Table of Contents
Buying exactly the bill-of-materials quantity can leave an assembly run short when setup, feeder rejects and destructive checks consume material. Adding one blanket percentage can create the opposite problem: expensive excess on lines that need little additional stock.
The calculation separates fixed losses from losses that scale with placement attempts. It also keeps supplier pack rounding visible, so commercial minimums do not masquerade as process consumption.
How do you calculate a line-level requirement?
Start with the number of successful placements required, divide by the expected placement yield, then add justified fixed consumption and subtract usable allocated stock. Round the resulting net requirement to the supplier’s permitted order increment.
For one BOM line, define:
- B: planned boards entering this assembly operation.
- P: placements of the part on each board.
- a: expected loss fraction per placement attempt, from 0 to less than 1.
- S: fixed setup/changeover consumption in units.
- T: destructive-test or retained-sample allocation in units.
- U: usable stock already allocated to this job.
A planning estimate is Qnet = max(0, ceil(B × P / (1 − a)) + S + T − U). If the loss percentage is defined against successful placements instead of attempts, use the corresponding model and label that denominator. The two conventions should not be mixed.
What does the calculation look like for a real order size?
For a 1,000-board run using four devices each, assume 1% loss per placement attempt, 30 setup units, 10 test units and 200 usable units already allocated. The estimate becomes ceil(4,000 / 0.99) + 30 + 10 − 200 = 3,881 units.
If the supplier requires multiples of 500, the purchase becomes 4,000 units, leaving 119 units above the modeled net requirement. That 119 is pack rounding, not predicted manufacturing loss. Keep it visible as potential residual stock.
Compare the financial consequence before accepting the increment. At an illustrative USD 8 per unit, the 119-unit rounding exposure is USD 952. A smaller permitted carrier could be economical if its additional handling and feeder costs are less than that exposure. The MOQ/MPQ quotation comparison covers the commercial terms separately. If the result feeds an end-of-life coverage decision, the last-time-buy quantity model combines production, service and this line-level allowance without double-counting.
Which inputs need evidence from the assembly process?
Use package- and process-specific records for variable loss, and actual changeover requirements for fixed consumption. A new line or new package can justify a provisional assumption, but it should be labeled and reviewed after the build.
| Input | Useful evidence | Common double count |
|---|---|---|
| Planned boards | Production release and board-yield convention | Adding board scrap twice |
| Placement loss | Attempted versus successful placements | Treating feeder rejects as both fixed and variable |
| Setup units | Feeder loading and changeover practice | Applying the same setup allowance per board |
| Test/sample units | Approved inspection and retention plan | Adding samples already included in assembly’s allowance |
| Usable stock | Released lot count and condition | Subtracting quarantined or untraceable material |
Carrier details affect the setup assumption. TI’s carrier-material lookup illustrates how packing quantities and pocket dimensions are tied to specific carrier options. Use the actual supplied carrier and the assembler’s feeder requirements, not a universal leader or loss allowance.
Close the loop after the build
Reconcile issued quantity against placements, rejects, test consumption, retained samples and returned material. Investigate unexplained differences before using the result as the next run’s attrition rate.
The record should separate preventable feeder/setup loss from normal planned consumption. If rejects rise after a carrier change, review the tape orientation and feeder handoff before increasing the purchasing percentage. Extra inventory can conceal an assembly problem without solving it.
The BOM sourcing input guide can carry these line-level quantities into a sourcing request. Keep the calculation, commercial rounding and final award quantity distinct so a future planner can see what changed.
Frequently Asked Questions (FAQ)
Should the same attrition percentage apply to every BOM line?
No. Feeder behavior, package, handling, changeovers and process history differ by line. Use justified line-level assumptions and replace them with measured production data when available.
Is board scrap the same as placement attrition?
No. Board scrap can consume successfully placed components across the whole assembly. Model it separately if the planned board quantity has not already included it.
Can the assembler’s unused allowance be returned to available stock immediately?
Only after the remaining quantity, lot identity, packaging and exposure history have been reconciled. An estimated unused balance is not yet released inventory.