What makes a spare part critical?

Two things together: its failure stops or endangers an asset that matters, and you could not get one fast enough if it failed. A part on a class A asset that a local supplier stocks within the hour may not need holding; a part on a class B asset that is obsolete usually does.

How do you calculate the reorder point for a spare part used only a few times a year?

Work out m = used a year x lead time in weeks / 52, then find the smallest stock k whose cumulative Poisson chance (k or fewer failures in a lead time) reaches your target. For m = 0.29 that is 1 at 90% or 95% and 2 at 99%. Excel gives the chance with POISSON(k, m, TRUE).

Why not use the safety stock formula with a z-score?

That formula assumes demand that is roughly normal, like a part used every week. A spare used once or twice a year is not: in most lead times it is used zero times. The Poisson model counts the chance of zero, one or two failures in a lead time directly. For steady demand, use the safety stock calculator.

How is this different from the equipment criticality analysis?

The equipment criticality analysis ranks assets on six weighted criteria and says class A assets need critical spares, but not which parts or how many. This list takes that class, or a quick four-question screen, and decides each part's class, policy, min and max.

What service level should critical spares have?

Your site decides. The template's example settings use 99% for Critical, 95% for Important and 90% for Routine parts, the three levels one plant used in EPRI report 1013472 (2006). The downtime cost columns show what a higher or lower target is worth.

Free template

Updated October 9, 2026

Critical spare parts list and stocking levels

A critical spare part is one you keep because waiting for it would stop or endanger a critical asset. The list starts with the equipment: each asset is scored on safety and environment, production, quality and repair time, or takes the class from your own criticality ranking. Each part then gets a class from what its failure does to that asset and how hard it is to get (single source, obsolete, long lead), and one of four policies: hold on site, repair and cycle, hold at a shared or regional store, or buy on demand. For the parts you hold, the minimum (the reorder point) comes from usage and lead time with the Poisson model, which suits parts used a few times a year, and the maximum is the minimum plus the order quantity. The Excel version works out the classes, policies, min and max, the chance a part is on the shelf when needed, what to order now, the stock value, and the downtime cost a year with and without stock from your own downtime cost per hour, then lists one action per part. The PDF has the paper sheets, a table that gives the minimum without a spreadsheet, and a worked example.

Free to use: print it, copy it and edit it for your team. Enter your name and work email once to download.

Page 1 of the critical spare parts list — Equipment criticality screenPart class from consequence and supplyHold, repair, shared store or buyMin and max from usage and lead timeChance on the shelf when neededDowntime cost of a stock-outOrder now and action listWorked example

When to use it

When to use a critical spare parts list

After an equipment criticality ranking, to decide which parts for the class A and B assets to keep and how many.
When a breakdown waited days for a part, to check the other parts on that asset before the next one.
In the yearly storeroom review, to find stock that is no longer needed and gaps that are.
When a supplier stops making a part or becomes the only source, to plan a replacement or a last-time buy.

How to fill it in

1
Rank the equipment first
Score each asset 1 to 3 on safety and environment, production and quality, and score its repair time from the hours it takes with the parts on hand. A total of 9 or more, or a safety score of 3, is class A in the example settings. Already ranked your assets? Type that class instead.
2
List the parts per asset, with real usage
Take the parts list for each class A and B asset and count how many of each were used on breakdowns in the last two or three years, from the work orders. Planned replacements are ordered for the job, so leave them out. A part on several assets goes on one row, on the most critical asset.
3
Enter what each part's failure does and how you would get one
Stops the asset, degrades it or minor. Then the lead time, unit cost, whether it is single source or obsolete, the fastest other way to get one (local supplier stock, a shared store, an expedited order) and its hours, and whether it can be repaired.
4
Let the class pick the policy
The part class is the lower of the asset's class and the part's effect, raised to Critical when a part that matters is hard to get. The first rule that applies sets the policy: used often, repair faster than buying, stock pays back, routine, fast enough to get, otherwise hold.
5
Set min and max
m = used a year x lead time in weeks / 52. Min is the smallest stock whose cumulative Poisson chance of covering m reaches the target for the class (99%, 95% and 90% in the example settings). Max = min + order quantity. Order up to max when on hand plus on order falls to min.
6
Act on the list, then review it
Order what is below min, stop ordering what is above max, find a second source or a replacement for single-source and obsolete parts, and confirm the shared store really holds what you rely on. Review when a part is used, a lead time changes, or an asset gets a backup, and at least once a year.

Four stocking policies

Which policy fits which part

The rules are the template's own, in the order the Excel applies them. Every limit in them is a setting you can change.

Policy	When the list chooses it	Stock levels
Hold on site	Used often (one or more expected in a lead time), or stock pays back within the limit, or a Critical or Important part takes longer to get than the wait you accept	Min from the Poisson chance at the class target; max = min + order quantity
Repair and cycle	The part can be repaired, it is Critical or Important, and repair is faster than buying a new one	Same rule, with the repair turnaround in place of the lead time; max is the number of spares in the loop
Hold at a shared or regional store	A shared store can deliver within the wait you accept, and holding your own does not pay back	None on site; confirm the store holds one and will reserve it
Buy on demand	Routine parts, and parts you can get within the wait you accept when stock does not pay back	None; min and max are zero

Hold on site
When the list chooses it
Used often (one or more expected in a lead time), or stock pays back within the limit, or a Critical or Important part takes longer to get than the wait you accept
Stock levels
Min from the Poisson chance at the class target; max = min + order quantity
Repair and cycle
When the list chooses it
The part can be repaired, it is Critical or Important, and repair is faster than buying a new one
Stock levels
Same rule, with the repair turnaround in place of the lead time; max is the number of spares in the loop
Hold at a shared or regional store
When the list chooses it
A shared store can deliver within the wait you accept, and holding your own does not pay back
Stock levels
None on site; confirm the store holds one and will reserve it
Buy on demand
When the list chooses it
Routine parts, and parts you can get within the wait you accept when stock does not pay back
Stock levels
None; min and max are zero

A filled-in example

Illustrative, not a benchmark

An example: a bottling line with 6 assets and 12 parts, 120 scheduled hours a week, a downtime cost of 3,000 an hour on the filler and capper and the template's example settings (illustrative numbers).

How many of the example's gearbox bearing sets to hold

m = uses a year × lead time in weeks ÷ 52

With S held, the chance one is there when needed = P(demand in the lead time ≤ S − 1), Poisson with mean m.

m =1.5 × 10 ÷ 52= 0.29

1 held74.9%below the target
2 held96.6%below the target
3 held99.7%meets the target

Stocking policy for the example's 12 parts

Held on site: 7 parts
Repair and cycle: 1 part
Shared store: 1 part
Buy on demand: 3 parts

Only 3 held meets the 99% target, so min 2 and max 3. One stock-out would wait 72 hours for an expedited set and cost 154,286 in lost production.

Filler drive gearbox bearing set: used 1.5 times a year, 10-week lead time, so m = 1.5 x 10 / 52 = 0.29. With 1 on the shelf the chance one is there when needed is 74.9%, with 2 it is 96.6%, with 3 it is 99.7%. The class target is 99%, so min 2 and max 3. One stock-out waits 72 hours for an expedited set: 51.4 production hours x 3,000 = 154,286.
Capper chuck: single source with an 8-week lead time, but a 2-week repair, so it is repair and cycle: m = 3 x 2 / 52 = 0.12, min 1, max 2 in the loop, and the action is to qualify a second source.
Filler main drive motor: 14,000, used once in 20 years, and the regional store can deliver in 6 hours, inside the 8 hours accepted for a Critical part. Holding one would take about 22 years to pay back, so it stays in the shared store.
Compressor air-end: 9,500 and single source, yet Routine, because a standby compressor carries the load. Buy on demand.
Case packer servo drive: obsolete, so the action is to find a replacement or make a last-time buy before the one on the shelf is used.

7 parts held on site, 1 in a repair loop, 1 at the shared store and 3 bought on demand. Stock value at max is 26,516; 5 parts need ordering now for 12,340. With nothing held, the downtime cost would be about 2,692,714 a year; with this policy it is at most 20,621.

Common mistakes

Stocking by price instead of by consequence
An expensive part on an asset with a standby unit can be routine, and a 240 label sensor that stops a line pays back in days. Start from what the failure does, then look at the price.
Holding one of everything
One on the shelf feels safe, but a part used 1.5 times a year with a 10-week lead time is missing about one time in four with one held. Work out the chance; sometimes one is far too few, and for rare cheap-to-get parts one is too many.
Counting planned replacements as usage
Parts changed on a schedule are known in advance and can be ordered for the job. Counting them as random demand inflates the minimum for every part on a PM.
Relying on a shared store nobody has checked
A regional store or a supplier's shelf only counts if it holds the part today and will deliver in the hours you assumed. Confirm it, and write down the agreement.
Setting it once
Lead times change, suppliers drop parts and assets get backups. A part that is now obsolete or single source can move from routine to critical without anyone noticing until it fails.

Download the template

Free to use: print it, copy it and edit it for your team. Enter your name and work email once to download.

Run this template in LeanSuite

In LeanSuite, technicians record the spare parts used in each repair on its emergency work order, next to what was repaired, and close it from any device, building a searchable repair history for every asset.

See it in a 45-minute demo

In the lean glossary

Free tools

Guides

FAQ