Emergency vs reactive vs preventive maintenance explained
- Trefnus
- Aug 13
- 18 min read
Published: 12 August 2026 | Last reviewed: 12 August 2026
Every maintenance team divides its work into jobs it chose to do and jobs it was forced to do. Preventive maintenance is planned before a failure occurs. Reactive maintenance addresses a fault that has already happened but can safely wait in a queue. Emergency maintenance demands an immediate response because delay would create an unacceptable safety, compliance or operational risk.
Reactive vs preventive maintenance is the comparison most organisations make, but treating emergency work as merely urgent reactive work hides the cost of losing control over your own schedule.
This guide sets out what separates the three approaches, what each genuinely costs once indirect effects are counted, and when each is the right choice. It also covers how to shift the balance of your workload towards planned work, without pretending unplanned work can ever be removed entirely.
What is the difference between emergency, reactive and preventive maintenance?
The three terms describe when a job is triggered, not how skilled the work is or which trade carries it out. The same engineer replacing the same bearing could be doing preventive, reactive or emergency work depending on the circumstances that put the job on the list.
Maintenance terminology at a glance Preventive maintenance: work carried out to prevent or delay a failure, triggered by time, usage or condition rather than by a fault. Corrective maintenance: work carried out to restore an asset after a fault or defect has been identified. Reactive maintenance: corrective work that was not planned in advance, where the fault can safely wait in a queue. Emergency maintenance: unplanned corrective work requiring an immediate response because delay is unacceptable. Run-to-failure: a deliberate decision to operate a suitable asset until it fails, rather than servicing it on a schedule. Planned and unplanned describe how the work is managed. Preventive, corrective and emergency describe why and when the work is triggered. The two dimensions overlap: a known defect can be fixed as planned corrective work, which is neither preventive nor an emergency. |
That final point is where most arguments about maintenance terminology start. A defect spotted during a routine inspection and fixed three weeks later on a booked visit is planned corrective work: it is not preventive, because the fault already exists, and it is not an emergency. Keeping the two dimensions separate stops you assuming that all planned work is preventive, which in most organisations it is not.
Preventive maintenance
Preventive maintenance is work carried out to prevent or delay failure, on a schedule you control. The trigger is a threshold rather than a fault: a date has arrived, an asset has run for a set number of hours, a filter has reached its service interval, or a statutory inspection is due.
Typical examples include annual boiler servicing, quarterly fire alarm testing, monthly emergency lighting checks, belt and filter replacement at fixed intervals, and lubrication schedules based on machine hours. Because the work is planned, parts can be ordered in advance, the right person can be booked, and the job can be timed for a quiet period.
Preventive maintenance is sometimes split further into time-based work, which runs on the calendar, and usage-based work, which runs on hours, cycles or mileage. Condition-based and predictive maintenance sit alongside these, triggered by a measured reading such as vibration or temperature rather than by elapsed time. For most small and medium businesses, time-based and usage-based schedules cover the majority of what is practical.
Reactive maintenance
Reactive maintenance is work carried out after something has failed or developed a fault, where there is no immediate danger and no immediate halt to operations. It is a subset of corrective maintenance, which is the broader term for any work that restores an asset after a defect has been identified. Corrective maintenance can be planned, if the defect is logged and scheduled, or reactive, if it is dealt with as it arises. Where the decision to wait for failure is deliberate rather than accidental, reactive work is described as run-to-failure.
A tap dripping in a staff kitchen, a flickering light in a corridor that still has adequate lighting, a damaged shelf edge in a stockroom, or a cooling fan running noisily but within tolerance are all reactive jobs. They need to be fixed, they will get worse if ignored, but they can be queued, batched and scheduled against your other commitments.
The important characteristic of reactive work is that you retain some control over timing. That control is what separates it from an emergency, and it is worth protecting.
Emergency maintenance
Emergency maintenance is unplanned work that cannot wait, because delay would create a risk to people, to the building, to legal or insurance compliance, or to the ability to trade. The initial response is dictated by the fault rather than by your schedule.
A burst pipe flooding a stockroom, a fire door that will not close, a failed refrigeration unit holding stock, a lift with a person inside, a power fault affecting a whole floor, or a gas leak are emergency jobs. So is a machine failure that stops a production line, even where nobody is in danger, because the commercial consequence of waiting is severe.
Emergency work is expensive not because emergency engineers charge more, although they usually do, but because every part of the job is decided under pressure. Parts may have to come from whoever can supply them fastest rather than from whoever offers the best price or the correct specification. Access is arranged at short notice. Other work is displaced. Records are written up afterwards, if at all.
Emergency response is not the same as emergency repair
One distinction saves a great deal of unnecessary cost, and it is missed surprisingly often. What has to happen immediately is usually the action that removes the danger, not the permanent fix.
A burst pipe needs isolating immediately. The pipework repair itself can often be scheduled once the water is off and the area is made safe. A fire door that will not close needs an interim control put in place immediately, which might be a temporary hold-open arrangement removed and the route managed, or the area taken out of use. The joinery work to correct the closer can then be planned properly.
Splitting the job into make safe and then repair converts most of the work from emergency rates back into planned work. It also produces a better repair, because the permanent fix is carried out by someone with the right parts and enough time. The mistake is to treat the whole job as an emergency simply because the first hour of it was.
Where the line between emergency and reactive work sits
Most organisations blur these two categories, and it costs them. If everything unplanned is logged as reactive, the emergency callouts disappear into a general repairs figure and nobody can see how much unplanned urgency is actually costing. If everything unplanned is treated as an emergency, the team spends its life responding at premium rates to jobs that could have waited until Thursday.
There is no universal definition that settles every case, so what you need is a triage rule you apply consistently. The one below works for most small and medium organisations. Treat an unplanned job as an emergency where delay would create a serious risk of harm, leave a legal or insurance requirement unsatisfied with no adequate interim measure in place, or cause operational or commercial disruption you are not prepared to accept. If none of those apply, it is reactive work and belongs in a queue.
Note the phrase about interim measures. A legal requirement does not automatically make immediate repair the only lawful response. Isolating the equipment, taking it out of service, or putting another control in place may satisfy the duty while the repair is scheduled properly. That is a judgement to make deliberately and record, not a loophole to lean on.
Applying the rule consistently matters more than getting every marginal case right. The point is to stop urgency being decided by whoever complains loudest.
Reactive vs preventive maintenance: comparing the true cost
Comparing the three approaches on the invoice alone will always flatter reactive maintenance, because doing nothing until something breaks has an obvious short-term saving and a set of costs that arrive later and land in other budgets.
Factor | Emergency | Reactive | Preventive |
Trigger | Fault, defect or failure requiring an immediate response | Fault or defect that can safely wait | Time, usage or condition threshold reached |
Response time | Immediate, often outside normal hours | Hours to days, queued by priority | Booked in advance to a date of your choosing |
Labour cost | Highest, with callout premiums and overtime | Standard rates, but travel and diagnosis are unplanned | Standard rates, and jobs can be batched by site or asset |
Parts | May have to come from whoever can supply fastest, not best value | Ordered on demand, with waiting time between visits | Known in advance and held or ordered to a schedule |
Downtime | Unplanned and usually at the worst moment | Unplanned, though it can sometimes be absorbed | Planned, and can be timed for quiet periods |
Risk conditions | Fault and working conditions may be uncontrolled at the point of attendance | Risk depends on the fault, and can rise the longer it is left | Work can be prepared, permitted and risk assessed in advance |
Documentation | Often completed retrospectively and under pressure | Varies, and defects are frequently recorded incompletely | Can be built into the planned workflow |
Why reactive work costs more than the invoice suggests
The direct repair cost is usually the smallest part of an unplanned failure. The costs that do the damage are the ones that are not itemised anywhere.
Lost output or lost trade while the asset is down, which for a small business can exceed the repair cost several times over in a single day.
Displaced work, because the team abandoned planned jobs to attend, pushing those jobs into next month and increasing the chance that they too become failures.
Secondary damage, where a component failure takes other parts with it, so a bearing replacement becomes a shaft and housing replacement.
Shortened asset life, since equipment run to the point of failure repeatedly tends to be replaced sooner than equipment maintained on schedule.
Premium sourcing, covering out-of-hours labour, expedited delivery and whatever part is available rather than the specified one.
Management time absorbed by chasing contractors, approving unbudgeted spend and explaining the disruption to others.
There is a widely repeated claim that reactive maintenance costs several times more than planned maintenance. The ratio varies so much by sector, asset type and how the costs are counted that it is best treated as a general direction rather than a figure to plan against. The reliable point is narrower and harder to argue with: the direct repair bill for an unplanned failure can be large, but the largest costs are frequently the indirect ones, and organisations that only track invoices will systematically underestimate them.
What preventive maintenance actually costs
Preventive maintenance is not free and it is not always cheaper. A schedule that services everything at the manufacturer's minimum interval regardless of how the asset is used will consume labour on work that was not needed, and every intervention carries a small risk of introducing a fault that was not there before.
The realistic cost of a preventive programme includes the labour to carry out the scheduled work, the parts consumed on schedule rather than on failure, the planned downtime while work is done, the time spent building and maintaining the schedule itself, and the administrative effort of recording completion.
That last item is where preventive programmes most often fail in smaller organisations. The schedule exists, the work gets done, but the record is a signature on a paper sheet in a folder in the plant room. When the insurer or the enforcing authority asks for evidence, the work might as well not have happened.
When each approach is the right choice
A maintenance strategy is not a choice between the three. It is a decision about which assets sit in which category, reviewed as circumstances change.
When emergency response is the only option
Some failures will always be emergencies no matter how good the preventive programme is. Fire safety equipment that fails, a security failure that leaves a building open, flooding, a gas escape, or an asset carrying life safety functions all demand immediate response.
The aim is not to eliminate emergency work. It is to make sure the emergencies that happen are genuinely unforeseeable, rather than the predictable end point of a service that was skipped three times.
When reactive maintenance is a rational decision
Running an asset to failure is a legitimate strategy, not an admission of defeat, provided the decision is made deliberately. It suits assets where the failure has no safety consequence, the consequence of downtime is minor, a replacement is cheap and quickly available, and there is no statutory inspection requirement.
Consumable fittings, non-critical light fittings, low-value hand tools, and duplicated equipment where a spare is immediately available all reasonably sit here. Servicing a £12 item on a quarterly schedule costs more in labour than replacing it when it fails. The goal is not maximum preventive maintenance. It is economically justified maintenance, sized to the consequence of failure.
The distinction that matters is between run-to-failure as a decision and run-to-failure as a default. The first is recorded, reviewed and defensible. The second is what happens when nobody has decided anything.
When preventive maintenance earns its place
Preventive maintenance is worth the investment where failure carries a consequence you are not willing to accept. In practice that means assets with a safety function, assets subject to statutory inspection, assets whose failure stops trading, assets with long lead times on parts or replacement, and assets where failure causes damage to something else, such as a water system in a building with stock or electronics below it.
It also earns its place where the failure pattern is predictable. Components that wear steadily, filters that clog, belts that stretch and seals that degrade all give reasonable notice on a time or usage basis. Components that fail randomly and without warning give preventive maintenance less to work with, which is where condition monitoring becomes the better answer if the asset justifies the cost.
What UK law requires
Maintenance obligations in UK law are generally written as outcomes rather than as instructions to run a particular schedule. Under regulation 5 of the Provision and Use of Work Equipment Regulations 1998, work equipment must be maintained in an efficient state, in efficient working order and in good repair. Regulation 5 of the Workplace (Health, Safety and Welfare) Regulations 1992 places a comparable duty on the workplace itself and on its equipment, devices and systems, and requires a suitable system of maintenance where appropriate.
Neither regulation tells you to service a specific asset every six months. HSE guidance on the Workplace Regulations notes that the frequency and nature of maintenance depend on the equipment concerned. What the law does is make the condition of relevant equipment and systems your responsibility. In practice, documented inspection, maintenance and remedial work is important evidence that those duties are being managed.
Certain equipment carries more specific inspection, examination or maintenance requirements on top of the general duty.
Lifting equipment is subject to thorough examination at set intervals under the Lifting Operations and Lifting Equipment Regulations 1998. Thorough examination is a separate exercise from routine servicing and does not replace it.
Most qualifying pressure systems require a written scheme of examination under the Pressure Systems Safety Regulations 2000. Regulation 12 of those regulations separately requires the user of an installed system, or the owner of a mobile system, to ensure it is properly maintained in good repair so as to prevent danger. The written scheme therefore sits alongside routine maintenance rather than substituting for it.
Gas appliances and flues carry specific safety check and maintenance duties under the Gas Safety (Installation and Use) Regulations 1998, with the detail depending on the type of premises, whether the property is let, and who has control of the appliance. HSE guidance for landlords makes the point directly: an annual service should not be assumed to satisfy the safety check requirement, and a safety check on its own should not be assumed to amount to effective maintenance.
Electrical systems and equipment are covered by their own duties and guidance, including requirements around inspection and testing appropriate to the installation.
These regimes generate planned work whether or not you have a preventive programme, and the dates are not negotiable. Putting them into the schedule first is the least controversial decision in maintenance planning.
Separately from the law, insurers commonly ask for evidence of maintenance when a claim is made, and may impose conditions in a policy that require servicing to be carried out and recorded. That is a contractual expectation rather than a statutory one, and the two are worth keeping distinct in your own records. Industry best practice, such as manufacturer-recommended service intervals, sits in a third category again: strong evidence of reasonable care, but not in itself a legal requirement.
How to shift the balance towards planned work
Teams that spend most of their time firefighting rarely got there through one decision. The route out is equally incremental, and attempting to build a complete preventive programme in one go is the most common way to fail.
Start with an asset register
You cannot schedule maintenance on assets you have not listed. A workable register needs the asset name, its location, a unique reference, its criticality to the operation, and any statutory inspection requirement attached to it. Serial numbers, install dates and warranty details are useful additions but not blockers.
This step is tedious and is the one most often skipped. It is also the one that makes everything after it possible.
Classify every job as it arrives
Apply the emergency, reactive or preventive label at the point the job is raised, not retrospectively. Retrospective classification is unreliable, because by the time anyone looks back the urgency has been forgotten and every job looks routine.
Within a month you will have something more useful than any industry benchmark, which is a picture of your own split and of which assets generate the unplanned work.
Apply preventive schedules to the critical few first
Rank assets by consequence of failure, then build schedules for the top of that list only. For most small operations that is somewhere between five and fifteen assets. Anything statutory goes on the list automatically, whatever its consequence ranking, because the schedule is not optional.
Manufacturer-recommended intervals are a reasonable starting point. Adjust them once you have your own history, because an asset running two hours a day does not need the same interval as an identical asset running twelve.
Protect planned time in the schedule
The reason preventive schedules collapse is almost never that anyone disagreed with them. It is that planned work is the only thing in the week that can be moved, so it moves every time something unplanned arrives.
Booking planned work into specific slots and defending those slots against anything short of a genuine emergency is the single most effective habit in maintenance planning. If planned work is repeatedly displaced, the schedule is telling you the workload is beyond current capacity, which is a resourcing conversation rather than a scheduling one.
Track the split and review it
Measure the proportion of hours going to planned versus unplanned work and watch the trend. A commonly cited target is that around eighty per cent of maintenance hours should be planned, with the remainder reactive and emergency. This is a rule of thumb that circulates widely in the maintenance industry rather than a figure supported by an authoritative UK source, and the appropriate ratio for a small facilities operation is not the same as for a manufacturing plant.
Treat the direction of travel as the meaningful signal. A team moving from twenty per cent planned to forty per cent over a year has clear evidence that its balance is changing, whatever benchmark it eventually settles on. Whether that shift is an improvement also depends on what happened to the backlog and to the number of failures over the same period, so read the ratio alongside those figures rather than on its own.
Where a CMMS fits
None of the above requires software. A whiteboard and a diary will run a preventive schedule for a handful of assets, and for a very small operation that may be the right answer.
The point at which a system starts to pay for itself is usually when one of three things happens. Recurring schedules become too numerous to hold in someone's head. Evidence of completion is needed for an insurer, an auditor or an enforcing authority, and a folder of paper is no longer adequate. Or more than one person needs to see the same list, and the diary lives with one of them.
A computerised maintenance management system holds the asset register, generates recurring preventive activities automatically, records completion with dates, notes and photographs, and keeps the unplanned jobs in the same place as the planned ones so the split is visible rather than inferred. That last point matters more than it sounds, because a maintenance history that only records the emergencies gives you no way to demonstrate the preventive work you did.
Keeping planned work planned with Trefnus CMMS Trefnus CMMS is a progressive web app built for small UK businesses, facilities teams and trades operations. It is designed around the same sequence this article describes. The asset register gives you the list you cannot schedule without. Recurring activities can be set on a calendar basis or on machine hours, so intervals follow actual usage rather than an arbitrary date. Defects are logged with before and after photographs, which gives a richer and more readily retrievable record of condition and completion than a folder of signed sheets alone. Service contracts and renewals sit in the same system, so the visits you are paying for are tracked alongside the work you do yourself. It runs in the browser and keeps your maintenance records stored locally on the device, so it continues working in a plant room or on a site with no signal, which is exactly where records tend to be captured. Trefnus describes the syncing as peer-to-peer, with an internet connection needed to verify your licence and to connect devices to one another, after which several computers or tablets can share the same records. It is a one-time purchase with no subscription. Find out more at: |
Frequently asked questions
What is the difference between corrective, reactive and preventive maintenance?
Preventive maintenance is carried out before a failure, triggered by time, usage or condition. Corrective maintenance is carried out after a fault or defect has been identified, to restore the asset. Reactive maintenance is corrective work that was not planned in advance. The distinction that catches people out is that corrective work can also be planned: a defect found during an inspection and fixed on a booked visit three weeks later is planned corrective maintenance, which is neither preventive nor reactive. Planned and unplanned describe how the work is managed, while preventive and corrective describe what triggered the work.
What is the difference between reactive and emergency maintenance?
Both are unplanned, but emergency maintenance must be dealt with immediately because delay would create a risk to people, a breach of legal or insurance requirements, or an inability to trade or occupy the building. Reactive maintenance addresses a fault or failure that carries no immediate danger, so the job can be queued and scheduled alongside other work. A burst pipe flooding a stockroom is an emergency. A dripping tap in a staff kitchen is reactive. The practical difference is control over timing, which is what makes emergency work so much more expensive.
Is preventive maintenance always cheaper than reactive maintenance?
No. Preventive maintenance is cheaper overall where failure carries a significant consequence, such as lost trade, safety risk, secondary damage or a statutory breach. It is more expensive where the asset is low value, quickly replaced and carries no safety or compliance requirement, because the labour spent servicing it exceeds the cost of simply replacing it when it fails. The correct approach is to apply preventive schedules selectively, based on the consequence of failure for each asset, rather than universally.
Is reactive maintenance ever the right strategy?
Yes. Running an asset to failure is a legitimate strategy where the failure has no safety consequence, the downtime cost is minor, replacement is cheap and readily available, and no statutory inspection applies. Consumable fittings, non-critical lighting and duplicated low-value equipment all reasonably sit in this category. What matters is that run-to-failure is a recorded decision rather than a default that happens because nobody has assessed the asset.
What percentage of maintenance should be planned rather than reactive?
A figure of around eighty per cent planned work is widely cited in the maintenance industry, but it circulates as a rule of thumb rather than as a benchmark backed by an authoritative UK source, and the right ratio varies considerably between a manufacturing plant and a small facilities operation. A more useful approach is to measure your own current split, then track whether the proportion of planned hours is rising over time. Direction of travel is a more meaningful indicator than any target figure.
Does UK law require preventive maintenance?
UK law generally sets an outcome rather than prescribing a schedule. Regulation 5 of the Provision and Use of Work Equipment Regulations 1998 requires work equipment to be maintained in an efficient state, in efficient working order and in good repair, and the Workplace (Health, Safety and Welfare) Regulations 1992 place a comparable duty on the workplace and its systems. Neither specifies intervals, though a suitable system of maintenance is required where appropriate. Specific equipment such as lifting equipment, pressure systems and gas appliances carries separate examination requirements with defined intervals.
Further reading and official guidance
The following sources cover the statutory maintenance and inspection duties referred to in this article. The sources below cover Great Britain. Businesses in Northern Ireland should check the legislation and guidance applicable there through the Health and Safety Executive for Northern Ireland.
The authoritative guidance on the maintenance duties that apply to work equipment.
A short employer-facing summary of PUWER, useful as a starting point.
Covers the duty to maintain the workplace itself, its equipment, devices and systems.
Explains the thorough examination regime that applies to lifting equipment under LOLER.
Covers which systems are caught by PSSR and the duties that follow, including maintenance.
Explains when a pressure system needs a written scheme of examination and what it must cover.
Sets out gas safety check and maintenance duties, and where the two differ.
The HSE landing page for equipment safety, inspection and maintenance topics.
The maintenance duty in its original wording, direct from the statute book.
Conclusion
Emergency, reactive and preventive maintenance are not three levels of quality. They are three different triggers, and the difference between them is how much control you retain over when the work happens. That control is what determines the cost, the safety exposure and the quality of the record you are left with afterwards.
No organisation eliminates unplanned work, and no organisation should try to schedule everything. The realistic goal is to make sure the emergencies that occur are genuinely unforeseeable, that reactive work is a decision rather than a default, and that the assets whose failure you cannot afford are on a schedule you control.
Start by classifying jobs as they arrive, build an asset register, and put preventive schedules on the small number of assets where failure would genuinely hurt. If you are ready to move that record out of a folder in the plant room, it is worth looking at what a simple maintenance system can hold on your behalf.
Disclaimer
The information in this article is intended for general guidance only and does not constitute professional legal, financial, or regulatory advice. Always consult a qualified professional for advice specific to your circumstances.
