Every maintenance storeroom has two problems at the same time. Too much money tied up in parts nobody has touched in years, and a stockout on the one part that would have saved a 14 hour outage.

Both come from the same cause. Nobody decided, part by part, why it is on the shelf. Parts got stocked because a vendor recommended a spares kit, or because someone got burned once, or because the equipment came with a box of them.

Here is the method I use to sort it out.

The three numbers that decide it

For any spare part, the stocking decision comes down to comparing two costs:

  • The cost of not having it when you need it
  • The cost of keeping it on the shelf until you do

The cost of not having it is roughly downtime cost per hour times the hours you would wait. The hours you would wait are mostly lead time. So you need three numbers per part: downtime cost, lead time, and carrying cost.

Downtime cost

This is the cost per hour when the equipment that uses the part is down. Not every machine is equal. A packaging line that feeds every shipment might cost $8,000 an hour in lost output and overtime to catch up. A backup air compressor might cost nothing if the primary is running.

Ask production and finance together. You will get two different numbers. Use the one you can both defend.

Lead time

Use real lead time, not the catalog number. Look at your last few purchase orders for the part, or call the supplier and ask what an emergency order would take today. For many OEM parts, especially on older or imported equipment, I see quoted lead times of 6 to 16 weeks.

Also check whether a local distributor carries it. A part you can get across town in two hours is a very different decision than one coming from overseas.

Carrying cost

Most companies use somewhere between 20 and 30 percent of part value per year for carrying cost. That covers capital, space, insurance, and the risk the part goes obsolete or degrades on the shelf. Rubber seals and belts do age, and some electronics have a shelf life too.

Rank criticality before you do any math

You do not need to run numbers on all 8,000 items. Start by ranking equipment, then work down to parts.

I use a simple three-level scale for equipment:

  1. A: failure stops production or creates a safety or environmental risk, and there is no backup
  2. B: failure slows production or there is a workaround that costs real money
  3. C: failure is an inconvenience, or there is a redundant unit

Then for the parts on A equipment, ask whether the part failing takes the machine down. A failed bearing on the main drive does. A cracked guard cover probably does not, even on the most important machine in the plant.

The parts that land on A equipment and stop it when they fail are your critical spares. In most plants I visit, that list is a few hundred items, not thousands.

Do the comparison

For each critical spare, compare:

Cost of a stockout = downtime cost per hour x hours to get the part (lead time plus install)

Cost of stocking = part cost x carrying rate x years you expect to hold it

A $3,000 gearbox on a line that costs $5,000 an hour, with a 4 week lead time, is not a hard call. Even one stockout costs more than carrying that gearbox for a decade.

A $40,000 spindle with a 2 day lead time from a local rebuilder is a different story. You might be better off with a standing agreement with the rebuilder than with a spindle on the shelf.

Use failure history where you have it

If your maintenance records show a part fails every 18 months, use that. If you have no history, use the manufacturer's guidance or what other plants with the same equipment experience. Then revisit after a year of tracking.

Min/max for parts that rarely move

Standard reorder point math assumes regular demand. Critical spares do not have regular demand. A part might be used zero times for three years and then twice in a month.

For these, I set min/max by judgment, not formula:

  • For most critical spares, min is 1 and max is 1. When you use it, you reorder one.
  • For parts with multiple installed units (say, the same motor on 6 conveyors), min might be 1 or 2 depending on lead time and how many could fail at once.
  • For parts that fail in pairs or sets, stock the set.

The point is that the reorder trigger fires on the first use. If your system treats a part with no movement in 12 months as a candidate for zero stock, exclude critical spares from that rule. I have seen automated "slow mover" cleanups remove exactly the parts that mattered most.

Protect the count

A critical spare that the system says is there but is not on the shelf is worse than no spare at all. Everyone relaxes. Count critical spares on a fixed schedule, at least quarterly, and check their condition while you are there. Look for rust, damaged packaging, or a part that was quietly swapped with a used one.

Parts for retired equipment

This is the easy money. When a machine is retired, its spare parts usually stay in the storeroom forever. Nobody connects the decommission with the parts list.

I have seen storerooms where 15 to 25 percent of the value was parts for equipment that no longer existed in the building.

To find them:

  1. Get a list of every machine retired, sold, or scrapped in the last ten years
  2. Pull the parts list (bill of materials or where-used) for each one
  3. Check whether any of those parts are used on equipment still in service
  4. Anything that is not used elsewhere is a candidate to sell, return, or scrap

Talk to the vendor first. Some will take back unused parts for credit. Surplus dealers will buy others. Even scrapping them frees shelf space for the parts you actually need.

Then add a step to your decommissioning checklist so it does not build up again.

Where to start

Pick your top five A machines. Build the critical spares list for those first, check the counts, and set min/max. That alone will cover most of your real outage risk, and you can work through the rest of the plant one line at a time.