Inventory production quantity model: formula and limits

The inventory production quantity model calculates an economic production lot when inventory is replenished gradually and demand continues during the run. Usually called economic production quantity, or EPQ, it balances setup costs against the cost of holding cycle stock. It is useful for comparing lot-size assumptions before production planning checks whether those lots fit the factory.

The EPQ formula

Economic production quantity, Q = √[2 × D × S / (H × (1 − D/P))]

Peak cycle stock = Q × (1 − D/P)

Average cycle stock = Q × (1 − D/P) / 2

Use these inputs:

P must exceed D. These rates describe production and demand on the same time basis; P is not the annual quantity actually produced after allowing for time spent making other products.

The starting cost tradeoff is the same one described in NC State's economic order quantity tutorial: larger lots reduce the number of orders or setups but increase average stock. EPQ changes the stock calculation because replenishment is gradual. J. E. Beasley's inventory-control notes identify production-rate replenishment as a case where goods arrive as they are produced rather than all together.

Why production changes the inventory peak

One production run takes Q/P years. During that time, demand consumes D × Q/P units. The remaining stock at the end of the run is therefore Q − DQ/P, which simplifies to Q × (1 − D/P).

Under constant rates, stock rises in a straight line during the run and falls in a straight line afterward. Average cycle stock is half the peak. The two annual costs that change with lot size are:

Annual setup cost = S × D/Q

Annual holding cost = H × Q × (1 − D/P) / 2

Annual relevant cost = Setup cost + Holding cost

Balancing those two cost terms gives the EPQ formula above. Constant unit production costs do not affect the choice because the annual demand quantity is unchanged. If unit costs change with lot size, include that difference in the comparison rather than applying the basic formula unchanged.

This cost view differs from inventory turnover, which measures observed inventory use. EPQ is a planning calculation driven by explicit assumptions.

Build the inputs from business records

Use shipment or consumption history to establish demand, with documented adjustments for known changes. Use production records to estimate the good-unit rate while running, rather than a machine's theoretical nameplate rate.

Setup records should identify changeover labor, consumables, and other costs that change when another run is added. Keep setup cost separate from the per-unit manufacturing cost. The holding-cost worksheet should explain its unit valuation, financing assumption, storage, and other relevant costs; the inventory's carrying value is a balance, not the annual holding cost itself.

Keep each item's assumptions together. An attractive number built from one item's demand and another item's setup cost is not a useful production recommendation.

A worked US manufacturing example

An illustrative Ohio component manufacturer plans one item using the inputs below. The operating calendar contains 240 days, so demand is 100 units per operating day and production while running is 400 good units per operating day.

The square-root input is 2,560,000, giving an exact lot size of 1,600 units. At that quantity, the annual number of runs is 24,000 / 1,600 = 15.

Compare nearby quantities instead of treating the formula as an instruction to the shop floor:

All quantities and costs in this example are exact; no rounding is required. The reported cost excludes unit production costs that are identical across these scenarios, and it excludes separately determined safety stock.

For a workbook with D in B2, P in C2, S in D2, and H in E2:

=IF(AND(B2>0,C2>B2,D2>0,E2>0),SQRT(2*B2*D2/(E2*(1-B2/C2))),NA())

The guard flags invalid inputs instead of silently returning a misleading quantity.

Prove the cycle and test one changed assumption

The 1,600-unit run lasts 1,600 / 400 = 4 operating days. Demand during the run is 4 × 100 = 400 units, leaving 1,200 units at the peak. Those units cover 1,200 / 100 = 12 further operating days, so the full cycle is 4 + 12 = 16 days.

The annual reconciliation is 15 runs × 1,600 units = 24,000 units, and 15 cycles × 16 days = 240 operating days. Setup and holding costs also match at the unconstrained optimum: $4,800 each. These checks validate the example's arithmetic, not machine availability for all the factory's other items.

Now change only setup cost from $320 to $80. EPQ becomes √[2 × 24,000 × 80 / (8 × 0.75)] = 800 units. Annual setup cost is 30 × 80 = $2,400, while holding cost is 300 × 8 = $2,400, for $4,800 total. Reducing the cost of starting a run supports smaller lots under the same demand and production assumptions.

Where it goes wrong

• Inconsistent rate units. Daily demand divided by annual production gives the wrong accumulation fraction. Use the same calendar and unit throughout.
• Production cannot exceed demand. P less than or equal to D makes the basic EPQ case infeasible. Resolve capacity before discussing lot economics.
• Variable demand treated as constant. Seasonality, downtime, and uncertain yields require separate availability and buffer decisions.
• Shared capacity omitted. A feasible lot for one item may conflict with other items, sequence-dependent setups, or maintenance windows.
• Accounting costs mistaken for relevant costs. An allocation that will not change with another setup may distort S. Document which costs actually move with the decision.

Use raw-material turnover to inspect upstream stock consumption and days inventory outstanding for the wider inventory investment. Neither measure replaces the capacity review behind a production plan.

Compare production assumptions on your own workbook

Upload the demand, production-rate, setup-cost, and holding-cost assumptions you want to compare; Covirage maps the columns, and its tools compute and check the supplied figures. The external AI model explains the comparisons with every figure traced to its rows. Create a free Covirage account and bring a sample of your own data; creating the account takes about a minute.

See Covirage for industrial manufacturers for the supporting business context.