Spare Parts and MRO Inventory: Criticality and Stock Levels

What spare parts and MRO inventory are

Spare parts are the items kept on hand to replace components that fail in your equipment. MRO (Maintenance, Repair and Operations) is the wider group: spares, maintenance consumables (lubricants, belts, seals, PPE) and tools, everything that keeps the plant running but never ends up in the product.

The central question is how much to stock of each item. Too much ties up capital and ages on the shelf; too little turns a $400 failure into days of downtime. The answer takes two steps: rank criticality first, then set the quantity.

Visual map of spare parts and MRO inventory

The map summarizes the criticality matrix, the policy per class and the worked example from this article.

Step 1: rank criticality

Criticality is not price and not usage. An $18 sensor can stop an entire line. Use two factors scored 1 to 5 and multiply them:

Criticality = Failure impact × Replenishment lead time

• Impact (1 to 5): what happens if the item fails and there is no part. 1 = no effect on production; 3 = partial loss or bottleneck; 5 = full stoppage, safety or environmental risk.
• Lead time (1 to 5): time to get a replacement. 1 = same-day from a local supplier; 3 = days to a few weeks; 5 = imported, made to order or special manufacture (months).

Take impact from failure analysis of the equipment: use the FMEA or the answers from RCM, not the opinion of whoever is in a hurry. With a score from 1 to 25, define three classes:

Step 2: set the stocking policy

Criticality decides whether and with what margin you stock; usage decides how much. That is why the top and bottom classes follow different logic:

• Regular-usage items (standard bearings, filters, belts): compute reorder point and safety stock as in safety stock and reorder point and economic order quantity. The ABC curve and the ABC-XYZ matrix help you focus attention.
• Slow-moving, high-impact items (drives, PLC cards, gearboxes, large motors): demand is a few units a year and averages mislead. Use the probability of failures within the lead time, as in the example below.
• Repairable items (motors, pumps, gearboxes): treat the repair loop as the lead time and keep one repaired unit in circulation.

For low-criticality items with unpredictable usage, consider moving the stock to the supplier through VMI or consignment.

Example: a 55 kW variable frequency drive

Fictional example from a bottling plant. The table shows the criticality of six items:

Now size the drive. Data from history and the supplier:

• Failures: on average 1 per year.
• Replenishment lead time (imported): 90 days.
• Price: $18,000; holding cost of 20% per year (capital, space, insurance, obsolescence) = $3,600 per unit per year.
• Downtime cost without a part: 40 h × $3,000/h = $120,000 per event.

Expected failures during the lead time:

λ = 1 failure/year × 90 ÷ 365 = 0.2466

With the Poisson distribution, P(X = k) = e^−λ × λ^k ÷ k!. For λ = 0.2466:

With a 95% service target, the part must be available for at least 95% of failures. When the total stock (on the shelf plus on order) is S, availability is P(X ≤ S − 1). S = 1 gives 78.15%; S = 2 gives 97.42%. Two drives meet the target.

Checking by cost: each extra unit prevents downtime in P(X = k) of failures, and the yearly gain is that value × 1 failure × $120,000:

The third unit costs more than it saves. Both criteria point to 2 drives. Before buying, compare alternatives: sharing a unit between sister plants, an express repair contract, or standardizing the rating so the same model fits other machines.

Keeping the inventory alive

The calculation ages. Keep it current with a simple routine:

1. Link each part to its equipment in the asset bill of materials in the CMMS. Without the link, nobody knows where the part is used.
2. Charge every issue to a work order. That is how real consumption shows up and the calculation self-corrects.
3. Review the classification once a year and after any change of equipment, supplier or lead time.
4. Deal with idle items: a low-criticality part with no movement for 24 months is a candidate for write-off, sale or return. An idle critical part is not excess, it is insurance: keep it, but record the justification.
5. Run cycle counts on critical items, with accuracy close to 100%. A spare the system says exists and does not is worse than none.

Also track the storeroom fill rate (parts requests served on the first ask) next to MTBF and MTTR: repair time includes waiting for the part. To plan the parts for a planned shutdown, work from the scope bill of materials and pull forward the long-lead items.

Common mistakes

• Ranking by price. A cheap item with high impact and long lead time needs stock.
• Using average usage on slow movers. With 1 failure a year, a "minimum of 0.5" does not exist; use probability.
• Ignoring the real lead time. The promised lead time is not the delivered one. Measure the last purchases.
• Duplicate stock across storerooms with no central view, the same item sitting in three places.
• Poor item coding. Free-text descriptions create duplicate records and phantom stock.
• Stocking without repairing. A repairable part outside the repair loop becomes expensive scrap.

Frequently asked questions

What does MRO stand for?

Maintenance, Repair and Operations: spares, maintenance consumables and tools used to keep operations running, outside the finished product.

How do I decide which parts to stock?

Score failure impact and replenishment lead time from 1 to 5 and multiply. High scores (15 or more) deserve risk-based stock.

How do I size stock for a slow-moving critical item?

Compute expected failures over the lead time (λ) and use Poisson to find the smallest quantity that reaches the service target, then check downtime cost against holding cost.

What does holding spare parts cost?

It adds tied-up capital, space, insurance, handling and obsolescence. Many companies use around 20% a year of item value, but the right number is your own.

Should a critical part idle for years be scrapped?

Not necessarily. If failure is unlikely but devastating and the lead time is long, the stock works as insurance. Review the justification, not only the turnover.

What is the difference between a spare and a consumable?

A spare replaces a failed equipment component. A consumable is used up continuously in operation, like lubricant, filters and PPE.

About the author

Vagner Soares

Lean Manufacturing & Behavioral Management Specialist

Over 20 years in the automotive and metalworking industries (GM and Dana), Lean Manufacturing practitioner since 2006. SENAI instructor and mentor in Brazil’s Brasil Mais Produtivo program, delivering consulting, training and audits for 50+ companies, combining quality, productivity and people development.

LinkedInInstagramFull profile →