Operations: Capacity, Inventory, Quality and Standard Costing
What you must know
Operations questions cluster around capacity and throughput, inventory management, lean thinking, quality economics and standard costing variances. Understand the Theory of Constraints sequence (identify the bottleneck, exploit it, subordinate everything else, elevate it) and Little's Law: work in progress = throughput × flow time. Know EOQ = √(2DS/H) and that reorder point equals demand during lead time plus safety stock. In lean, recognise the seven wastes, JIT, kanban, value-stream mapping and 5S. For quality, separate conformance costs (prevention, appraisal) from non-conformance costs (internal and external failure), and know Cp versus Cpk. For variances, be fluent in material price and usage, labour rate and efficiency, and the reconciliation of standard to actual contribution.
Traps
A local efficiency gain at a non-bottleneck builds inventory; it does not raise throughput. EOQ minimises ordering plus holding cost only, so a quantity discount must be tested on total cost, not on EOQ logic. Calculate the material price variance on quantity purchased and the usage variance on quantity used. A favourable variance is not automatically good: cheap material often raises scrap and rework. External failure is normally the most expensive quality cost, which is why prevention is argued for. Cpk can never exceed Cp; Cp shows potential capability, Cpk shows capability after centring. Volume variances and capacity variances are frequently confused with spending variances.
Worked example
Annual demand is 36,000 units, ordering cost 50 per order and holding cost 4 per unit per year. EOQ = √(2 × 36,000 × 50 ÷ 4) = √900,000 ≈ 949 units. A supplier offering a 2% discount on orders of 3,000 units must be evaluated by comparing purchase cost savings against the higher average inventory holding cost - not by rejecting the larger order on EOQ grounds.
Rule of thumb
The system's output equals the bottleneck's output; protect and elevate the constraint before optimising anything else.