Equipment Criticality Analysis for Small Plants: The Simple ABC Method

Why criticality ranking matters

Most small plants treat every machine the same way: everything gets a monthly PM, or nothing does. Both approaches waste effort. If your team of three techs spends an hour a month greasing a backup exhaust fan that could sit broken for a week without anyone noticing, that's an hour not spent on the compressor that stops the whole line when it goes down.

The math is simple. A team with 40 hours of PM capacity per week and 80 assets can't give everything deep attention. Without a ranking, attention gets allocated by habit — whichever machines are closest to the shop, loudest, or newest. The machines that actually hurt you when they fail get the same treatment as the ones that don't.

A criticality ranking fixes the allocation. It tells you where to spend PM hours, which spare parts to stock, and which overdue work order to chase first on a busy Monday. You don't need failure-rate data or an RCM study to get 90% of the value. You need three questions and an afternoon.

The simple ABC method

Every asset gets one of three grades:

A — Production stops or someone gets hurt. If this machine fails, the line stops, product is scrapped, or there's a real safety or environmental risk. There is no workaround. Think: the single air compressor feeding the whole plant, the boiler, the only CNC that runs your highest-volume part, the ammonia system in a food plant.

B — Production degrades or a workaround exists. Failure hurts, but you limp along. You have two packaging lines and one can absorb the other's volume at overtime cost. The forklift fleet has a spare unit. A conveyor failure means manual handling for a shift — painful, not fatal.

C — Run to failure is acceptable. Failure is an inconvenience you fix when you get to it. Office HVAC, a utility fan, the second-string bench grinder, lighting circuits. If it broke Friday and got fixed Wednesday, nobody outside maintenance would know.

The point of only three grades is speed and honesty. Five-tier systems invite arguments about whether something is a 3 or a 4. With three tiers, the question is blunt: does failure stop us, slow us, or just annoy us?

How to score: three questions per asset

Grading by gut feel works for the obvious cases, but you'll hit assets people disagree about. For those, score each asset on three questions, 1–5 each:

1. Downtime impact (1–5). If this fails right now, what happens to production? 1 = nothing for days. 3 = one area slows down. 5 = the whole plant stops or product is scrapped.
2. Safety and environmental risk (1–5). Can a failure hurt someone or trigger a reportable release? 1 = no plausible risk. 3 = minor injury possible. 5 = serious injury, fire, or environmental incident.
3. Repair cost and lead time (1–5). How expensive and slow is recovery? 1 = parts on the shelf, fixed in an hour. 3 = parts in a few days, moderate cost. 5 = specialist vendor, 8–12 week lead time, five-figure repair.

Sum the three scores. Total 11–15 is an A. Total 7–10 is a B. Total 3–6 is a C. One override rule: any asset that scores a 5 on safety is an A regardless of the total.

Here's what that looks like for a few typical assets:

Do the scoring as a group: the maintenance lead, one senior tech, and someone from production. Production's input matters because they know which machine failures actually stop shipments — maintenance sometimes overrates the machines they enjoy working on. Budget two minutes per asset. Eighty assets is under three hours, coffee included.

What to do with each class

The ranking only pays off if it changes how you maintain things. Here's the playbook:

A assets get the full treatment. Meter-based PMs (runtime hours, cycle counts) plus calendar-based PMs as a backstop. Stock the critical spares on site — bearings, belts, contactors, anything with a lead time over a week. Attach the manuals and PM checklists to the asset so any tech can do the work. When an A asset's PM comes due, it doesn't slip. When an A asset has an open work order, it's at the top of the board.

B assets get calendar PMs. A sensible interval — monthly or quarterly depending on duty — with a standard checklist. You don't need meter tracking or a deep spares inventory; you need consistency. If the week gets crushed, a B asset's PM can slip a few days. An A asset's can't. That's the operational difference the letters buy you.

C assets get inspection-only or planned run-to-failure. “Run to failure” is a strategy, not neglect — the difference is the plan. For each C asset, know how you'll fix it when it fails: which vendor, roughly what cost, roughly what downtime. A quick visual check on a quarterly walk-through is plenty. If a C asset starts failing every other month, that's your signal to re-score it, not to quietly start PMing it off the books.

Common mistakes

Everything gets rated A. The most common failure. If 60% of your assets are A, you don't have a priority list — you have the same undifferentiated pile with letters on it. A useful distribution for a small plant is roughly 15–20% A, 30–40% B, and the rest C. If your A list is longer than that, re-ask the blunt question: does this failure really stop production, or does it just feel important?

Ranking once and never revisiting. Criticality changes. You add a second compressor and the first one drops from A to B overnight. A customer doubles orders on one line and its equipment moves the other way. A ranking from three years ago quietly misallocates every PM hour you spend.

Ignoring lead times. A $400 gearbox with a 14-week lead time is more dangerous than a $4,000 motor your supplier stocks locally. Teams score cost and forget time. The third question exists precisely because recovery time, not repair price, is what turns a failure into a crisis.

Keep it alive

Put a recurring task on the calendar: review the criticality list once a year. Also re-score whenever you add or remove equipment, and after any surprise failure that hurt more than its rating said it should — that's the ranking telling you it's wrong. The whole review takes an hour once the first pass exists. That hour is the cheapest reliability work you'll do all year, because it points every other maintenance hour at the machines that actually matter.