What is VED analysis in spare parts management?

VED classifies parts by the consequence of a stockout: Vital parts stop production or create a safety hazard when unavailable, Essential parts degrade output or quality without an immediate stop, and Desirable parts cause inconvenience only. It answers whether to stock a part, independent of its price. A $12 sensor can be Vital; a $9,000 gearbox can be Vital too, even at near-zero usage.

What is the difference between ABC and VED classification?

ABC ranks parts by annual usage value, A-items carry most of the spend, and determines how tightly to control what you stock. VED ranks parts by production consequence and determines whether to stock at all. Used together as a matrix, they produce a stocking policy per part: Vital items get stocked regardless of value, cheap Desirable items may not be stocked at all.

How do you set min and max levels for spare parts?

Min equals expected usage during the supplier lead time plus safety stock, with the safety buffer sized by demand variability, lead-time variability, and criticality. Max equals min plus a sensible order quantity shaped by price breaks, packaging multiples, and shelf life. Review both at least annually, usage patterns shift as equipment ages and PM schedules change.

How much inventory does spare parts carrying cost?

Carrying cost, capital, space, shrinkage, and obsolescence combined, is commonly cited at roughly 15–30% of inventory value per year. Even at the low end, a $500,000 storeroom costs about $75,000 annually just to hold, which is why maxes and slow-mover reviews matter as much as avoiding stockouts.

Should you stock a spare part that has never been used?

Sometimes yes. For a Vital part with a long lead time, a drive, a gearbox, a custom fabrication, zero usage history does not mean zero risk; it means the one failure that occurs takes the line down for the full lead time. The decision should be explicit: stock it, or put a written supplier availability or consignment agreement in place instead.

What is phantom inventory in a storeroom?

Phantom inventory is when the system shows stock that the shelf does not have, usually because parts left without being checked out. It silently defeats min-max, since the reorder never triggers. The fixes are disciplined check-in/check-out and regular cycle counting so record errors are caught in weeks, not at the annual count.

Spare Parts Inventory: Stock the Right Parts Without Freezing Cash

A practical system for deciding which spares to stock, how many, and when to reorder, using VED criticality, ABC value, and min-max levels.

Spare parts inventory management is the practice of deciding which parts to stock, how many, and when to reorder, based on how critical each part is to production and how hard it is to get. Done well, it prevents stockout downtime without freezing cash in shelves of parts that never move.

Every storeroom tells the same story from two directions. Maintenance remembers the night a $40 seal was out of stock and the line sat for nine hours waiting on a courier. Finance remembers the annual count that found six figures of parts nobody has touched in five years. Both are right, and both problems have the same root cause: parts were stocked by anecdote instead of by classification. This guide covers the two classifications that matter, the min-max mechanics, and the habits that keep the storeroom honest.

Why is spare parts inventory hard to get right?

Because the cost of a stockout and the cost of overstock are paid by different people at different times. A stockout on a critical part turns a two-hour repair into a multi-day outage, the most expensive form of machine downtime because the clock runs while the part travels. Overstock is quieter: cash sits on shelves, parts shelf-age past usability, and carrying costs, commonly cited at roughly 15–30% of inventory value per year once you count capital, space, shrinkage, and obsolescence, accumulate invisibly.

Guessing splits the difference badly. You end up overstocked on cheap, easy-to-get parts (because they are easy to buy) and understocked on the expensive, long-lead-time parts that actually take lines down. Classification fixes the guess: stock decisions follow criticality and lead time, not purchasing convenience. Spare parts are the largest chunk of most plants' MRO inventory so this is usually where the money is.

How do you classify spare parts by criticality?

You classify each part on two independent axes: how badly production suffers without it (criticality) and how much money flows through it (value). The standard tools are VED and ABC analysis, used together.

VED scores consequence: Vital parts stop a line or create a safety hazard when unavailable; Essential parts degrade output or quality but don't stop production immediately; Desirable parts cause inconvenience only. ABC scores annual spend: A-items are the small fraction of parts carrying most of the annual usage value, C-items are the long cheap tail. Neither axis alone is enough, a $12 proximity sensor can be Vital, and a $9,000 gearbox can sit at zero usage for years and still be the most important part in the building.

A working criticality matrix. VED (consequence of stockout) sets whether to stock; ABC (annual value) sets how tightly to control what you stock.

How do min-max levels actually work?

Min-max is a reorder rule: when on-hand quantity drops to the minimum, you order enough to return to the maximum. The min covers expected demand during the resupply lead time plus a buffer; the max caps how much cash sits on the shelf.

The two numbers come from four inputs: average usage rate, usage variability, supplier lead time, and lead-time variability. In practice:

Min = expected usage during lead time + safety stock. The buffer exists because both demand and lead time wobble; the wobblier they are, the bigger the buffer, for Vital parts especially.
Max = min + a sensible order quantity, usually shaped by price breaks, packaging multiples, and shelf life.

The min-max sawtooth. Stock drains with usage, an order fires at min, and safety stock absorbs demand during the lead-time gap. Deeper dips into the safety band mean the min is set too low.

Two failure modes dominate. First, mins set once and never revisited: usage changes when equipment ages or a PM schedule changes, and a min that was right in 2022 quietly becomes wrong. Second, phantom inventory: the system says three on hand, the shelf has zero, and the min never triggers. That is a records problem, and cycle counting plus disciplined check-in/check-out is the only cure.

What does a working spare parts program look like?

It looks like a short sequence of decisions applied to every stocked part, then maintained. Here is the framework:

Build the parts catalog. One record per part with description, location, supplier, cost, and lead time. Kill duplicates, the same bearing under three part numbers is how you stock nine and find none.
Score VED criticality with maintenance and production in the room. The question is "what happens to output if this is unavailable for its lead time?", not "how expensive is it?"
Run ABC on annual usage value. Most systems can produce this in an afternoon. The A-list is where tight control pays; the C-list is where control costs more than it saves.
Set min-max from lead time and usage, with safety stock sized by criticality. Vital + long-lead parts get the deepest buffers; Desirable + short-lead parts may get none.
Decide the no-stock list explicitly. Not stocking a part is a decision with an owner, not an accident. For Vital A-items you choose not to stock, put a supplier agreement or consignment arrangement in writing.
Maintain the loop. Cycle count, review slow movers quarterly, and re-score criticality when equipment is added or retired. Track stockout events against downtime so the storeroom sees its own consequences.

What numbers should you track?

Track a small set that exposes both failure modes, stockouts and frozen cash:

Stockout-driven downtime. Every downtime event extended by a missing part gets flagged. This is the single most persuasive number in the building, and it also feeds your MTTR figure, since waiting on parts is usually the longest slice of repair time.
Inventory value and turns. Spares turn slowly by nature, many healthy storerooms sit around one turn per year or less, but the trend matters: value creeping up with no service-level improvement means the maxes are wrong.
Carrying cost. Commonly cited at roughly 15–30% of inventory value per year across industry; even the low end means a $500,000 storeroom costs $75,000 a year just to exist.
Reactive share of maintenance. The U.S. Department of Energy's Federal Energy Management Program O&M Best Practices guidance estimates plants that reduce reliance on reactive maintenance can see savings opportunities exceeding 30–40%. Fewer emergency repairs means calmer, more predictable parts demand, planned work orders name their parts in advance, which is the whole premise of maintenance planning and scheduling.

The common thread is connection: parts data means little in isolation. When storeroom records, work orders, and downtime logs live in one connected layer, Harmony's inventory and shortage intelligence module does exactly this kind of gap-spotting across systems (see the platform), the "$40 seal, nine hours down" story becomes a report instead of a legend, and the fix becomes obvious.