A heavy equipment parts room has two jobs that pull against each other. It should never stop a machine for a $40 part, and it shouldn’t tie up cash in parts nobody touches for years. Min/max levels and a short critical spares list do most of that work.
Sort parts before setting levels
Different parts need different rules. Sort every part number into one of four groups first.
- PM consumables. Filters, oils, grease and DEF filters. Demand follows the PM schedule, so you can forecast it.
- Wear parts. Bucket teeth, edges, tires and undercarriage. Demand follows hours and ground conditions.
- Critical spares. Hoses on main circuits, starters, alternators, key sensors and belts. Failure is hard to predict, but a stockout parks the machine.
- Insurance spares. Final drives, swing motors and transmissions. Expensive, rarely needed, and often better sourced than stocked.
Most of the shelf should be the first two groups. The second two are where the judgment calls live.
Decide what earns a shelf
A part earns a place as a critical spare when a stockout costs more than holding it.
Stockout cost = lead time in days × downtime cost per day
Expected stockout cost per year = chance of a failure in a year × stockout cost
Holding cost per year = part price × annual carrying rate
The carrying rate covers capital, space, handling and the chance the part becomes obsolete. Your controller can give you a figure; the examples here use 20% a year.
Worked example one: a sensor. A $400 sensor fits eight machines in your fleet and takes 5 days to arrive. A machine down costs $1,200 a day in rental and crew time. Across eight machines, you expect to need one most years. Stockout cost is 5 × $1,200 = $6,000. Holding cost is $400 × 20% = $80 a year. Stock it.
Worked example two: a final drive. A $15,000 final drive takes 10 days to arrive, so a stockout costs 10 × $1,200 = $12,000. You expect one failure across those machines every ten years or so, so the expected stockout cost is about $1,200 a year. Holding cost is $15,000 × 20% = $3,000 a year. Don’t stock it. Know which supplier has one, and have a rental plan for the 10 days.
Decision rule. Stock a part when its expected stockout cost is higher than its holding cost. Favor parts that fit more than one machine, because one spare then covers several chances of failure.
The min/max formulas
For parts you use regularly, three numbers run the shelf.
Usage per week = units used over the period ÷ weeks in the period
Min (reorder point) = usage per week × lead time in weeks + safety stock
Max = min + order quantity
Safety stock covers the weeks when usage runs high or the supplier runs late. A simple version is usage per week × safety weeks, with more safety weeks for a less reliable supplier or a longer haul from the warehouse.
Worked example. A shop runs 12 excavators on the same engine. Each runs about 40 hours a week and needs an oil filter every 250 hours. That is 12 × 40 = 480 engine hours a week, or about 2 filters a week. Lead time is 1 week, and you want 1 week of safety stock.
- Min = 2 × 1 + 2 = 4 filters
- Filters come in boxes of 6, so the order quantity is 6
- Max = 4 + 6 = 10 filters
When the count drops to 4, order a box. The shelf never holds more than 10, and it never runs dry unless usage doubles or the supplier misses by two weeks.
Let the PM schedule drive demand
History is a fair guide for wear parts. For PM consumables you can do better, because you already know which services are coming.
Count the services due inside your lead time plus safety window and add up their parts. If six excavators reach 500 hours and three loaders reach 1,000 hours in the next four weeks, the parts lists for those nine services are your demand. The PM interval planner projects each machine’s next four services from its current hours and weekly use, which is the input this needs.
Build kits for each model and interval: a 500-hour excavator kit, a 1,000-hour loader kit. Reserve a kit when the work order opens, so the parts are bagged and labeled before the mechanic needs them. A mechanic who spends 20 minutes hunting for a fuel filter costs more than the filter.
When a machine’s hours jump, because it moved to a busier job or a second shift, its services come sooner and its parts demand follows. Forecasting from the schedule picks that up; forecasting from last year’s history doesn’t.
Wear parts follow the ground
Bucket teeth, edges and tires wear by the hour, but how fast depends on the material. Track wear life per job type, not per model, and the forecast becomes simple.
Wear parts per week = machine hours per week ÷ wear life in hours
Worked example. Four wheel loaders on a caliche pit near Belen each run 50 hours a week, 200 hours in total. On that pit, a set of bucket teeth lasts about 400 hours. That is half a set a week. With a 2-week lead time and 1 week of safety stock, the min is 0.5 × 2 + 0.5 = 1.5, so round up to 2 sets. If the teeth come one set to a carton, order 2 sets at a time, for a max of 4.
Move the same loaders to a sand job and the teeth may last twice as long. The levels should change with the job.
Build the first critical list
Most shops can build a usable critical spares list in an afternoon:
- List your models, grouped by how many machines share each one.
- Ask the mechanics which parts stopped a machine for more than a day in the last two years.
- Check lead times with your suppliers for each of those parts, today, not from memory.
- Run the test above: expected stockout cost against holding cost.
- Start small. Ten to twenty parts that clearly pass, reviewed each quarter.
Slow movers and multiple yards
For parts used a few times a year, the formulas round to zero. Use judgment instead:
- critical spares: min 1, max 1, or 2 if many machines share the part
- review slow movers every quarter against the machines you still own
- keep spares at the yard nearest the machines that use them, and know what each yard holds
A filter in Albuquerque doesn’t help a loader near Lovington today. If you run two yards, decide which parts live at both and which travel on the service truck.
Service truck stock is inventory too. A truck carrying filters, hoses and fittings for the Permian jobs holds real money, and it belongs in the counts and the min/max like any shelf.
Count and review on a rhythm
- Cycle counts. Count high-value and critical parts monthly, regular movers quarterly, everything else twice a year.
- Stockout log. Record every time a machine waited on a part, with the hours lost.
- Dead stock list. Parts with no movement in 12 months. Return what you can, and stop reordering parts for machines you sold.
- Fill rate. Work order lines filled from stock ÷ total work order lines.
Worked example. Last month the shop’s work orders had 400 parts lines, and 360 came off the shelf, a fill rate of 90%. The 40 that didn’t are the list to read. Some are one-off repairs nobody could predict. Some are a min set too low, and those get raised this week.
The stockout log is the most useful of the four. A part that stopped a machine twice in a quarter has earned a shelf, whatever the formula said.
Mistakes in heavy equipment parts rooms
- Set once, never revisited. The fleet changed and the levels didn’t.
- Parts for sold machines. Unique parts become dead stock the day the machine leaves.
- Duplicate numbers. The same filter stocked under the manufacturer’s number and an aftermarket number.
- No bin locations. A part on the shelf that nobody can find is a stockout.
- Parts without work orders. Shrink you only discover at the annual count.
- Off-spec substitutes. A cheaper filter that doesn’t meet the manual’s specification can cost you a warranty claim.
A starting checklist
- Every part number sorted into one of four groups
- A critical spares list built with the stockout and holding cost test
- Min and max set for PM consumables and wear parts
- PM kits built for each model and interval
- Kits reserved when work orders open
- Cycle counts on a calendar
- A stockout log the shop actually fills in
- Service truck stock counted with everything else
Tie parts to work orders
Parts and inventory reserves parts against each work order when it opens and updates the reorder point, so the min/max math runs off real demand. It sits within maintenance scheduling, on the Shop plan, beside preventive maintenance and warranty tracking.
