What Is Corrective Maintenance?

Corrective maintenance is repair work carried out after a fault has been recognised, intended to restore an asset to a state in which it can perform its required function. That phrasing follows the maintenance-terminology standard EN 13306, and the load-bearing word in it is after. The fault might surface as a breakdown mid-job, a defect spotted during an inspection, or a problem raised on a service ticket - what makes the work corrective is that the failure or defect already exists when the job is created. When the repair happens is a separate decision entirely.

This guide covers the precise definition and the terms it gets confused with, the four types of corrective maintenance, what triggers a job, how corrective work compares with preventive and predictive maintenance, its real advantages and costs, what belongs on a corrective work order, the metrics that show whether reactive work is under control, and how to convert unplanned repairs into planned ones.

What you will learn

• What counts as corrective maintenance
• The types of corrective maintenance
• What triggers corrective work
• Corrective vs preventive vs predictive
• Where corrective maintenance earns its place - and where it hurts
• The corrective maintenance work order
• Measuring corrective maintenance
• Turning unplanned corrective work into planned work
• Corrective maintenance examples across asset types
• FAQ

What counts as corrective maintenance

The test is simple: was there a known fault before the job existed? A pump stripped and rebuilt because its bearing is screaming is corrective. The same pump stripped and rebuilt on a 2,000-hour schedule while running perfectly is preventive. The work on the bench looks identical; the trigger is what classifies it.

One disambiguation: “corrective maintenance” is also a category in software maintenance, meaning defect fixes after release. This page covers the equipment sense.

The surrounding vocabulary is where most confusion lives:

• Breakdown maintenance is the subset of corrective work that follows an actual failure. Every breakdown job is corrective; not every corrective job follows a breakdown.
• Reactive maintenance describes the posture rather than the work - unplanned, urgent, driven by whatever just stopped.
• Remedial maintenance is a common synonym for corrective, used more in facilities and IT support contracts.
• Unplanned maintenance is the scheduling label: any job that was not on the plan, which in most organisations is mostly corrective work. See planned vs unplanned maintenance for the split.
• Corrective action (as in CAPA) is a quality-management term about fixing the cause of a nonconformity in a process. Related in spirit, but it is not a repair job on an asset.

The types of corrective maintenance

Corrective work splits into four very different experiences, and calling them by the right name is what stops “we do a lot of maintenance” from hiding a serious planning problem.

Immediate or emergency corrective maintenance follows a failure that stops something important. One of the trucks is off the road with a load booked, so the repair jumps every queue and costs whatever it costs. This is what people usually mean by reactive maintenance, and it is the type that should be rare.

Planned corrective maintenance covers a fault that is already known but whose repair is scheduled deliberately. A slow hydraulic leak found at inspection: the defect is logged, the seal kit is ordered, and the machine goes into the workshop on a day it was not needed anyway. The fault is real, so the work is corrective - but the timing belongs to you, which is why the cost profile looks nothing like the emergency version.

Deferred corrective maintenance is a logged defect that has been postponed, and two very different situations hide under one label. Deferred by choice means the fault is genuinely low-impact and the repair waits for a sensible moment. Deferred by constraint means the part, the budget or the person is missing and the job sits because it must, not because it should. Both accumulate into a maintenance backlog, and a backlog nobody ages is a list of future emergencies.

Temporary fix versus permanent repair. A strap, a jumper wire or a hire-in replacement can get you through the week, but none of them is the repair. Standards vocabulary calls the temporary version palliative and the real one curative; the practical point is that a temporary fix must leave a follow-up job behind. If it does not, it silently becomes permanent, and six months later nobody remembers that the guard on that machine is held on with a cable tie.

What triggers corrective work

Corrective jobs arrive through a small number of doors, and knowing which one a job came through tells you a lot about how urgent it really is:

• An operator or user reports a symptom - a noise, a smell, a warning light, a tool that will not hold charge. The most common entry point by far.
• An inspection or condition report flags a defect before anything has stopped. This is the door that produces planned corrective work.
• A test, calibration or safety check fails - the item is now unusable in the eyes of a regulator even if it still runs.
• A monitoring reading crosses a limit - a temperature, a pressure, a vibration trend. See predictive maintenance for where that data comes from.
• A supplier, recall or warranty notice arrives, naming a defect you had not found yourself.
• A customer, tenant or client complains - in facilities and rental work, this is frequently the first anyone hears of a fault.

Two things follow from this list. First, the trigger sets the urgency, not the fault itself: an identical bearing failure is an emergency on a machine running a shift and a routine job on a spare unit in the store. Second, the quality of the first report decides how much of the repair can be planned. Which asset, what symptom, since when, a photo, and whether it is still safe to use - five fields that turn a vague complaint into a job somebody can order parts for. A regular inspection schedule is the cheapest way to move faults from the first door to the second.

Corrective vs preventive vs predictive

Preventive maintenance services working equipment on a schedule; corrective maintenance repairs equipment with a known fault. Neither is “the right one” - they are positions on a spectrum, and the sensible mix depends on the asset.

Predictive maintenance sits between prevention and correction, using condition data to time work just before failure. Run-to-failure is corrective by deliberate choice rather than by accident, and it is a perfectly respectable strategy when the decision inputs support it: low replacement cost, low criticality, a spare already in the cupboard, no safety consequence, and redundancy elsewhere. Fail any of those tests and it is not a strategy, it is a gamble.

You will see planned-to-reactive ratios quoted as targets - 80/20 is the number most often repeated in maintenance guidance. Treat it as a rule of thumb rather than a measured law: the right split depends entirely on asset criticality, and a workshop full of cheap hand tools has no business chasing the same ratio as a production line. The expensive failure mode is not choosing corrective maintenance - it is defaulting to it for critical kit because nobody ever set up anything else.

Where corrective maintenance earns its place - and where it hurts

The advantages are real. There is no scheduling overhead - no calendar to maintain, no routes to plan, no reminders to chase. Nothing is spent servicing components that were perfectly fine, and every cheap part gives up its full remaining life instead of being binned at 60% wear. For a large population of low-value items - light fittings, hand tools, cables, a spare monitor - a preventive programme would cost more than simply fixing things when they break. Corrective-first is the rational answer there, not a failure of discipline.

The disadvantages compound, which is why they are underestimated. You do not control the timing, so downtime lands at the moment the asset is busiest. And the invoice is only part of the bill. The real cost anatomy of an unplanned corrective job includes:

• Rush parts and expedited freight - next-day pricing on a part that would have cost half as much on a normal order.
• Overtime for the people doing the repair, and idle labour for everyone downstream waiting on it.
• Hire-in replacement to cover the gap, often at day rates that dwarf the repair.
• Secondary damage - one failure taking something else with it, because a cheap component let go inside an expensive assembly.
• The lost job - the delivery not made, the tenant not satisfied, the site day written off.
• Safety exposure, when a faulty asset stays in use because nobody quarantined it.

There is a compounding effect on top of all this: a team permanently in reactive mode never gets the slack to become a planned one. Every hour spent on today’s emergency is an hour not spent removing the cause of tomorrow’s. That is the trap maintenance teams and facilities management operations describe most often, and it is why the ratio of planned to unplanned hours matters more than any single repair bill.

The corrective maintenance work order

A corrective job that goes well follows the same shape almost everywhere. Each step is really a set of fields on a work order.

1. Report. Whoever finds the fault records which asset, what the symptom is, when it started, and whether the asset is still safe to use - with photos, because a photo settles in one second what three messages cannot. A report naming no asset is not a work order, it is a rumour.

2. Triage and priority. Someone decides how urgent this is. A plain four-level scale works better than an elaborate matrix:

Rank that against how critical the asset is, and the two together give a defensible order of work. Triage is also where a faulty asset gets quarantined or tagged out so nobody signs it out mid-repair - set its status to in repair and the problem disappears. Skipping this step is how a machine with a known fault ends up back on site.

3. Repair. In-house or sent out, parts sourced, the work done. This is where the most common real-world state deserves an honest name: awaiting parts. A job stalled on a supplier is not “in progress”, and pretending otherwise is how backlogs become invisible.

4. Close-out. What was found, what actually failed and why, parts consumed, labour hours, total cost, whether the asset returned to service, and whether a follow-up job is needed - which is exactly where a temporary fix gets converted into a real one. The close-out is the step teams skip and later regret; it is the only part that pays off the next time the same asset misbehaves.

Measuring corrective maintenance

A handful of numbers tell you whether corrective work is a manageable part of the operation or the whole of it. None of them exist unless jobs are logged against the asset.

• MTTR (mean time to repair) - how fast repairs land. A long MTTR is usually a parts or access problem rather than a skills problem, so chase spare parts management before training.
• MTBF (mean time between failures) - whether the same asset keeps coming back. A short MTBF on one item says the repairs are treating symptoms.
• Planned maintenance percentage - the share of maintenance hours that were scheduled rather than reactive. This is the single number that tracks whether you are climbing out of the hole or sliding further in.
• Backlog size and age - how many corrective jobs are open, and how old the oldest is. Age matters more than count: a small backlog full of six-month-old defects is worse than a large fresh one. See maintenance backlog.
• Repeat-fault rate per asset - how many jobs on the same item, same failure mode, inside a year. Anything above two is a review, not a repair.
• First-time-fix rate - the share of jobs closed on the first visit. Low numbers usually mean the initial report was too thin to prepare for.
• Corrective spend against replacement value - the input to the repair-or-replace call, and the trigger for refurbishment or retirement.

Read alongside downtime and asset uptime, these say what to change rather than just what happened. A rising planned percentage with a flat backlog is progress. A flat planned percentage with a growing backlog means the deferrals are doing the work that scheduling should.

Turning unplanned corrective work into planned work

The goal is almost never “less corrective maintenance”. It is the same repairs, happening on your schedule instead of the failure’s. Six habits do most of the work:

1. Make faults trivial to report. Most defects are noticed days before they stop anything, by someone who did not have an easy way to say so. Remove the friction and faults surface while they are still planned corrective work.
2. Review repeat faults instead of re-fixing them. When the same item fails the same way a third time, stop and ask why several times over - a 5 Whys pass, or a small FMEA on the critical assets. If the answer is a design, environment or usage problem, the repair was never going to help.
3. Convert recurring faults into scheduled tasks. Any failure mode that has now happened three times has earned a preventive task with a date on it.
4. Stock the spares that actually stop work, not every part. A short list of critical spares does more for MTTR than a large stores budget spread evenly.
5. Set the repair-or-replace rule before you need it. Repair cost against replacement value, age against useful life, repeat-fault history, and whether the item is still under warranty - decide the thresholds calmly, so the call is not made in a panic with a job on the line. Warranty tracking belongs in that decision, because an in-warranty repair paid for out of your own budget is pure waste.
6. Work the backlog by age, not by volume. The loudest requester is rarely the oldest defect.

None of this requires a heavyweight system. It requires that faults get logged against the asset they belong to, which is the same habit that makes equipment tracking work in the first place.

Corrective maintenance examples across asset types

The pattern is identical across industries; only the kit changes.

• Vehicles and plant. A driver reports brake noise on a van (emergency corrective - safety, off the road until fixed). A slow hydraulic leak found on a trailer at inspection (planned corrective - seal ordered, booked in).
• IT hardware. A laptop whose battery no longer holds charge (planned corrective - loan machine issued, battery ordered). A docking station that stopped charging, swapped for a spare from the shelf (run-to-failure by design). A monitor with a dead port (deferred - it still has two working ones).
• Facilities. A heater that will not fire in October (emergency). A fire door that sticks (P1 despite looking cosmetic, because it is a life-safety item). A dripping tap (deferred by choice).
• Tools and site kit. A cracked guard on a grinder among the power tools (emergency - quarantine the tool, no guard, no use). A generator that will not start on site, covered by a hire-in while it is repaired (temporary fix, with a follow-up job attached). An extension lead that failed its electrical test (planned corrective - re-terminate or bin).
• Catering. A fridge running two degrees warm (emergency in food service, whatever the thermometer looks like).
• Lab and medical. An instrument that failed calibration (planned corrective - unusable for reporting until adjusted, but nothing is broken).

Note how often the same physical fault lands in different types depending on context and redundancy. That is the taxonomy doing its job.

FAQ

What are the types of corrective maintenance?

Four, in practice. Immediate or emergency corrective work follows a failure that stops something important and jumps every queue. Planned corrective work covers a known fault whose repair is booked into a slot, with parts ordered first. Deferred corrective work is a logged defect that has been postponed, either by choice or because parts, budget or people are missing. And a temporary fix is a workaround that keeps the asset usable until the permanent repair happens - it only counts as corrective maintenance done if it leaves a follow-up job behind.

What is the difference between corrective and preventive maintenance?

Preventive maintenance services working equipment on a schedule, before anything has gone wrong. Corrective maintenance repairs equipment that already has a known fault. The trigger is the difference: a date, a meter reading or a usage count starts preventive work, while a failure or a reported defect starts corrective work. Most teams run both, and the mix should follow how critical the asset is rather than a fixed ratio.

What should a corrective maintenance work order include?

At the point it is raised: which asset, the symptom in plain words, when it started, photos, whether the asset is still safe to use, and a priority. At close-out: what was found, the failure mode and likely root cause, parts consumed, labour hours, total cost, whether the asset went back into service, and whether a follow-up job is needed. The close-out half is the part teams skip and later regret, because it is the only part that helps the next time the same asset misbehaves.

Is corrective maintenance the same as reactive maintenance?

They overlap but are not identical. Reactive maintenance usually means unplanned response to a breakdown - the asset stops, someone scrambles. Corrective maintenance is the wider term for any repair of a known fault, including planned corrective work where a defect is logged during an inspection and the repair is scheduled for a convenient slot. All reactive work is corrective; not all corrective work is reactive.

What is an example of corrective maintenance?

A delivery van develops a grinding noise, the driver reports it, and the brakes are repaired - that is corrective maintenance triggered by a breakdown symptom. So is replacing a cracked guard found during a routine inspection, swapping a laptop battery that no longer holds charge, fixing a leaking site generator, or re-terminating a damaged extension lead that failed its test. The common thread is that the fault exists before the work is raised.

When is a corrective-only approach acceptable?

Run-to-failure is a legitimate strategy for assets that are cheap to replace, non-critical, or backed by spares - light fittings, basic hand tools, a spare monitor. It becomes a problem when it is the unplanned default for critical or safety-related equipment, where a failure means stopped work, a missed job, or a hazard rather than a shrug and a replacement from the cupboard.

Tools that make this easier

AMPthilly puts the corrective record where it belongs - on the asset. Anyone can report an issue against an item with a description, photos and a category (damage, missing, needs maintenance, needs replacement), and scanning the asset’s QR label with a normal phone camera opens its profile in the browser with no app to install, so the person holding the broken thing raises the ticket on the spot rather than remembering to later.

From there, tickets sit in a queue with statuses including in review, in progress, awaiting parts and resolved - so a job stalled on a supplier is visible instead of quietly forgotten. Each ticket carries comment threads, and repair invoices or receipts attach directly to it. Set the asset’s status to in repair and it will not be signed out mid-repair. Purchase price, supplier, warranty end date and expected useful life live on the same record, which is where the repair-or-replace call actually gets made. Ticket history and the full audit trail stay on the asset permanently, and everything exports to CSV when you want to count it elsewhere. The free plan covers 3 users and 25 assets with no card required, which is enough to start logging faults before you formalise anything.

The takeaway

Corrective maintenance is any repair triggered by a fault that already exists - and the useful question is never whether you do it, but whether it arrives as an emergency or as a booked job. Name the four types so deferred defects cannot hide, capture enough on the first report to plan the repair, close jobs out with the failure mode and cost, and watch the planned percentage alongside the backlog’s age. Do that for a few months and the pattern shows itself: a short list of assets generating most of the emergencies, each one a candidate for a scheduled task, a stocked spare, or a replacement.

Related terms

• Predictive Maintenance - condition data times the work just before failure
• Preventive Maintenance - scheduled servicing of equipment that has not failed
• Planned vs Unplanned Maintenance - the scheduling split corrective work sits across
• Work Order - the scheduled unit of repair work a fault report becomes
• Service Ticket - the report that starts most corrective jobs
• Maintenance Backlog - where deferred corrective work accumulates
• MTTR - mean time to repair, the headline downtime measure for corrective work
• MTBF - mean time between failures, computed from logged corrective history