Odour assessment benchmarks and offensiveness categories

Benchmark selection is the point in an odour assessment where the whole conclusion is decided. Before any dispersion model runs, the assessor classifies the odour as most offensive, moderately offensive or less offensive, and that choice sets the exposure benchmark the modelled results are compared against. Get it wrong and a scheme that would have passed at 6.0 odour units fails at 1.5, or worse, passes against a benchmark the environmental health officer will not accept.

The indicative benchmarks used in England derive from Environment Agency odour management guidance and are carried into the Institute of Air Quality Management's planning framework. They are expressed as the 98th percentile of hourly mean concentrations over a modelled year, written as ouE/m3 C98, 1-hour.

What the benchmark means

The benchmark is an exposure level at a sensitive receptor below which odour is unlikely to cause unreasonable loss of amenity. It is not a statutory limit and it is not a nuisance threshold under Part III of the Environmental Protection Act 1990. Officers treat it as the yardstick for a planning judgement, which means an exceedance invites refusal or a condition rather than automatic prosecution.

Three figures are in common use:

Those figures do the heavy lifting, so the reasoning that places a particular process in a particular row has to be written down, not assumed.

Classifying the odour by offensiveness

Offensiveness is about hedonic tone, the unpleasantness of the odour at a given concentration, rather than its strength. Two processes can emit identical odour unit concentrations and sit in different categories because one smells of bread and the other of putrefaction.

Mixed sources complicate this. A waste transfer station handling both dry recyclables and food waste does not have one hedonic tone, and the honest approach is to assess against the most offensive component present in meaningful quantity. An anaerobic digestion plant with a digestate store and a combined heat and power unit has at least two source types with different tones.

Food processing sits awkwardly between the rows. Frying fats and roasting meat are usually argued as moderately offensive, while a bakery or a coffee roaster falls into the less offensive category. The practical consequence is a fourfold difference in the concentration that counts as acceptable at the nearest window.

And the category can shift with operating condition rather than with the process description. A well-run poultry house smells different from the same house during a muck-out, so the assessment has to say which condition it modelled.

Where the classification is arguable, assess against the more precautionary benchmark and show the sensitivity run at the less precautionary one. That gives the officer the range rather than a single number they cannot interrogate. It also protects the applicant if the position is challenged on appeal.

Source strength data and its provenance

A benchmark comparison is only as good as the emission rate feeding the model. Odour emission rates should come from olfactometry to BS EN 13725 on the process in question, or on a closely comparable operating plant, with the sampling date, number of samples and panel details recorded.

Literature values are sometimes unavoidable for a plant that does not yet exist. When they are used, state the source publication, the process it described and why it is comparable, then carry the uncertainty into the conclusion. A single literature emission rate applied to a novel process and compared against a 1.5 ouE/m3 benchmark is a weak evidence base.

Two provenance questions decide credibility. Was the sampled plant operating at the capacity the application proposes? And were abatement systems running as they will run in practice, rather than in a commissioning condition that cannot be sustained?

Meteorological data years and sensitivity

The C98 statistic comes from a modelled year of hourly meteorology, so the choice of year matters. Standard practice is to run at least five recent years from a representative station, report the worst-case year and show the spread.

Distance from the station, surface roughness and terrain all shift results. A site in a valley modelled on data from a coastal airport 40 km away produces numbers that look precise and are not. Say which station was used, how far it is and why it is representative, and run a sensitivity test where the choice is open to question.

Building downwash is the other sensitivity worth reporting. Where the emission point sits close to a tall building, including or excluding that building can move predicted concentrations at nearby receptors by a large margin.

Receptor selection and sensitivity

Receptors are the places where people experience the odour: dwellings, schools, hospitals, workplaces and some outdoor amenity areas. Residential gardens are normally assessed at the dwelling, though an officer may ask about a garden where it is the main amenity space.

Receptor sensitivity modifies the judgement rather than the benchmark figure. A dwelling and a short-stay industrial office do not deserve equal weight, and the assessment should say which receptors drive the conclusion. Include receptors with planning permission but not yet built, because a consented housing allocation downwind will be raised by the authority if you leave it out.

An illustrative case makes the point. On a proposed in-vessel composting facility, the nearest existing dwelling sits 320 m to the north-east and a consented 60-home allocation sits 180 m to the south-west. Modelling only the existing dwelling gives a comfortable pass against a 3.0 ouE/m3 benchmark. Including the allocation does not. These figures are illustrative and not from a real assessment.

Reporting against the benchmark honestly

The results section should give the predicted C98 at each receptor, for each modelled year, against the selected benchmark, with the classification reasoning stated. Rounding 3.04 down to 3.0 and declaring compliance is the kind of thing that gets picked apart by a consultee.

Where an exceedance is predicted, set out the mitigation and re-run rather than arguing the benchmark should be relaxed. Enclosure, biofilters, carbon filters, stack height and operational controls all have modelled effects, and an odour management plan secured by condition is the usual way the authority makes them stick.

Give the officer the inputs as well as the outputs. A results table without the emission rates, stack parameters, building dimensions and receptor coordinates behind it cannot be checked, and an unverifiable assessment tends to attract a holding objection from environmental health rather than a conclusion. Appending the model input file, or a clear summary of it, removes that argument in one step.

Finally, be explicit about what the assessment does not cover. Construction-phase dust, traffic emissions and operational noise are separate matters with their own criteria, and bundling a sentence about each into an odour report leaves all three half-answered.

An Odour Assessment is a report councils expect from a suitably qualified author, normally an air quality or odour specialist with olfactometry experience, and professional review is strongly advised before submission. An Odour Assessment drafted on PlanScope structures the benchmark selection, source inventory and receptor sections and flags the modelling inputs that must be supplied, and a Ventilation and Extraction Statement often sits alongside it where cooking or process extraction is involved. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval. For specialist technical reports, qualified professional input may be required.

Odour rarely arrives alone in a validation list. Waste-handling schemes usually need the storage and collection evidence covered in waste storage calculations for a Waste Management Strategy, and on larger proposals the amenity findings feed the wider picture described in health determinants inside a Health Impact Assessment.