Spare Parts and MRO Inventory: Balancing the Cost of Downtime Against the Cost of Holding

Spare parts and MRO (maintenance, repair and operations) items follow a different logic from sale goods. Their value is not in being sold but in preventing downtime. A cheap part whose absence stops a production line can matter more than an expensive one whose shortage creates no problem. So classification by monetary value alone is misleading.

The first criterion is criticality: what happens if this part is missing? Complete line stoppage, reduced capacity, or just a repair delay. The second is substitutability: is there an equivalent or otherwise usable part? The third is lead time: a part buyable in a day needs a different policy from a three-month import.

Combining these criteria gives a simple matrix. Critical parts with no substitute and a long lead time are the hardest group and are usually stocked even if unused for years. Non-critical parts with substitutes and fast supply need no stock and can be bought when required. Between them lie parts needing item-by-item decisions.

To decide on holding, two costs must be compared: holding cost (capital, space, obsolescence) against expected downtime cost (failure probability times cost per hour of downtime times wait time for supply). When downtime cost is far above holding cost, holding even an unused part is rational. Many organizations never do this calculation and decide by habit or fear.

Stock sharing is a valuable solution. If several units or plants use an expensive, low-usage part, one central shared stock may cover the risk with less capital. A consignment agreement or a fast-delivery arrangement with a supplier can also move part of the stock to the supplier.

A common mistake is managing spare parts with sales rules (such as removing items with no usage in a year). This rule deletes low-usage critical parts and creates a shortage precisely on the day they are needed. Another mistake is lacking a part record (criticality, substitute, supplier, lead time) and relying on the memory of a few people who may leave the organization.

Coordination with maintenance is essential. The preventive maintenance schedule says part needs in advance, and they can be drawn from the schedule rather than a statistical forecast. Only emergency repairs are truly random and need safety stock.

A practical starting point: select 50 high-value or critical spare parts and build a record for each with four fields: criticality, substitute, lead time and approximate downtime cost. Then write explicit hold or don't-hold decisions for each.