Spare parts strategy that follows asset criticality

A spare parts strategy is one of the few asset management decisions that simultaneously affects safety, availability and the balance sheet. When it works, the right part is on the shelf when a critical asset fails, and the storeroom is not carrying years of obsolete stock for equipment that has already been retired. When it does not, both failure modes coexist: a multi-million-pound storeroom that still cannot supply the bearing the planner needed last night. The cause is almost always the same. Stocking decisions have drifted out of sync with asset criticality, and nobody owns the gap.

This piece is about how to design a spare parts strategy that follows asset criticality, holds together in the CMMS rather than in a parallel spreadsheet, and survives a year after the consultant has left.

What “criticality-led” actually means

Most stocking policies in mature operators have three drivers underneath them: historical demand, supplier lead time, and the price of the part. Each is a sensible input. None of them, on their own, says anything about how much it would cost the operation if the asset that needs the part went down for the duration of an unplanned procurement.

A criticality-led strategy adds the missing input. The stocking decision for any given part starts from the criticality of the assets it serves, not from the part’s own turnover or unit cost. A low-cost, low-turnover gasket on a single critical compressor is a different stocking case from the same gasket on a non-critical utility skid, even though the part number is identical and the historical demand looks the same.

The principle is simple. The downside of being out of stock is set by the asset, not by the part. A stocking policy that does not reflect that is, in effect, optimising the wrong objective.

Why most stocking policies drift from criticality

Programmes that intend to be criticality-led but end up looking like everything else usually fail in a recognisable set of places.

Criticality and inventory live in different systems

The criticality register sits with reliability engineering. The stocking policy sits with the storeroom and procurement, in inventory parameters nobody outside MRO ever opens. No field says “this min-max came from this criticality decision”. When criticality is reassessed, the inventory parameters do not change because nothing connects them.

ABC and XYZ classifications run on their own

ABC analysis on annual spend and XYZ analysis on demand variability are useful and well understood. They are also self-referential. A part that has not been used because the asset has been reliable looks like a low-volume, low-spend item, regardless of what would happen if the asset failed. Pure spend-and-demand classification systematically under-stocks insurance spares for critical equipment.

Bills of material are incomplete or stale

Without a working bill of material on critical assets, the link from “this asset is critical” to “these parts protect it” is missing. Storerooms compensate by stocking everything that has ever been issued historically, which is how a storeroom acquires its long tail.

Insurance spares are managed by personality, not policy

High-value, low-turnover insurance spares (a spare motor, a long-lead transformer, a pressure vessel) are often held because a senior engineer decided years ago that the risk of being without one was unacceptable. Without a documented criticality basis, the same item is impossible to retire when the asset is decommissioned, and equally impossible to defend in a working capital review.

The retirement loop is broken

When assets are retired or replaced with a different make, nobody systematically reviews the parts that supported them. Stock that protected a class of pumps replaced five years ago is still on the shelf, classified the same way it was on day one. Industry analysis of oil and gas operators puts working capital tied up in dead stock at 20 to 35 per cent of total MRO inventory value where this loop has not been actively managed.

Building a stocking policy that follows asset risk

The pattern in operators who get this right is not exotic. It is methodical and supported by a small number of structural decisions.

Start from a current criticality register. A stocking strategy can only follow criticality if criticality is genuinely current. The wider piece on asset criticality assessment frameworks sets out what this looks like; the relevant property here is that the register is reviewed on a defined cycle and that operations agrees with the rating.
Define stocking categories that map to criticality bands. A workable scheme has four categories: insurance (held for high-criticality assets regardless of demand), critical (held to a defined service level for high or medium-high criticality), routine (managed on min-max from demand), and consumable (managed by simple reorder points). Every stocked part sits in one category, and the category is recorded against the inventory record in the CMMS.
Use VED on top of ABC and XYZ, not instead of it. Vital, Essential, Desirable classifies parts by operational impact if unavailable. ABC ranks by annual spend, XYZ by demand predictability. The stocking decision combines all three. A Vital, low-spend, low-predictability part on a critical compressor still gets stocked. A Desirable, high-spend, high-predictability item is a candidate for consignment or vendor-managed inventory, not for sitting in working capital.
Hold the policy in the CMMS. Maximo, SAP PM and any comparable system carry inventory items, BoMs against assets, and stocking parameters. The criticality category, the service-level target and the reasoning for any insurance spare belong in those records, not in a parallel spreadsheet. If the storeroom and the maintenance planner are looking at different sources of truth, the policy is already drifting.
Run a retirement loop alongside the new-asset loop. When an asset is decommissioned, the parts that uniquely supported it are flagged. When a new asset is commissioned, the parts that protect its critical functions are added with their stocking category set on day one. Both loops belong to the same governance forum.

These steps depend on the same operational ownership described in the wider piece on asset data governance. A spare parts strategy without named stewards behaves like every other policy without named stewards.

Where the policy lives day to day

A criticality-led stocking policy earns its keep in three operational forums.

The first is the weekly maintenance review. When a stockout disrupts a job on a critical asset, the policy is the basis for asking the right question: was the part categorised correctly, was the service-level target wrong, or did the supply chain fail to meet a correct target? Each answer points at a different fix.

The second is the asset register governance forum, where criticality reassessments, BoM corrections and retirement decisions converge on a single ownership group. Without that forum, criticality changes happen in isolation and the storeroom is the last to know.

The third is the capital review. Insurance spares for new assets, refurbishment programmes and obsolescence cases sit better in capital papers when the criticality basis is in the inventory record. ISO 14224-coded failure data, where available, underwrites the assumed failure rate that drives the stocking quantity.

Measuring whether the policy is doing its job

Two metrics together tell the truth about a spare parts strategy. Stockout rate on critical-asset jobs measures whether the policy is protecting the operation. Inventory value held against retired or non-critical assets measures whether the policy is being maintained. A storeroom can move either metric on its own; only a criticality-led policy moves both in the right direction at once.

A programme that has taken this seriously for a year tends to look the same across sectors. The criticality register and the inventory parameters reference each other. Insurance spares carry a documented basis. ABC, XYZ and VED are visible against every stocked item. Retired assets trigger a parts review. The MRO budget conversation is held against operational risk rather than against the storeroom’s history.

If the stocking policy is changing in response to criticality changes, the strategy is working. If it is not, the storeroom is being audited and the plant is still being run on memory.