boundary. A part rejected at machining and mentioned again in the final inspection report is not two scrapped parts. Several defects on one part are also one lost unit, although each defect may matter to the investigation.

Operation-level and whole-route results need separate definitions. A machining scrap rate uses the units entering that operation; an end-to-end rate follows the original production population across its route. Adding the operation percentages does not produce a reliable factory-wide scrap rate.

Separate final scrap, rework and material waste

A failed inspection is a reason to decide what happens next. The part may be reworked, held for investigation or finally scrapped. Preserve that distinction in the records feeding the KPI.

Also distinguish rejected products from expected process waste. Machining chips, cutting trim and startup material may belong in a material-loss measure even when every finished part passes inspection. Mixing them into a piece-based reject count makes the result difficult to interpret.

If you measure material scrap by weight, use weight on both sides of the calculation. For example, imagine a closed cutting run with 500 kg of material input, 40 kg classified as scrap and the remaining 460 kg accounted for as useful output. The material scrap rate is 40 ÷ 500 × 100 = 8%. State whether that scrap includes normal trim, abnormal rejects or both, and account separately for reusable remnants or material still in process.

That 8% is not directly comparable with the batch’s 5% rejected-piece rate. They describe different losses. A heavy rejected casting can matter far more to material consumption than a small rejected fitting.

Aggregate quantities, not unrelated percentages

For comparable products measured on the same basis, add the scrapped quantities and total quantities before dividing. Do not take a simple average of batch percentages when batch sizes differ.

Keep different products visible when their material value, routes or consequences differ. A factory average can fall simply because more of an easy product was made. Read the rate alongside the underlying quantities, product mix and cost of loss before concluding that the process improved.

Put a value on what was lost

The scrap rate describes quantity, not the whole financial consequence. A part rejected after several operations has consumed more work than one rejected before machining starts.

Return to the imagined 50 scrapped parts. Suppose each contains $12 of material and has accumulated $8 of conversion cost at rejection. The material loss is $600, and the accumulated conversion cost is $400. If recovery of the scrap material yields $2 per rejected part, the net recorded loss on this simplified basis is:

50 × ($12 + $8 − $2) = $900

Use your own costing rules for the costs accumulated at the rejection point. Recycling value can reduce the net financial loss, but it does not turn the rejected parts back into deliverable products or restore their consumed production time.

Keep the historical loss separate from the forward cost of recovery. Making replacement parts needs additional material and operations. Do not count that future work twice by including it in both a fully loaded loss figure and a separate recovery estimate without explaining the boundary.

For the immediate decision, identify what changes: extra material purchases, overtime, outside processing or expedited transport, for example. Freed machine time can be valuable even when it does not immediately remove a wage or equipment payment. Treating every avoided scrap dollar as an immediate cash saving overstates the result.

Turn the shortfall into a production plan that can still deliver

In the example, 1,000 original units have produced only 950 accepted parts. If the customer needs 1,000 and no suitable stock is available, the remaining requirement is 50 good parts. That is a delivery decision as well as a quality result.

First, establish which material is genuinely available for that work. Then approve the replacement route, quantity and required inspections. Use realistic operation durations and account for setups. If further yield losses are expected, make the production allowance explicit rather than assuming every replacement will pass.

Imagine that the approved replacement run needs six machining minutes per part, with no further rejects assumed for this illustration. The 50 parts require 300 machining minutes, or five hours, before additional setup and inspection time. If that machine has only three free hours before the next operation must begin, the replacement run does not fit there as planned.

Putting it in an unofficial urgent-work queue hides the conflict; it does not create the missing capacity. You need to decide whether another eligible machine can help, extra capacity is justified, another order can move or a delivery commitment must change.

Keep material consumption and replacement work connected in SkyPlanner

SkyPlanner lets operators report scrap as faulty pieces when stopping the job timer. With material tracking configured, consumption includes the material used by those faulty units and updates stock balances. That helps prevent the next job from relying on blanks that have already been used to make rejected parts.

For example, if ten blanks are consumed to obtain eight good pieces because two pieces are scrapped, the material record needs to reflect ten blanks, not eight. In this situation, the two faulty pieces represent confirmed final rejects. Keep repairable parts and unresolved quality holds distinguishable when reconciling a final scrap KPI; a generic failed-quality quantity alone does not establish the final outcome of every part.

Then bring the approved replacement operations into SkyPlanner’s shared production schedule. Arcturus, SkyPlanner’s scheduling AI, calculates their placement against configured capacity, operation dependencies and priorities. The Gantt timeline shows the resulting sequence and completion dates alongside the other orders using those resources. With material-aware scheduling enabled, material availability also affects that calculation.

For the five-hour replacement run, you can review the effect on the remaining operations and other customer orders before committing to a recovery date. You control the priorities and protected scheduling decisions; Arcturus handles the connected scheduling calculation instead of leaving you to rebuild each affected machine queue by hand.

This is why SkyPlanner is a useful next step when scrap repeatedly disrupts production: the recorded loss, consumed material and work still required become part of the planning response, rather than separate figures for quality, stores and the supervisor to reconcile later. SkyPlanner integrates with any ERP or MES through its open REST API, so existing order and quality records can remain part of that connected workflow.

What material visibility changed at Eskomatic

Contract machining company Eskomatic previously organized work queues largely by hand in an older ERP and production-control system. Limited material visibility meant shortages could surface too close to a job’s planned start.

With SkyPlanner connected to its existing ERP, the company gained a clearer Gantt work queue, material-shortage visibility and automatic load scheduling. Eskomatic reports more accurate delivery-date setting and fewer production stoppages caused by missing material.

Its experience shows the practical value of aligning material readiness with the work sequence: fewer surprises when production is about to start, and a better basis for the date given to the customer.

Reduce scrap by fixing a confirmed cause

There is no single scrap percentage that makes every manufacturing process acceptable. A dedicated repeat process, a new-product launch and a high-value machining route have different loss patterns. Set a target against comparable production, customer requirements and the consequences of failure, not a universal pass mark such as 5%.

Use the records behind the rate to choose an investigation. Group losses by product, operation, material lot and confirmed reason where that information is available. Rank the issues by lost quantity, cost and delivery consequence; the largest pile of scrap is not automatically the most urgent production problem.

A practical response follows the evidence:

• Repeated dimensional rejects: check tool condition, fixtures, offsets and process settings against the actual failed requirement.
• Losses concentrated in one material lot: investigate incoming material and supplier evidence before blaming the machine or operator.
• Startup or changeover rejects: verify setup and first-off approval before committing the whole run.
• Damage between operations: examine handling, storage and protection at the point where damage occurs.

Give the corrective action an owner and check subsequent comparable batches under the same acceptance rules. Earlier inspection may stop you from adding more value to an already defective part, even before the defect source is eliminated. It is not an improvement if a lower rate merely reflects relaxed inspection or missing rejects.

Quality improvement prevents the next loss. A realistic material and capacity plan helps you handle the loss already incurred. Keeping both responses moving is how the percentage becomes better production control.

See what a rejected batch means for your next delivery

Bring one recent final rejection to a SkyPlanner demo: the original and accepted quantities, the material consumed, the approved replacement operations and an affected customer order.

Use that situation to explore a concrete recovery plan. Account for the missing material, fit the work around existing commitments and see which delivery decisions need attention before the customer has to ask.

Book a SkyPlanner demo to discuss your production, or try SkyPlanner with demo data to explore first.