Certified by the following organisations
Lines Open (Monday - Friday)
9:00AM - 5:30PM
9:00AM - 5:30PM
Every commercial Energy Performance Certificate (EPC) in England and Wales comes from the same government methodology: the Simplified Building Energy Model (SBEM), driven by the National Calculation Methodology. If a rating has come back worse than you hoped, or you cannot tell whether it is good or bad, this page explains how the number is arrived at and lets you test it on an interactive rating explorer further down.
Start here, because it catches almost everyone. A commercial EPC rating is a number where a low number is good. Band A runs from 0 to 25. Band G starts at 151. A building rated 30 is doing well; a building rated 130 is not.
A domestic certificate runs the other way, from 1 to 100 with 100 the best. Same A to G letters, same coloured bars, reversed arithmetic. If you own both a house and a shop you have two certificates whose numbers mean opposite things.
The reason is how the commercial figure is built. It is an index:
So the benchmark itself always lands on exactly 50, the boundary between bands B and C. Emit half the benchmark and you rate 25, which is the worst rating still inside band A. Emit twice the benchmark and you rate 100, the worst rating still inside band D. There is also a band A+ below zero, for a building that exports more energy than it imports over the year.
The second difference matters more in practice. A domestic SAP rating is a cost index: broadly what the dwelling costs to run under standard assumptions. A commercial rating is a carbon rating. Nothing in it responds to the price of fuel.
That single difference makes the two systems reward opposite decisions. Electricity is expensive per unit and comparatively clean; mains gas is cheap per unit and comparatively dirty. So electrifying the heating lifts a commercial rating and pulls a domestic one down, on the same building fabric, for the same money.
Anyone who works across both should expect that, and anyone applying domestic intuition to a commercial building will reach the wrong answer. Our SAP EPC ratings explained page covers the domestic side of the same question.
A commercial assessment does not ask how you run the building. It looks up the activities inside it in the National Calculation Methodology's own database, and that database fixes how many people occupy each space, for how many hours, how brightly it must be lit, how much hot water is drawn and what temperature it is held at. An office and a restaurant in the identical shell are given entirely different assumptions, and neither is negotiable.
The effect on emissions is large. The effect on the rating is small, and the reason is the benchmark. Each use is measured against a benchmark for that use, so a restaurant is compared with other restaurants and not with offices. The benchmark for the sample building's five uses, in kilograms of CO2 per square metre per year:
| Use | Benchmark, kgCO2/m2 a year |
|---|---|
| Office | 13.67 |
| Shop | 18.72 |
| Warehouse | 14.39 |
| Restaurant | 59.51 |
| Industrial unit | 14.63 |
A restaurant is allowed several times an office's emissions before it is judged as harshly, because kitchens and their extract are unavoidable. This is the most useful thing to understand about a commercial rating: it tells you how the building compares with its peers, not how much carbon it emits. Both numbers appear in the explorer, and they move independently.
Every figure in the explorer is a real SBEM calculation, not an approximation. The settings below combine 14,580 ways and all 14,580 were run through the government engine in advance. Even so, they are a thin slice of a real assessment, which takes hundreds of inputs rather than eight.
Start with the building's age. An assessment does not begin by measuring walls, it begins by establishing when the building was built, because the methodology holds a table of what a building of each vintage is assumed to be built to and that assumption stands until an assessor evidences something better. The age setting here carries the engine's own figures for the walls, the roof, the glazing and the air leakage. It deliberately leaves the heating alone: buildings get re-boilered, so the age of the building says nothing about the age of its plant.
The sample is a synthetic two-storey commercial block, 40 by 20 metres, 1,600 square metres over 12 zones. It is not a real building and no client's building is published here; its shape and zoning were modelled on the range found across 27 real assessments. The same shell carries all five uses, which is what isolates the effect of use, though a real warehouse would not be two storeys at 3.5 metres.
Above the certificate sits a Building Regulations check, answering a different question: not what the rating is, but whether the same specification would comply if this were new work. The section below the tool explains why those two answers diverge.
The same office throughout, with only the specification changing. Every figure is calculated by the government SBEM engine for the explorer's sample building, so the table and the tool agree row for row. The first two rows are the identical building: the only difference between them is whether anything on site evidences the age and efficiency of the boiler.
| Specification | Rating | Emissions1 | Criterion 12 |
|---|---|---|---|
| Uninsulated, single glazed, gas boiler, T8 fluorescent, untested, no solar | 90 D | 24.57 | fails |
| Same again, with nothing to evidence the boiler's age or efficiency | 115 E | 31.48 | fails |
| Part L backstop fabric, gas boiler, unevidenced LED, untested, no solar | 35 B | 9.55 | fails |
| Same again, with the lighting specified and evidenced | 32 B | 8.75 | fails |
| Same again, on direct electric heaters | 27 B | 7.28 | fails |
| New-build fabric, air source heat pump, evidenced LED, tested at 3, no solar | 15 A | 4.02 | fails |
| Same again, with a 32 kWp roof array | 7 A | 1.90 | passes |
Every figure is calculated by the government SBEM engine (v6.1.e.1) for the explorer's synthetic sample office, so this table and the tool agree row for row. Illustrative, and not a guarantee for any specific building or report.
1 Building Emission Rate, kgCO2 per square metre per year.
2 Whether the specification would meet both Approved Document L 2021 criterion 1 targets, carbon and primary energy, if this were new work.
On a house, heating dominates and everything else is a footnote. On a commercial building lighting is the same order of size, because the building is occupied in daylight hours, lit to a level the activity database sets rather than to taste, and has a lot of floor area for its heating demand. Which of the two is larger depends on the specification, and in the explorer both orderings occur.
What makes lighting worth singling out is that the calculation penalises missing evidence. If the model cannot state a measured luminaire efficacy it has to fall back on a conservative default. Holding everything else at the tool's opening specification:
| Lighting | Lighting energy1 | Heating energy1 |
|---|---|---|
| T8 fluorescent tubes, efficacy not evidenced | 47.1 | 20.2 |
| LED, efficacy not evidenced | 25.9 | 24.7 |
| LED, specified at 95 lumens per circuit watt, with daylight and occupancy controls | 16.5 | 27.0 |
| The notional building this design is judged against | 7.6 | - |
1 kWh per square metre per year, for the sample office with its fabric, heating, airtightness and solar held constant.
The middle two rows are the same light fittings. The only difference is whether the specification was documented, and it is worth more than a fifth of the lighting energy. Producing a lighting schedule is among the cheapest things on this page and it is routinely left undone.
The heating column moves the other way, and that is not an error. Inefficient lights pour waste heat into the building, and the calculation credits it: cut the lighting load and the heating system has more to do. The building still comes out ahead, because it is trading expensive electric lighting for cheaper gas heat, but the gain is smaller than the lighting saving alone suggests.
A commercial building faces two separate assessments and they are constantly confused. The rating is the A to G number on the certificate, and it compares the building with others of the same use. Building Regulations compliance is a separate calculation reported on a BRUKL Output Document and submitted to building control when a building is put up, extended or fitted out. A BRUKL is never lodged on the EPC register, so searching the register tells you nothing about whether one exists.
Compliance under Approved Document L 2021 turns on numbered criteria. The explorer reports the first two:
Both targets in criterion 1 come from a fictitious notional building: the same size and shape as the design, built to a specification written into the regulations. Whatever the twin achieves becomes the figure the real building has to beat. Because SBEM writes that twin out as a file, its specification can simply be read off:
| Element | What the notional building is built to |
|---|---|
| Walls | U 0.18 |
| Flat roof | U 0.15 |
| Glazing | U 1.4, solar transmittance 0.29 |
| Air permeability | 3 |
| Lighting | 95 lumens per circuit watt, with daylight and occupancy control |
| Rooftop solar | 32 kWp on the gas-heated version of this building |
The last row is the one that surprises people. The notional building has solar panels on it. A new commercial building with a bare roof is not starting level with its target, it is starting behind it and has to find the difference somewhere else. On the sample building, fitting a matching array is what turns a criterion 1 failure into a pass.
How much solar the notional building carries is not fixed, which is worth knowing before budgeting for it. On the gas-heated version of this building it is 32 kWp. Specify a heat pump instead and the notional array falls to a little over 1 kWp, because the heat pump itself already counts towards the renewable contribution the notional building is built to achieve. So the size of the solar hurdle depends on what else you have specified.
The scale of that is worth stating. Of the 14,580 specifications in the explorer, 231 meet the carbon target and 242 meet the primary energy target, but only 225 meet both. Separately, 360 clear every criterion 2 element backstop. Passing one test is not passing the other.
One caveat. Criterion 2's limiting values apply to new work, so an existing building being certified for sale or letting is not judged against them, and the explorer shows them for comparison rather than as an obligation every building carries. Criterion 3 is a solar gain check, which is often mistaken for an overheating assessment: overheating in dwellings is Approved Document O and a different test entirely. Compliance in the end rests with building control, not with a calculation.
Our accredited non-domestic assessors produce commercial EPCs and SBEM calculations for offices, shops, warehouses, restaurants and industrial units across the UK.
Get a quoteIn rough order of how much they tend to shift the number on an existing building:
A commercial rating is not only a document at the point of sale. Under the Minimum Energy Efficiency Standards it has been unlawful since April 2018 to grant a new letting of commercial property in England and Wales rated F or G, and since April 2023 that applies to continuing leases as well. The penalty runs to £150,000. A building sitting in band F or G is a building you may not be able to let.
Both bands are reachable in the explorer above, and the route to them is instructive. Set the sample building to a warehouse and put it in the oldest age band, leaving everything else as it opens. It rates 103 on an evidenced gas boiler: band E, and lettable. Change nothing about the building except removing the evidence for the boiler, so the calculation falls back to its oldest-boiler default, and the same warehouse rates 138, which is band F. Put that unevidenced boiler on oil and it rates 203, deep into band G. Nothing physical has changed between the first of those and the second. The difference is what could be proved.
What comes next is proposed rather than settled. The government's interim response of June 2026 proposes that privately rented non-domestic buildings over 1,000 square metres reach band B by 2031 where that is cost effective. Buildings under 1,000 square metres would stay at band E, and an earlier proposal for a band C milestone in 2027 has been dropped. None of that is law yet: it needs secondary legislation, and a full response and the regulations themselves are still to come.
One further trap. Certificates are valid for ten years, and lodging a new one supersedes any earlier certificate for the same building. Because the 2021 carbon factors changed what a given building scores, a building that was band B under an older methodology can honestly rate several bands worse today without anything about it having got worse. If a valid certificate already exists, it is worth checking what it says before commissioning a new one.
Lower is better, which is the opposite of a domestic certificate. A commercial rating divides the building's carbon emissions by a benchmark for that type of building and multiplies by 50, so the benchmark always lands on exactly 50 and band A starts at zero. A domestic SAP rating runs from 1 to 100 with 100 best. The letters are the same and the arithmetic is reversed, which is the single most common misunderstanding of the two systems. There is also a band A+ below zero, for a building that exports more energy than it imports.
Not directly. SBEM calculates the rating from the data entered, and that data comes from the building. What an assessor can do is tell you which inputs are worth changing, and make sure nothing is being assumed against you for want of evidence. On a commercial building the usual candidates are documenting the lighting specification instead of letting the software fall back on a default, an air pressure test in place of an assumed leakage rate, and establishing real plant efficiencies rather than accepting the defaults.
No, it helps it, and this is where commercial and domestic certificates part company. A commercial rating counts carbon, and a unit of grid electricity now causes less CO2 than a unit of mains gas. A domestic SAP rating counts cost, and electricity costs about four and a half times as much per unit as gas. So switching to a heat pump lifts a commercial rating and pulls a domestic one down. Even direct electric heating rates better than a gas boiler on a commercial certificate: the methodology gives a fanned electric room heater a system efficiency close to a good gas boiler's, so the same heat takes about the same energy and the cleaner fuel wins. The catch is primary energy, the second Building Regulations target, where direct electric is worse than gas. It can improve your certificate and make compliance harder at the same time.
They are two different tests with different pass marks. The rating compares your building with others of the same use. Building Regulations compare it with a notional twin of itself built to a specification written into the regulations, and that notional twin is demanding. It carries rooftop solar panels, walls at 0.18, an airtightness of 3 and a fully specified lighting installation. A building can hold a respectable rating and still miss both of the criterion 1 targets.
Because it has to assume something, and an assumption that flattered the building would let anyone claim anything. Where the assessor cannot evidence a value, the National Calculation Methodology supplies its own, and those defaults are set deliberately low. A boiler whose age and efficiency nobody can establish is treated as the oldest band the methodology holds, a seasonal efficiency of 65% against the 91% a documented modern boiler earns. A lighting installation with no specification is given a conservative efficacy whatever is actually fitted. A building with no air pressure test is assumed to leak at the rate expected of pre-1995 construction. Every one of those is displaced by evidence, and the evidence is usually cheaper than the work people consider instead: a nameplate photograph, a commissioning certificate, a lighting schedule, a pressure test. This is the single largest thing an owner controls.
A BRUKL Output Document is the Building Regulations compliance report for a new or altered non-domestic building, produced by the same software that produces the EPC and submitted to building control. It is a different document from the certificate, it is not lodged on the EPC register, and a register search will never show whether one exists. If you are building, extending or fitting out, you need one. If you are selling or letting an existing building unchanged, you need the certificate.
Related reading: SAP EPC ratings explained (the domestic counterpart to this page), EPC ratings explained for existing homes, commercial EPCs, BRUKL and SBEM for new build, and EPC changes explained.