Supply Chain & InventoryFree calculatorPDF, Excel, Word export

Safety Stock Calculator with Demand and Lead Time Variability

Most online safety stock calculators assume the supplier is always on time. This one adds lead time variability, which is often the larger driver. Enter your numbers, pick a service level and read safety stock and reorder point.

Your numbers

units/day

Mean demand per day over a recent stable period.

units

Standard deviation of daily demand.

days

Order release to available in stock.

4 days

Measured from actual receipts versus order dates.

Your results

Safety stock
284
Buffer units above expected lead time demand.
Reorder point
1,124
Z score
1.6449
Standard normal value for the chosen service level.
Std deviation of demand during lead time
172.3833
Combines demand and lead time variability.
Expected demand during lead time
840
Safety stock if lead time were fixed
106
Share of buffer caused by lead time variability
62.8%

Assumes demand during lead time is approximately normal and that demand and lead time are independent. Not suitable for very lumpy or intermittent items.

Save results:

Safety stock vs lead time variability (days)

Sweeps Std deviation of lead time from 0 to 15, holding your other inputs fixed.

Check your ERP safety stock against real receipts

Optional: an expert review of how to compare loaded safety stock values with the numbers your own history supports.

Optional. The tool and every export work without this. No spam.

How to use the calculator

Take daily demand statistics from issue history and lead time statistics from actual receipts: receipt date minus order date, averaged across the last 12 to 24 receipts. Choose the cycle service level, which is the probability of not stocking out in a replenishment cycle. The chart sweeps lead time variability from 0 to 15 days with your other inputs fixed, so you can see what negotiating a more reliable supplier is worth in units of buffer.

The formula

Safety stock = Z x sigma_LTD, where sigma_LTD = sqrt(LT x sigma_d^2 + d^2 x sigma_LT^2). The first term is demand variability accumulated across the lead time, the second is the effect of the lead time itself being uncertain, scaled by average demand. Reorder point = d x LT + safety stock. The calculator also shows the buffer you would need if lead time were fixed, so you can isolate the supplier effect.

  • d: average demand per day
  • sigma_d: standard deviation of demand per day
  • LT and sigma_LT: lead time mean and standard deviation in days
  • Z: normal score for the service level

Reading the result

With the default values, the lead time term contributes a large share of the buffer even though four days of variation on 21 looks modest. When the share is high, spending effort on supplier delivery performance, shorter transit or a second source often reduces inventory more cheaply than lowering the service target. Moving from 95% to 99% raises Z from about 1.64 to 2.33, roughly 40% more buffer, so confirm the higher target is worth the carrying cost.

Common mistakes

Using planned lead time from the item master instead of actual receipts is the biggest one. Another is mixing time units: demand per day with lead time in weeks. Using monthly standard deviation scaled incorrectly is the third. If you only have monthly figures, convert both mean and standard deviation to the same period, standard deviation scales with the square root of time, not linearly. Finally, do not apply this formula to sporadic demand, results will be unreliable.

Frequently Asked Questions

What is the difference between cycle service level and fill rate?

Cycle service level is the chance of no stockout in a replenishment cycle. Fill rate is the share of demand filled from stock. They are different numbers. See the service level and loss function table for converting between them.

Why is my safety stock higher than the simple formula?

The simple formula Z x sigma_d x sqrt(LT) ignores supplier variability. If your suppliers are late or early by several days, the extra term adds buffer, and it can dominate for fast moving items.

How many receipts do I need to estimate lead time variability?

A dozen is a workable minimum. Fewer than that and the standard deviation is noisy. Separate suppliers or ship methods if they behave differently, one blended figure can hide the real driver.

Does this apply to make items?

Yes, use the manufacturing lead time and its variability from work orders, with demand from dependent or independent usage. For items driven by MRP on firm demand, a time-phased approach may suit better than a statistical buffer.

If you want safety stock recomputed from your ERP receipts and issues and compared with what is loaded in the item master, Netray can build a read-only analysis over your ERP data.