Home / Kitchen Extraction Cleaning / Guide
Kitchen extraction - the standard
How often should a kitchen extract be cleaned? The recognised standard does not leave it to guesswork - it sets frequency by how much a kitchen cooks, with specific intervals for heavy, moderate and light use. Knowing what the standard actually specifies is the starting point for getting your own frequency right.
The short answer
The recognised standard for kitchen extract cleaning frequency, TR19 Grease, bases it on cooking volume measured in hours per day. As a baseline: heavy use (around 12 to 16 hours a day) points to cleaning roughly every 3 months; moderate use (around 6 to 12 hours) roughly every 6 months; and light use (around 2 to 6 hours) roughly every 12 months. These are baseline recommendations - more frequent cleaning may be needed depending on the cooking type, the system's condition, and insurance or fire risk assessment requirements.
Why a standard exists
The question of how often to clean a kitchen extract could be left to opinion, but it is not - there is a recognised standard, TR19 Grease from BESA, that sets out frequency guidance. This matters because extract cleaning is a fire-safety measure, and a fire-safety measure benefits from a recognised, consistent basis rather than each operator or provider guessing. The standard gives an accepted framework for how often the system should be cleaned, which is what insurers, fire risk assessors and operators can reference in common.
The standard's approach is to tie frequency to cooking volume, because that is what determines how fast the fire risk builds. Grease accumulates in proportion to how much a kitchen cooks, so the more cooking, the sooner the system reaches a level of grease that carries fire risk. Rather than one universal interval, the standard therefore bands kitchens by their cooking volume and sets a frequency for each band - a risk-based approach grounded in the actual driver of the risk. Knowing what the standard specifies is the foundation for setting a defensible cleaning frequency.
What the standard specifies
The standard's frequency guidance works through three broad usage bands, defined by cooking hours per day. Heavy use - a kitchen cooking for long hours, around 12 to 16 hours a day - points to cleaning approximately every 3 months. Moderate use - around 6 to 12 hours a day - points to cleaning approximately every 6 months. Light use - around 2 to 6 hours a day - points to cleaning approximately every 12 months. So the interval shortens as the cooking volume rises, from a year for light use down to quarterly for heavy use.
These bands give a clear, specific starting point for any kitchen. Rather than a vague sense that busy kitchens need cleaning more often, the standard puts numbers on it: identify roughly how many hours a day the kitchen cooks, and the corresponding band gives an indicative interval. A high-volume operation cooking most of the day falls into heavy use and its quarterly interval; a light user cooking a few hours falls into the annual one. The banding translates cooking volume into a concrete cleaning frequency, which is exactly what a standard is for.
Baseline, not a ceiling
Crucially, these intervals are baseline recommendations, not fixed maximums - the standard is clear that more frequent cleaning may be required depending on specific circumstances. The type of cooking matters: heavy frying, grilling and char-cooking throw far more grease into the extract than lighter cooking, so a kitchen doing a lot of grease-heavy cooking may need cleaning more often than its hours alone suggest. The condition of the system matters too, as does what insurance and fire risk assessments require.
So the banded intervals are the starting point, to be shortened where conditions demand. A kitchen at the heavy-use band that also does intensive frying, or has a system prone to fouling, may warrant more frequent cleaning than the quarterly baseline. The standard's own position is that the intervals are minimum recommendations and that actual frequency should reflect the specific kitchen's operation, cooking type and the grease levels found. Treating the bands as a floor that circumstances can raise - not a ceiling - is how the standard is meant to be applied.
Verified by measurement
Underpinning the frequency guidance is the standard's measurable definition of acceptable cleanliness, which is what ultimately determines whether the frequency is right. TR19 Grease specifies that grease levels should be controlled so as not to exceed a mean of 200 microns between scheduled cleans, with areas at 500 microns or more needing immediate attention. So the real test of a cleaning frequency is whether it keeps the grease below these levels - the interval is right if it does, and too long if the grease exceeds them before the next clean.
This ties the frequency back to measurement rather than the calendar alone. The banded intervals are the recommended starting frequency, but the grease-deposit limits are the actual standard the system must meet - so the frequency should be set, and adjusted, to keep the grease within the limits for the particular kitchen. A kitchen finding its grease approaching the limit before the scheduled clean needs a shorter interval than the band suggested. So the standard combines a use-based frequency recommendation with a measurable cleanliness limit, and the two together define what proper extract cleaning frequency means.
The takeaway
The recognised TR19 Grease standard sets extract cleaning frequency by cooking volume, with baseline intervals of roughly 3 months for heavy use (around 12 to 16 hours a day), 6 months for moderate use (around 6 to 12 hours), and 12 months for light use (around 2 to 6 hours). It is a risk-based framework, because grease - and the fire risk - builds in proportion to how much a kitchen cooks.
But the bands are baseline recommendations, not ceilings: more frequent cleaning may be required depending on the cooking type, the system's condition, and insurance or fire risk assessment requirements, and the real test is keeping the grease below the standard's measurable limits. So knowing what the standard specifies is the starting point, and applying it to your own kitchen - placing it in a band and adjusting for its specifics - is how you turn the standard into the right frequency for your operation.
Questions
TR19 Grease bases it on cooking volume in hours per day, with baseline intervals of roughly 3 months for heavy use (12 to 16 hours a day), 6 months for moderate use (6 to 12 hours), and 12 months for light use (2 to 6 hours).
Because grease - and the fire risk it creates - accumulates in proportion to how much a kitchen cooks. The more cooking, the sooner the system reaches a level of grease that carries fire risk, so a risk-based frequency follows the cooking volume.
No - they are baseline recommendations. The standard is clear that more frequent cleaning may be required depending on the cooking type (heavy frying throws more grease), the system's condition, and what insurance and fire risk assessments require. The bands are a floor, not a ceiling.
By grease measurement. TR19 specifies grease should not exceed a mean of 200 microns between cleans, with 500 microns or more needing immediate attention. The frequency is right if it keeps the grease below these limits, and too long if the grease exceeds them before the next clean.
You may need cleaning more often than your hours alone suggest. Heavy frying, grilling and char-cooking throw far more grease into the extract than lighter cooking, so a kitchen doing intensive grease-heavy cooking may warrant a shorter interval than its usage band's baseline.
Place your kitchen in a band by its cooking hours for the indicative interval, then adjust for its specifics - cooking type, system condition, insurance and fire risk requirements - and verify by keeping the grease within the standard's limits. The band is the start; your kitchen sets the rest.
We clean extract systems to the TR19 Grease standard on a frequency matched to your cooking volume - and verify the grease is within the standard's limits, so the interval is genuinely right.
