Commercial Kitchen Fire Risk in 2026: Where the Evidence Says to Spend
Fires in food and drink premises have stopped falling while the compliance environment around them has tightened. And 86 per cent of cooking appliance fires trace to human factors, not equipment failure, which should change how operators spend money.
INFIRISK Team·9 min read·
Shrinks99, CC0, via Wikimedia Commons
On 22 July 2026 the Ministry of Housing, Communities and Local Government published fire and rescue incident statistics for the year ending March 2026. Buried in the detailed tables is a number that anyone responsible for a commercial kitchen should sit with for a minute.
Fires in food and drink premises in England have stopped falling.
There were 1,275 primary fires in food and drink premises in the year to March 2026, against 1,285 the year before. That is a fall of eight tenths of one per cent, which in a dataset of this size means no change at all. Against the pre-pandemic year of 2019/20, when there were 1,422, the sector is down 10.3 per cent. Against 2010/11, when there were 1,731, it is down 26 per cent. The long decline has flattened out.
Everything else about the compliance environment around those kitchens has tightened. That gap is the story.
What the numbers say
Food and drink premises is the government's property classification covering restaurants, cafes, public houses and takeaways. It sits inside the wider "other buildings" category, which is everything that is not a dwelling.
In the year to March 2026 there were 13,679 primary fires in other buildings across England. Food and drink premises accounted for 1,275 of them, a 9.3 per cent share, making it the fourth largest building type behind private non-residential buildings (3,205), other public buildings (2,427) and industrial premises (1,806), and ahead of retail premises (1,195). Hotels, boarding houses and hostels are counted separately and recorded 551.
Of the 1,275, some 1,145 were accidental and 130 deliberate. The sector also produced 103 non fatal casualties and one fire related fatality, down from 125 casualties the previous year.
A note on sourcing, because it trips people up. These are MHCLG statistics, not Home Office statistics. Responsibility for fire and rescue transferred on 1 April 2025. Anyone still writing "Home Office fire statistics" in 2026 is citing a department that no longer publishes them.
The finding that should change how you spend money
The cause detail sits in a separate table, currently on the year to March 2025 and next updated in autumn 2026. It is worth the read.
Cooking appliances were the source of ignition in 1,349 fires in other buildings in England in 2024/25. That is 10.3 per cent of all 13,134 other building fires, and the third largest identified source after the unhelpfully broad "other or unspecified" category (5,987) and electrical distribution (2,126).
The breakdown within cooking appliances:
Cooker including oven: 428
Deep fat fryer: 294
Microwave oven: 169
Ring or hot plate as a separate appliance: 143
Other cooking appliance: 121
Grill or toaster: 99
Barbecue: 71
Camping stove: 24
Deep fat fryers alone account for roughly one in five cooking appliance fires in commercial buildings.
Now the part that matters most. Of those 1,349 fires, 1,165 were attributed to human factors. That is 86 per cent. Non human factors accounted for 136, with 48 unspecified.
Read that plainly. Commercial kitchen fire risk in this country is overwhelmingly a supervision, housekeeping and training problem, not an equipment failure problem. An operator who responds to a fire risk assessment by buying better kit is addressing roughly one seventh of the observed cause. The other six sevenths are about who was watching the fryer, whether the extract was cleaned on schedule, and whether anyone told the alarm receiving centre before testing.
Correcting a widely repeated error about TR19 and DW/172
A great deal of trade coverage states that TR/19 replaced DW/172, or treats DW/172 as superseded guidance. That is wrong, and assessors who believe it end up commissioning the wrong scope of work.
The two documents cover different stages of a system's life and both remain live.
DW/172, Specification for Kitchen Ventilation Systems, second edition, published 1 October 2018, covers the design and installation of commercial kitchen ventilation. It is still on sale from BESA with no withdrawal or supersession notice.
TR19 Grease is a standalone specification for fire risk management of grease accumulation within kitchen extraction systems. It covers ongoing cleaning once the system is in use. BESA states that it expands on and replaces what had been Section 7 of the second edition of TR/19. What it replaced was a section of TR/19, not DW/172.
TR/19 itself has also split. General ventilation cleanliness now sits in TR19 Air, released in April 2024, while grease and kitchen extract sit in TR19 Grease. Naming the specific document rather than the bare legacy label "TR/19" avoids a lot of confusion on site.
Neither document is legislation. They are the industry benchmark that insurers, building control and enforcing authorities use to judge whether the responsible person discharged the duty under article 9 of the Fire Safety Order. Failing to follow TR19 Grease is not itself an offence. Failing the article 9 duty is.
The cleaning thresholds, and the insurance exposure
BESA's guidance sets the grease accumulation limit: grease levels must be controlled so as not to exceed a mean average of 200 microns between scheduled cleans, verified by wet film thickness test or deposit thickness test.
Baseline minimum cleaning frequencies are:
Heavy use, 12 to 16 hours a day: every 3 months
Moderate use, 6 to 12 hours a day: every 6 months
Light use, 2 to 6 hours a day: every 12 months
BESA is explicit that these are baselines subject to risk assessment, not fixed entitlements. A busy fryer station in a small unit may need more frequent attention than the headline figure suggests.
The commercial consequence is worth quoting directly. BESA warns that failure to observe a proper cleaning and maintenance regime can invalidate commercial liability and property insurance policies, and that in the event of a fire, insurers will investigate whether highlighted uncleaned or inaccessible areas in previous inspection reports were addressed.
That last clause is the trap. An inspection report that flags an inaccessible section of ductwork, filed and never acted on, is worse than no report. It is documented knowledge of an unmanaged risk.
Section 156 of the Building Safety Act 2022 amended the Fire Safety Order with effect from 1 October 2023. Government guidance states that you must now record the fire risk assessment in full, including all the findings, and record your fire safety arrangements, in all circumstances.
The old exemption for small premises and businesses with fewer than five employees is gone. A single site cafe with three staff is in exactly the same recording position as a hotel group.
Section 156 also requires recording the identity of any person engaged to assist with the assessment, recording and sharing responsible person contact details, and cooperation between multiple responsible persons. That last point is directly relevant to a restaurant or takeaway occupying the ground floor of a residential block, where the operator and the building's freeholder both hold duties.
Competence: what is true today, and what is coming
This is where precision matters, because the position is frequently overstated.
There is a legislative requirement that where the responsible person appoints someone to make or review the fire risk assessment, that person must be competent. Government guidance is clear that this requirement will be brought into force at a later date. It is not in force now.
So today there is no mandatory licence, no statutory qualification and no approved provider list for fire risk assessors. The responsible person remains legally liable for a suitable and sufficient assessment and must satisfy themselves that whoever they appoint is up to it. In practice that is evidenced through third party certification schemes and professional registers, which enforcers recognise as indicators but which are not legal requirements.
That is changing. MHCLG ran a consultation on regulating the fire risk assessor profession from 26 March 2026 to 18 June 2026, covering the future development of the profession, the powers needed to regulate it, and implementation approaches for competency requirements. It applies to England. The consultation has closed and the government response has not yet been published.
Alongside it, BS 8674:2025 sets a tiered competence framework, and the Institution of Fire Engineers is launching Level 3 Award, Level 3 Certificate and Level 4 Certificate qualifications in fire risk assessment for the October 2026 exam session, aligned to that framework.
The practical implication for an operator commissioning an assessment now is simple. Ask what framework your assessor works to and whether they are on a path to the new qualifications. In eighteen months that question may be answered by law rather than by preference.
Getting the suppression standards right
A quick note, because US-origin content circulates widely and cites the wrong documents.
BS 6173:2020 governs installation and maintenance of gas fired catering appliances. It superseded the withdrawn 2009 edition.
BS EN 16282-7:2017+A1:2021 is the current standard for installation and use of fixed fire suppression systems in commercial kitchen ventilation.
LPS 1223 is the LPCB product certification standard for fixed fire extinguishing systems for catering equipment. LPCB listing is typically what insurers look for.
LPS 1263 covers LPCB approval of grease filters in commercial kitchen extract.
UL 300 and NFPA 96 are American. They are not UK requirements and should not be cited as such.
A short checklist
If you operate or assess foodservice premises, the current position reduces to six questions.
Is the fire risk assessment recorded in full, including findings and arrangements, regardless of how small the site is?
Does the cleaning schedule for kitchen extract reflect actual daily hours of use, and is grease depth being measured rather than assumed?
Has every item flagged as inaccessible or uncleaned in a previous report been closed out in writing?
Is fryer supervision an actual documented procedure with named responsibility, given that 86 per cent of cooking appliance fires trace to human factors?
Where the premises sits inside a larger building, do you and the other responsible persons know who holds what?
Does the suppression system carry LPCB listing, and is it maintained to BS EN 16282-7?
None of these require capital spend. All of them are where the evidence says the risk actually lives.
