Fire incident data
Hot work fires in factories and warehouses in England: the 2026 figures
In the year ending March 2026, fire and rescue services in England attended 74 fires in factories, industrial manufacturing and processing buildings and warehouses whose source of ignition was recorded as welding and cutting equipment or a blow lamp. That is 6.4% of the 1,154 fires recorded in those buildings in the year. Over the ten years to March 2026 there were 769, 6.0% of 12,763.
Noon Elite, which publishes this page, makes a sign in kiosk for factories and warehouses and publishes a free hot work permit template.
Source: Ministry of Housing, Communities and Local Government, Fire and rescue incident statistics, England, year ending March 2026, published 22 July 2026. Figures derived 13 October 2026.
Year ending March 2026, and ten years
Four figures
74
fires in factories, industrial manufacturing and processing buildings and warehouses in 2025/26 whose source of ignition was recorded as welding and cutting equipment or a blow lamp. That is 6.4% of the 1,154 fires recorded in those buildings, and about one a week.
769
in the ten years from 2016/17 to 2025/26, 6.0% of the 12,763 fires in the same buildings. 754 were welding and cutting equipment and 15 were blow lamps.
85%
of them were in a building occupied at the time, against 66% of all fires in those buildings over the same ten years. Hot work happens while the site is working.
51
of the 769 involved a fatality or a non-fatal casualty, 6.6% of hot work fires against 3.5% of all fires in the same buildings. The dataset does not say who was hurt.
16 years of fires
Hot work fires by year, 2010/11 to 2025/26
Each row counts the dataset’s rows for the four building types across every fire and rescue service in England, and the rows whose source of ignition is one of the two hot work categories. Years run from April to March.
Fires in factories and warehouses by source of ignition, England, 2010/11 to 2025/26. Source: Other building fires dataset (incident level), year ending March 2026 edition.
| Year | All fires in the four building types | Welding and cutting equipment | Blow lamp | Hot work fires | Share of all fires |
|---|---|---|---|---|---|
| 2010/11 | 2,176 | 131 | 8 | 139 | 6.4% |
| 2011/12 | 2,103 | 140 | 7 | 147 | 7.0% |
| 2012/13 | 1,749 | 118 | 3 | 121 | 6.9% |
| 2013/14 | 1,687 | 101 | 6 | 107 | 6.3% |
| 2014/15 | 1,667 | 108 | 0 | 108 | 6.5% |
| 2015/16 | 1,549 | 95 | 0 | 95 | 6.1% |
| 2016/17 | 1,478 | 111 | 1 | 112 | 7.6% |
| 2017/18 | 1,438 | 80 | 2 | 82 | 5.7% |
| 2018/19 | 1,461 | 97 | 1 | 98 | 6.7% |
| 2019/20 | 1,356 | 87 | 2 | 89 | 6.6% |
| 2020/21 | 1,188 | 61 | 1 | 62 | 5.2% |
| 2021/22 | 1,255 | 65 | 0 | 65 | 5.2% |
| 2022/23 | 1,194 | 66 | 2 | 68 | 5.7% |
| 2023/24 | 1,144 | 50 | 3 | 53 | 4.6% |
| 2024/25 | 1,095 | 65 | 1 | 66 | 6.0% |
| 2025/26 | 1,154 | 72 | 2 | 74 | 6.4% |
The count fell from 139 in 2010/11 to 74 in 2025/26, with the most in 2011/12 (147) and the fewest in 2023/24 (53). All fires in the same buildings fell over the same 16 years, from 2,176 to 1,154, so the hot work share has stayed between 4.6% and 7.6%. Against 2024/25, 2025/26 rose, from 66 to 74.
Over the ten years, welding and cutting equipment was the eighth most recorded source of ignition in these buildings, behind “Unspecified”, “Manufacturing equipment”, “Wiring, cabling, plugs”, “Other industrial equipment”, “Other”, “Kiln, oven, furnace” and “Apparatus - batteries, generators”. Hot work is not the leading cause of fires in factories and warehouses; it is a steady one.
The publisher’s own table FIRE0602 gives “Blowlamps, welding and cutting equipment” as a source of ignition for all other building fires together: 210 fires in 2025/26, of 13,679 in all buildings that are not dwellings. It does not break that figure out by building type, which is what this page does: 74 of the 210 were in the four industrial types counted here.
How they are recorded
The cause of fire, and where it started
The dataset records a main cause for each fire as well as a source of ignition. These are the publisher’s own categories, for the 769 hot work fires of 2016/17 to 2025/26 beside all 12,763 fires in the same buildings.
Cause of fire, hot work fires and all fires, factories and warehouses, England, 2016/17 to 2025/26.
| Cause of fire | Hot work fires | Share | All fires | Share |
|---|---|---|---|---|
| Other accidental | 298 | 39% | 5,359 | 42% |
| Faulty appliances and leads | 76 | 10% | 2,699 | 21% |
| Faulty fuel supplies | 13 | 2% | 1,453 | 11% |
| Deliberate | 0 | 0% | 1,067 | 8% |
| Placing articles too close to heat | 204 | 27% | 1,047 | 8% |
| Misuse of equipment or appliances | 163 | 21% | 687 | 5% |
| Careless handling of fire or hot substances | 15 | 2% | 339 | 3% |
| Chip/ fat pan fires | 0 | 0% | 77 | 1% |
| Playing with fire | 0 | 0% | 19 | 0% |
| Unspecified cause | 0 | 0% | 16 | 0% |
27% of hot work fires are recorded under “Placing articles too close to heat” and 21% under “Misuse of equipment or appliances”, against 8% and 5% of all fires in the same buildings. “Faulty appliances and leads” is the other way round: 10% of hot work fires and 21% of all fires. Where the fire service could name a cause, a hot work fire was more often recorded as how the job was done than as the equipment failing.
| Location | Hot work | All fires |
|---|---|---|
| Process/ Production room | 447 | 5,630 |
| Other | 109 | 2,420 |
| External fittings and structures | 61 | 1,060 |
| Roof/ Roof Space | 47 | 354 |
| Shop floor/ Showroom/ Display hall | 42 | 520 |
| Garage | 23 | 181 |
| Item first ignited | Hot work | All fires |
|---|---|---|
| Structural/ Fixtures/ Fittings | 206 | 3,542 |
| Other/ Not known/ Animal/ Decoration/ Celebration/ Vegetation/ None | 161 | 3,406 |
| Explosives/ Gas/ Chemicals | 100 | 770 |
| Foam/ Rubber/ Plastic | 93 | 1,267 |
| Rubbish/ Waste/ Recycling | 61 | 977 |
| Wood | 61 | 825 |
58% of hot work fires started in a process or production room, against 44% of all fires. Roofs and roof spaces account for 6% of hot work fires and 3% of all fires. The item first ignited was recorded as explosives, gas or chemicals in 13% of hot work fires, against 6% of all fires.
When
Time of day and day of week, 2016/17 to 2025/26
Time of day and day of week over the ten years, in the dataset’s own bands. The publisher defines morning as 6am to midday, afternoon as midday to 6pm, evening as 6pm to midnight and night as midnight to 6am, allocated by the time of the call.
| Time of day | Hot work | Share | All fires | Share |
|---|---|---|---|---|
| Morning | 291 | 38% | 3,809 | 30% |
| Afternoon | 340 | 44% | 4,105 | 32% |
| Evening | 100 | 13% | 2,911 | 23% |
| Night | 38 | 5% | 1,938 | 15% |
| Building type | Hot work | All fires | Hot work share |
|---|---|---|---|
| Industrial Manufacturing (not including Factory) | 416 | 5,291 | 7.9% |
| Factory | 194 | 3,123 | 6.2% |
| Warehouses and bulk storage | 82 | 2,882 | 2.8% |
| Industrial Processing (not including Recycling Centre) | 77 | 1,467 | 5.2% |
| Day | Hot work | Share | All fires | Share |
|---|---|---|---|---|
| Monday | 120 | 16% | 1,994 | 16% |
| Tuesday | 149 | 19% | 2,139 | 17% |
| Wednesday | 140 | 18% | 2,200 | 17% |
| Thursday | 150 | 20% | 2,100 | 16% |
| Friday | 127 | 17% | 1,930 | 15% |
| Saturday | 53 | 7% | 1,335 | 10% |
| Sunday | 30 | 4% | 1,065 | 8% |
82% of hot work fires were in the morning or the afternoon, against 62% of all fires in the same buildings. 11% were at the weekend, against 19%. These are fires attended while people were working, which is also when they are seen.
Discovery and outcome
How soon they were found, and how far they had got
How long after ignition the fire was discovered, how big it was when the fire service arrived, and whether an alarm was present, over the ten years.
| Ignition to discovery | Hot work | Share | All fires | Share |
|---|---|---|---|---|
| Immediately | 369 | 48% | 3,545 | 28% |
| Under 5 minutes | 215 | 28% | 3,947 | 31% |
| 5 to 30 minutes | 107 | 14% | 2,718 | 21% |
| Over 30 minutes and up to 2 hours | 23 | 3% | 565 | 4% |
| Over 2 hours | 20 | 3% | 386 | 3% |
| Not known | 35 | 5% | 1,602 | 13% |
| Occupied | Hot work | Share | All fires | Share |
|---|---|---|---|---|
| Occupied | 655 | 85% | 8,426 | 66% |
| Not occupied | 108 | 14% | 4,062 | 32% |
| Not known | 6 | 1% | 275 | 2% |
| On arrival | Hot work | Share | All fires | Share |
|---|---|---|---|---|
| No fire damage | 147 | 19% | 3,112 | 24% |
| Limited to item 1st ignited | 305 | 40% | 4,406 | 35% |
| Limited to room of origin | 157 | 20% | 2,425 | 19% |
| Limited to floor of origin (not whole building) | 88 | 11% | 1,542 | 12% |
| Limited to 2 floors | 11 | 1% | 202 | 2% |
| Roofs/ Roof spaces | 28 | 4% | 366 | 3% |
| Whole Building/ Affecting more than 2 floors | 33 | 4% | 710 | 6% |
| Alarm | Hot work | Share | All fires | Share |
|---|---|---|---|---|
| Alarm Present and raised the alarm | 301 | 39% | 5,818 | 46% |
| Alarm Present but did not raise alarm | 65 | 8% | 962 | 8% |
| Alarm Present but did not operate | 98 | 13% | 1,317 | 10% |
| Alarm Absent | 305 | 40% | 4,666 | 37% |
Of the hot work fires with a discovery time recorded, 6% were found more than 30 minutes after ignition, against 9% of all fires in the same buildings. 59% of hot work fires had done no damage or were limited to the item first ignited when the fire service arrived, and 33 (4%) had reached the whole building or more than two floors. No alarm was present in 40% of hot work fires.
By fire and rescue service
Hot work fires in the 43 services, 2016/17 to 2025/26
Ten-year totals for each of the 43 fire and rescue services in England, as the dataset names them. The counts are small: West Midlands 73, West Yorkshire 56, Greater London 36, Greater Manchester 34, and 12 services with fewer than ten. One year on its own would be smaller still, which is why no single-year figure by service is given.
The service is where the fire was. It is the territory the fire took place in, not necessarily the service that attended, and the publisher asks for care in commenting on particular services. The table gives counts and a share; it does not rank anybody.
Each row has its own address, so you can link to one service: add its name to the end of this page’s address, as in /factory-hot-work-fires/#derbyshire.
Hot work fires and all fires in factories and warehouses by fire and rescue service, England, 2016/17 to 2025/26. Source: Other building fires dataset (incident level), year ending March 2026 edition.
Ten-year totals, 2016/17 to 2025/26. Download this table as a CSV file.
Caveats
How to read these figures
- The source of ignition is what could reasonably be supposed. The publisher’s guidance says fire and rescue services are not necessarily certain what the source of ignition was, only that it was one that could be reasonably supposed given the evidence available, and that it may change as a fire investigation produces more evidence. The same words apply to the cause of fire. Nothing on this page says a fire was caused by anything.
- Two categories, and only two. A hot work fire here is a fire whose source of ignition is recorded as “Welding/Cutting equipment” or “Blow lamp/Paint remover”, the dataset’s own values. Grinding, torch-applied roofing, heat guns and burning off have no category of their own and are not counted, so the true number of fires started by hot work is higher than any figure here. The dataset records no permit and no job: it cannot say whether a permit was in place.
- The 2025/26 year straddles two recording systems. The publisher says fire data were collected by the Incident Recording System until November 2025 and by the Fire and Rescue Data Platform since, and that this edition holds the records that had reached the new platform by 1 May 2026. Treat the latest year with that in mind.
- Four building types, as the dataset names them. “Factory”, “Industrial Manufacturing (not including Factory)”, “Industrial Processing (not including Recycling Centre)”, “Warehouses and bulk storage”. The publisher says some categories have been combined for personal data reasons. Vehicle repair, recycling centres and public utilities are industrial but are left out, and the first of those matters: vehicle repair buildings recorded 266 hot work fires over the same ten years.
- The fire and rescue service is where the fire was, not who attended. The publisher’s guidance says the service named is the territory the fire took place in, that another service occasionally attends over the border, and that users should be careful about general comments on particular services. The table by service compares counts over ten years and ranks nobody.
- Fires the fire service attended. The dataset covers primary fires attended by a fire and rescue service in England. A hot work fire put out on the spot and never reported is not in it.
- These are official statistics. They are published by the Ministry of Housing, Communities and Local Government, which took responsibility for them from the Home Office on 1 April 2025.
The permit
Where a hot work permit comes in
None of these figures says whether a permit was in place, and the dataset has no field that could. What the guidance asks for is the same either way. The Fire Protection Association’s RC7 recommends a separate permit for every hot work job in an established building, a continuous fire watch while the work goes on and for at least 60 minutes after it stops, then checks until two hours have passed. HSE’s guidance on permit to work systems says hot work is one of the tasks that may need a permit form of its own. Both are quoted, with the figures, on the free hot work permit template.
Noon Elite makes a sign in kiosk for factories and warehouses. A hot work permit is one permit type in its Permit to Work module, which is in early access and has not yet run at a customer site; the page on hot work permit software says what it does and what it does not. The kiosk does not make a site compliant with anything, and nothing on this page is changed by whether a site uses it.
Sources and method
Where each figure comes from
- Other building fires dataset (incident level), year ending March 2026 editionMinistry of Housing, Communities and Local Government, Fire and rescue incident statistics, England, year ending March 2026. Published 22 July 2026. Every count on this page is a count of its rows.
- Other building fires dataset guidanceThe publisher’s description of each field, including the words quoted in the caveats on the source of ignition, the cause of fire, the building type and the service.
- FIRE0602: Primary fires, fatalities and non-fatal casualties by source of ignitionMinistry of Housing, Communities and Local Government. Published 19 August 2026. The cross-check: its row for “Blowlamps, welding and cutting equipment” in all other building fires.
Derived on 13 October 2026. Every figure on this page is a count of the publisher’s own rows, with no reclassification. The incident-level datasets are listed on GOV.UK under fire statistics incident level datasets, and the published tables under fire statistics data tables. The audit record for the same buildings is on the page on factory and warehouse fire safety audits.
Contains public sector information licensed under the Open Government Licence v3.0.
Use these figures
Quote them freely
These figures may be quoted freely. Please link to this page as the source of the analysis, and name the Ministry of Housing, Communities and Local Government as the source of the data.
The table by fire and rescue service is available as a CSV file, with the source and the licence on every row.
If a figure here is wrong, say so and it will be corrected: [email protected].
Next step
A question about these figures?
Email is the quickest way to raise a correction or to ask how a figure was derived. If you run a factory or a warehouse and would rather talk, the call is fifteen minutes.
Book a 15 minute callWould rather email? [email protected]