Building Energy Intensity (EUI) Position Checker
An operational rating is a comparison, not a measurement — and every error that distorts one moves the rating without moving the energy. The building consumes exactly what it consumed; the score shifts because the comparison shifted. This checker asks eight questions about the inputs behind your rating, tells you which would distort it and in which direction, and computes nothing. It applies CIBSE TM46, which underpins the Display Energy Certificate.
A DEC rating is your consumption measured against a fixed benchmark for your building’s category. Get the category, the floor area, the separable loads or the adjustments wrong and the rating changes while the energy does not — and the benchmark table itself has been frozen since 2008.
Eight questions, about a minute. An operational rating is a comparison, not a measurement — and every error this page checks for moves the rating without moving the energy. The building consumes what it consumed; the score shifts because the comparison shifted. This tells you which of your inputs would distort it, and which way. It computes no rating.
- The category was reached by judgement rather than allocated from the TM46 building-type list, which means the consumption is being compared against a benchmark that may belong to a different kind of building. This is the one error on the page with no sanity check available to you: a wrong floor area produces a figure somebody may notice looks unlike the building, and a wrong category produces a perfectly plausible letter over a perfectly plausible number. Allocate from the list rather than by resemblance — a specialist facility scored as a general office is the standard case, and it is not a close call once the list is in front of you.
- The floor area includes space that does not serve the building’s primary function — a car park is the usual case — and floor area is the denominator of every intensity figure on the certificate. An inflated denominator lowers the measured intensity and flatters the rating, which is the direction that goes unquestioned: nobody queries a result that came out better than expected. Re-measure to the internal face of external walls, excluding areas outside the primary function.
- No degree-day correction has been applied, so a cold year is being read as a badly run building. The correction scales the fossil-thermal benchmark by the ratio of the location’s actual heating degree-days to its long-run average, on a 15.5 °C base, using data organised across fourteen UK regions. Without it the rating moves with the winter rather than with the management, and it moves in whichever direction the weather went — so a mild year flatters exactly as much as a cold one penalises.
- The building runs substantially longer than the standard hours its category assumes, and no occupancy adjustment has been applied. A twenty-four-hour control centre or an extended-hours leisure facility consumes more for reasons of legitimate use rather than inefficiency, and the adjustment raises the benchmark in proportion to the actual operating pattern. Without it the rating penalises the building for being open.
- An atypical load is present with no sub-metering, so the separables provision is unavailable rather than unused. Only around 0.3 per cent of certificates in the national database make any use of it, and missing sub-metering is the stated reason — which makes this a metering decision with a rating consequence attached, and one that pays back on every certificate afterwards rather than once.
- Gaps in the twelve months have been filled by estimate. For non-metered fuels — oil, coal, wood — the expected route is reconstruction from delivery records rather than estimation, and delivery records usually exist even where meters do not.
- This rating is displayed, claimed or used to compare buildings, so any distortion above travels to a reader who cannot see it and has no way to test it. That does not change what to fix; it changes how long it is reasonable to leave it unfixed.
What this rating cannot tell you
- The TM46 benchmark table was fixed in 2008 from stock data of that era and has never been revised, while LED lighting displaced fluorescent and halogen, IT loads shifted and the grid decarbonised. A genuinely efficient modern building can rate merely average because the benchmark did not move with the stock — so a rating is partly a statement about the table, and improving against it is not the same as improving.
- The benchmarks are annual and fixed, which means they cannot distinguish a summer from a winter within the year. Where seasonality is the story, the rating is the wrong instrument for telling it.
- Twenty-nine categories cannot capture every building. Performing-arts venues, specialist research facilities and unusual mixed uses sit awkwardly in the nearest category, and the separables provision only partly mitigates that.
Position register — each input, where you stand, and the rule
| Input | Your position | The rule |
|---|---|---|
| Your position | A public building over 1,000 m², renewed annuallyAnnual renewalmet | A DEC is required for qualifying public buildings: validity is one year above 1,000 m² and ten years for 250–1,000 m² |
| TM46 category | Assumed — nearest category by judgementAssumedopen | The category selects the benchmark; mixed uses are built up by area rather than forced into one of the 29 categories |
| Useful floor area | A gross figure including areas that do not serve the primary functionGrossopen | Useful floor area is measured to the internal face of external walls and excludes areas not serving the primary function |
| Atypical loads | Present with no sub-meteringNo sub-meteringopen | TM46 allows separately sub-metered atypical uses to be excluded from the benchmark comparison; the energy still appears in the certificate’s carbon footprint |
| Weather correction | Not appliedAbsentopen | The 15.5 °C degree-day correction scales the fossil-thermal benchmark to the location and year — and does not apply to electricity |
| Hours adjustment | Extended hours, adjustment not appliedAbsentopen | TM46 benchmarks assume standard operating hours for each category and provide an adjustment where the building runs substantially longer |
| Consumption data | Gaps filled by estimateGaps estimatedopen | The rating requires a complete twelve months per fuel; non-metered fuels are reconstructed from delivery records rather than estimated |
| What the rating is used for | Displayed publicly on a certificateCarried outward | A distorted rating displayed, claimed or compared carries the distortion to a reader who cannot see it |
Why this verdict: The building has been allocated to a TM46 category by judgement rather than from the building-type list, or a mixed building has been forced into a single one — and the category is what selects the benchmark. That makes this the error that invalidates the entire comparison rather than shifting it. The consumption is real and the arithmetic is right; the number it is being measured against belongs to a different kind of building. Our standards page names it first among the errors that distort a certificate, and describes the result exactly: a rating that reflects the mismatch rather than the building. What makes it top this register rather than the floor-area finding below is that it offers the reader no sanity check. An implausible floor area can look wrong to somebody who knows the building. A wrong category produces a perfectly plausible letter over a perfectly plausible number, and there is nothing in the output to catch. It is also the one finding whose direction cannot be predicted. A forced mixed-use allocation errs in whichever way the dominant use pushed it, so the rating cannot even be treated as conservative.
This register states no rating, no band letter and no benchmark figure. Those depend on the 29-category TM46 table, heating degree-day data organised across fourteen UK regions, and the three adjustments — an assessor’s calculation rather than ours. It also does not decide whether a Display Energy Certificate discharges an ESOS obligation: that is a different question, answered by the Energy Audit Obligation Checker. Every finding here changes what the rating says, not what the building uses.
The TM46 benchmark table has been fixed since 2008 and its revision is the single change that would move every rating in the country at once. We will email you when something this register applies changes — not otherwise.
This register applies CIBSE TM46 and the Display Energy Certificate regime as our reviewed standards page records them. It reads nothing of yours and calculates nothing — no operational rating, no band, no kWh/m². It reports which of your inputs would distort a rating and in which direction, not how efficient your building is and not whether an assessor would agree.
What this verdict means
Seven outcomes, five of them findings. They are ordered by how much of the rating each one moves — and, just as importantly, by whether you would ever notice.
Category mismatch
The consumption is being compared against a benchmark that may belong to a different kind of building. The only error here with no sanity check available to you, and the only one whose direction cannot be predicted.
Denominator unsound
Floor area is the denominator of every intensity figure on the certificate, so an error scales all of them at once. An inflated area flatters the rating — the direction nobody queries.
Separables unclaimed
An atypical load is being scored as normal for the category. The rare finding whose correction changes the rating and no consumption at all.
Adjustments unsound
The weather or hours correction is missing or misapplied, so a cold winter or long opening hours is being read as poor management.
Data incomplete
The twelve months behind the rating is not whole. The most tractable finding here, and the one with no justification available — the readings exist.
Position sound
Every input holds. A statement that the comparison is fair, not that the building is efficient.
The seventh is an exit. Not rated means no TM46 comparison is being made, so there is nothing for an error to distort. That is a statement about today: the DEC duty follows public-authority occupation, frequent public visits and floor area, so a change of tenancy or use carries it in without anything about the energy changing.
Every finding on this page changes what the rating says, not what the building uses. That is not a caveat — it is the whole shape of the instrument. If you are looking for a number that describes consumption, energy per square metre without a benchmark behind it is a different and often more useful measure, and it is not an operational rating.
A rating is a comparison, not a measurement
The distinction sounds pedantic and decides everything. TM46 provides a benchmark for each of 29 building categories. Your consumption is set against the benchmark for your category, adjusted for weather and operating hours, and the result is expressed as an operational rating on an A–G scale.
So the rating has two inputs, not one. Your energy is the first. The comparison — which benchmark, over which floor area, covering which loads, adjusted how — is the second, and it is the one that is easier to get wrong and harder to notice.
That is why the errors this page checks for share a property that is genuinely unusual in carbon and energy work: none of them requires you to use a different amount of energy to fix. Correcting the category, re-measuring the floor area, separating a sub-metered server room, applying a degree-day correction — each changes the rating and leaves consumption untouched.
A gross floor area inflates the denominator and lowers measured intensity. A weather correction wrongly applied to electricity inflates the electricity benchmark and improves the comparison. Both push the rating the right way, and a result that comes out better than expected is the one nobody goes back to check.
Why the category outranks everything else
Our standards page lists the errors that distort a certificate, and wrong category allocation is first. The reason it tops this register too is narrower than “it matters most” — it is that the reader has no way to catch it.
Consider the alternative. An implausible floor area is a number somebody who knows the building can look at and say that is not right. Missing consumption months are visible as gaps. An unseparated server room is a thing you can walk past.
A wrong category produces none of those signals. It produces a perfectly ordinary letter over a perfectly ordinary number, arrived at by correct arithmetic on real consumption. The rating is internally consistent and completely valid — for a different kind of building.
Two shapes of the error recur. A specialist facility scored as a general office is the first. The second is a mixed building forced into a single category, where TM46 expects the categories to be built up by area into a weighted benchmark. That second one has a further property worth knowing: its direction depends on which use dominated the choice, so the resulting rating cannot even be treated as conservative.
The provision used by 0.3% of certificates
TM46 allows energy uses that are unusual for a building’s category, or highly variable between buildings, to be removed from the benchmark comparison. A server room in an otherwise ordinary office. A commercial kitchen. Major process equipment. The condition is that the load is separately sub-metered and meets the criteria in the standard.
Analysis of the national DEC database found that only around 0.3 per cent of certificates make any use of separables at all. The reasons given are missing sub-metering and limited awareness of the provision — not that buildings rarely carry atypical loads, which they plainly do.
The consequence is direct. A building with a data centre inside it, unseparated, is scored as though that load were normal for its category. It is not, and the rating is worse for a reason that has nothing to do with how the building is run.
The separated energy still appears in the building’s carbon footprint on the certificate. It is excluded from the benchmark comparison only. Nothing is hidden and no energy is removed — which is precisely why our standards page can describe it as one of the few legitimate ways to improve a DEC without changing actual consumption.
Where the sub-metering already exists, this is available now and costs nothing but the assessor’s attention. Where it does not, this is a metering decision with a rating consequence attached — and one that pays back on every certificate afterwards rather than once.
The two adjustments that make it fair
A fixed benchmark compared against raw consumption would penalise buildings for things they do not control. TM46 corrects for two of them.
Weather. Heating energy depends heavily on outside temperature, so the fossil-thermal benchmark is scaled by the ratio of the location’s actual heating degree-days in the assessment year to its long-run average, on a 15.5 °C base, using data organised across fourteen UK regions. A building assessed in a cold winter is judged against a proportionately higher benchmark.
Two errors follow from it, and they are opposite. Omitting the correction reads a cold year as poor management. Applying it to electricity is an error in its own right — TM46 does not weather-correct electricity, and doing so inflates the electricity benchmark and flatters the result.
Operating hours. The benchmark for each category assumes standard operating hours. A 24-hour control centre or an extended-hours leisure facility consumes more for reasons of legitimate use rather than inefficiency, and the occupancy adjustment raises the benchmark in proportion to the actual pattern. Without it, the building is penalised for being open.
Both errors move the rating in whichever direction the year or the timetable happened to go. A mild winter flatters exactly as much as a cold one penalises, which is why neither can be waved through as conservative.
The benchmark has not moved since 2008
This one is not a finding against anybody, and it is the most important thing on the page to understand before acting on a rating.
The TM46 benchmark table was fixed in 2008 from stock data of that era and has never been revised. Since then LED lighting has displaced fluorescent and halogen, IT loads have shifted, heating controls have improved, and decarbonisation of the electricity grid has changed the carbon weighting of the electricity column relative to gas.
The result is that a genuinely efficient modern building can rate merely average, because the benchmark it is measured against did not move with the stock. CIBSE maintains a separate Energy Benchmarking Tool precisely to provide more current data-driven benchmarks, but the statutory DEC continues to use the 2008 table.
A good rating is worth somewhat less than it looks and a poor one may be worth more. And improving against a frozen benchmark is not the same as improving — a fact worth stating out loud before a rating is used to support a target or a claim.
Two further limits sit alongside it. The benchmarks are annual and fixed, so they cannot distinguish a summer from a winter within the year — academic work on the University Campus category found large discrepancies against actual measured consumption, prompting proposals for monthly thermal benchmarks the static table cannot accommodate. And 29 categories cannot capture every building: performing-arts venues, specialist research facilities and unusual mixed uses sit awkwardly in the nearest one, which the separables provision only partly mitigates.
What this checker does not decide
It computes no rating. No operational rating, no band letter, no kWh/m² benchmark figure. Those depend on the 29-category table, heating degree-day data across fourteen UK regions and the three adjustments — an accredited assessor’s calculation, not ours. Quoting a benchmark value here would invite exactly the category error the page warns about.
It does not decide whether a DEC discharges an ESOS obligation. A Display Energy Certificate is one of the UK scheme’s four compliance routes, and whether it discharges your duty is scope-dependent. That is a different question with a different answer, and our Energy Audit Obligation Checker works through it.
It does not tell you whether your building is efficient. It tells you whether the rating means what it appears to mean. Those are separate findings and a page that ran them together would be less useful, not more.
And it is not an embodied-carbon tool. If the question is the carbon in the building rather than the energy through it, our whole-building LCA calculator and EN 15978 conformance checker cover that side.
How we keep this current
This register applies CIBSE TM46 and the Display Energy Certificate regime as our reviewed standards pages record them, and it moves with those pages rather than ahead of them.
The single change that would move every rating in the country at once is a revision of the 2008 benchmark table. It has been anticipated for years and has not happened. If it does, every input on this page keeps its meaning and every rating derived from it changes — which is the clearest possible demonstration that a rating is a statement about a comparison.
The re-check control in the result panel is wired to that. We email when something this register applies changes — not otherwise.
Frequently asked questions
That is the normal case rather than the strange one, and it is what this page exists for. A rating is your consumption compared against a fixed benchmark, so it moves when the comparison moves. A colder winter with no degree-day correction applied, a re-measured floor area, a change of assessor who allocated a different category, or a new server room left inside the comparison will each move the rating with consumption flat. Work through the inputs before concluding anything about the building — and note that two of the errors push the rating the right way, so a rating that improved unexpectedly deserves the same check.
Yes, and TM46 explicitly provides for it. Where a building carries an atypical load — a server room, a commercial kitchen, major process equipment — that is separately sub-metered and meets the standard’s criteria, it can be excluded from the benchmark comparison so the building is not scored as though that load were normal for its category. The separated energy still appears in the certificate’s carbon footprint; only the comparison changes. It is also strikingly under-used: around 0.3 per cent of certificates in the national database make any use of the provision, with missing sub-metering and limited awareness given as the reasons.
Measured to the internal face of external walls, excluding areas that do not serve the building’s primary function. A car park included in the figure is the standard error. It matters more than it sounds because floor area is the denominator of every intensity figure on the certificate, so getting it wrong scales the whole rating rather than shifting one number. The direction is worth knowing too: an inflated area lowers measured intensity and improves the rating, which is the direction least likely to prompt anyone to go back and check.
No, and applying it to electricity is an error in its own right rather than a harmless extra. TM46 corrects the fossil-thermal benchmark only, scaling it by the ratio of the location’s heating degree-days in the assessment year to its long-run average on a 15.5 °C base, using data organised across fourteen UK regions. Electricity is not weather-corrected. Applying the correction to both inflates the electricity benchmark and flatters the result — so this is one of the two errors on this page that make a building look better than it is.
You do not pick one. TM46 expects mixed uses to be built up by area across categories, producing a weighted benchmark rather than a single one. Forcing a mixed building into one category means part of the estate is judged against a benchmark that was never meant for it, and the resulting rating reflects the mismatch rather than the building. The unusual thing about this error is that its direction depends entirely on which use dominated the choice — so unlike most measurement errors, the result cannot be treated as conservative in either direction.
Not if the occupancy and hours adjustment is applied. The TM46 benchmark for each category assumes standard operating hours, and a building running substantially longer consumes more for reasons of legitimate use rather than inefficiency. The adjustment raises the benchmark in proportion to the actual operating pattern, which makes the comparison fair rather than mechanical. If it has not been applied, the rating is penalising the building for being open — and that is a correction to ask the assessor about rather than an outcome to accept.
The rating requires a complete twelve months per fuel, and the expected route for non-metered fuels — oil, coal, wood — is reconstruction from delivery records rather than estimation. Delivery records usually exist even where meters do not, which is what makes reconstruction the standard answer rather than a concession. Estimating where metered data does exist is the one data problem on this page with no justification available at all, and a rating built on it cannot be defended afterwards by pointing at the meter.
It means your consumption sits where it sits against a benchmark fixed in 2008 and never revised. Since then LED lighting has displaced fluorescent and halogen, IT loads have shifted, controls have improved and the grid has decarbonised — which also changed the carbon weighting of electricity against gas. A genuinely efficient modern building can rate merely average because the benchmark did not move with the stock, so a good rating is worth somewhat less than it looks and a poor one may be worth more. Improving against a frozen benchmark is not the same as improving, and that is worth saying before a rating is used to support a target or a public claim.
That is a different question and this page does not answer it. A Display Energy Certificate is one of the UK scheme’s four compliance routes, and whether it discharges the duty depends on scope — on whether what the certificate covers matches what the obligation reaches. Our Energy Audit Obligation Checker works through both regimes properly. This page is only about whether the rating on the certificate means what it appears to mean.