Why “critical spare” is not a useful category
Almost every spare parts list eventually gets sorted into two piles: critical and non-critical. The split is usually made by whoever knows the plant best, takes about an afternoon, and is never revisited. It fails for a predictable reason — criticality is not one property. It is at least four, and they interact.
A part can be cheap and utterly critical (a EUR 12 seal on a pump that feeds the whole process). A part can be expensive and completely non-critical (a spare drive for a machine that has been idle for two years). Sorting by any single dimension produces a list that either starves the real risks or ties up capital in the wrong places.
The four dimensions worth scoring
A criticality matrix scores each part on four questions, then combines them. The dimensions below are the ones that consistently predict which spares cause trouble.
| Dimension | Question | Score 1 | Score 3 | Score 5 |
|---|---|---|---|---|
| Consequence of failure | What stops if this part fails? | Cosmetic, or a redundant unit takes over | One line slows or partially stops | Whole plant stops, or safety/environmental impact |
| Lead time to replace | How long to get a working part in position? | In the local store today | 2–6 weeks from a distributor | Obsolete, or 3+ months, or unquoted |
| Failure likelihood | How often does this actually fail here? | No recorded failures in 5 years | 1–2 events in 3 years | Multiple events per year |
| Detection ability | Will you see it coming? | Condition monitoring or clear warning | Operator notices degradation | Fails with no warning |
Multiply rather than add. A part scoring 5-5-5-5 is a genuine emergency risk; a part scoring 3-5-1-3 is a slow-moving item with a long lead time, which is a planning problem rather than a crisis. Multiplication forces the distinction that averaging hides.
Reading the scores into stocking policy
The matrix is only worth building if it changes what you stock. A workable mapping:
- Score above 300 — hold at least one spare on site, plus a documented alternative source. Review the holding every six months. These are the parts where a stockout stops production.
- Score 80–300 — hold a spare if the item is also subject to a predictable replacement interval; otherwise hold a spare at a supplier under a consignment or hold-stock agreement.
- Score 20–80 — no on-site spare. Keep the part number, the source and the lead time recorded so a purchase can be raised the same day it is needed.
- Score under 20 — buy on failure. Anything else is capital sitting on a shelf.
The thresholds are less important than applying them consistently. A plant that has a written rule and follows it will outperform one with a better rule that is applied by feel.
The input that is usually missing: your own failure history
Most matrices are built from opinion because the data is not in a usable form. The fix is unglamorous: pull the last three years of work orders, filter to corrective maintenance, and count events per part. Even a rough count beats a good guess.
Two adjustments make the numbers more honest:
- Normalise by operating hours, not calendar time. A machine that runs two shifts a year is not the same as one that runs continuously, even if both are “in service.”
- Include near-misses. A failure that was caught during a planned stop is still evidence about failure likelihood. Excluding it systematically understates risk.
Turning min/max into something that survives contact with reality
Once criticality is scored, min/max levels follow from lead time and consumption rather than from habit. The calculation that works in practice:
- Establish average consumption per period from actual issues, not from the vendor’s recommended interval.
- Establish the realistic replenishment lead time — including the internal time to raise the order, which is often longer than the supplier’s quoted lead time.
- Set the reorder point at consumption during lead time, plus a safety stock sized to the criticality score.
- Set the maximum at the reorder point plus one economic order quantity, not at “a year’s worth.”
Review the levels annually against actual issues. A min/max that has never triggered in three years is telling you that the part does not need to be stocked, or that the level is set above the highest consumption the plant will ever see.
The output that makes the exercise pay
A criticality-scored spare parts list produces three things beyond the obvious: a defensible stocking budget, a shortlist of parts that need a qualified second source, and a shortlist of parts that should be examined for obsolescence risk before the maker announces end-of-life. That third output is where the money actually is — see our guide to sourcing obsolete and discontinued parts for what to do when a critical part is already out of production.
How we help
We supply industrial and automation spare parts from China and we are used to working from a criticality list: send us the parts you have flagged as high consequence and long lead time, and we will come back with availability, lead time and cost so you can decide what to hold and what to leave to the supply chain.
Need a specific reference? Contact us freely at zhuaotech.com/contact-us — send the part number and we will confirm what we can supply.
Related reading
- MRO spare parts inventory planning: criticality, min/max and reorder points
- Last-time buy planning: how to estimate lifetime demand
- Second source qualification checklist for legacy spare parts
Need a specific reference? Contact us freely at zhuaotech.com/contact-us — send the part number and we will confirm what we can supply.