Table of Contents
A warehouse can contain thousands of components while a production order remains short. Some material may be quarantined, assigned to another program, packed in an unusable format, or tied to an unapproved ordering-code suffix. A buffer model that counts all of it as available begins with the wrong number.
For replenishable electronic components, set a reorder point from demand during the time needed to obtain usable material, then add a separately justified safety stock. The examples below are planning illustrations, not recommended inventory targets.
Define the event that ends replenishment time
Use the interval from the relevant ordering decision to release for production. Depending on the purchasing arrangement, this can include supplier processing, manufacturing, shipping, customs, receiving, inspection, programming, and repacking.
SAP’s explanation of reorder-point planning, accessed September 15, 2026, distinguishes safety stock from average demand during replenishment lead time. That distinction prevents the same normal consumption from being funded twice.
Measure lead time using consistent start and end timestamps. A supplier’s shipment date is not the same event as quality release. For a split delivery, record quantities and release dates separately; the first small arrival cannot represent the whole order.
Count an inventory position the factory can use
A practical inventory position is usable on-hand stock plus qualifying open supply minus unfulfilled demand commitments. Define which receipts qualify and how reservations are counted so a commitment is not subtracted twice.
| Quantity category | Base planning treatment | Reason |
|---|---|---|
| Released, approved stock | Count | Available to satisfy the modeled demand |
| Quarantine or unresolved deviation | Track separately | Release date and usability are uncertain |
| Confirmed inbound supply | Include by usable date | A late arrival cannot cover an earlier build |
| Unconfirmed seller offer | Exclude from firm coverage | An offer is not an accepted supply commitment |
| Stock reserved to another program | Exclude unless formally reassigned | Ownership of the quantity matters |
| Alternate not yet qualified | Exclude from approved coverage | Technical acceptance is incomplete |
Use the BOM risk assessment to identify which lines need this discipline first. A low-cost component can still be the sole cause of a line stop.
A reorder-point calculation
For independent daily demand with a stable mean and standard deviation, and a fixed replenishment interval:
Safety stock = z × daily demand standard deviation × √lead-time days.
Reorder point = mean daily demand × lead-time days + safety stock.
Assume demand averages 100 units per day, its standard deviation is 20 units per day, lead time is fixed at 25 days, and the chosen cycle service factor is 1.645. Safety stock is 164.5 units, rounded up to 165; the reorder point is 2,665 units.
This is a continuous-review approximation under a normal demand model. It does not set the purchase lot size. Periodic review requires coverage of the review interval as well as replenishment, and the chosen service measure must remain explicit.
When lead time also varies
If daily demand and lead time are independent, a common approximation adds their variance contributions:
Safety stock = z × √(L × σd² + d² × σL²).
Here L is average lead-time days, d is average daily demand, σd is daily demand standard deviation, and σL is lead-time standard deviation in days.
Keeping the earlier assumptions but allowing a four-day lead-time standard deviation produces about 678.3 units of safety stock, rounded to 679. The reorder point becomes 3,179 units. That large change shows why delivery variability deserves measurement.
SAP’s safety-stock planning documentation also treats demand and procurement forecast errors separately and makes an independence assumption in their combination. The calculation here is an illustrative analytical model, not a claim to reproduce a particular SAP configuration.
Where the model stops being useful
Lumpy service demand, new-product ramps, correlated demand and lead-time shocks, and impending discontinuation can make the simple model misleading. Use explicit build and repair scenarios where the historical distribution no longer represents the planning task.
For a component reaching end of life, use a last-time-buy demand model rather than assuming future replenishment. For continuing industrial production, review actual stockouts, excess stock, and usable delivery performance together.
The model should lead to a decision: revise a buffer, improve inspection turnaround, qualify a second source, or change the order cadence. A larger number alone cannot resolve a material or design constraint.
Frequently Asked Questions (FAQ)
Is a 95% cycle service level the same as filling 95% of demand?
No. Cycle service level is the probability of avoiding a stockout during a replenishment cycle under the model. Fill rate measures the share of demand supplied immediately; the two require different calculations.
Should quarantined components count as available inventory?
Exclude them from immediately usable stock until the required release is complete. A planning system can show them separately with an expected release date and uncertainty.
Can a normal-distribution formula model an allocation crisis?
Usually not adequately on its own. Correlated delays, rationing, intermittent demand, and lifecycle events require scenario planning and supply actions in addition to a statistical buffer.
