Most of what we do rests on modelled performance — SAP, HEM, U-values and PSI-values that predict how a home should behave. This month the government put its weight behind measuring how a home actually behaves. The Department for Energy Security and Net Zero (DESNZ) has published a Smart Meter Enabled Thermal Efficiency Ratings (SMETER) Strategic Guide, along with a supporting evidence base and case studies. It won’t change a compliance calculation tomorrow, but it’s a clear signal of where things are heading.
What SMETER actually measures
A SMETER assessment looks at how well an occupied home holds onto heat. Most methods combine actual energy-use data, indoor and outdoor temperatures and local weather to work out how quickly heat escapes the building. The result is expressed as a Heat Transfer Coefficient (HTC) — in watts per kelvin — for the whole dwelling.
The rule of thumb is simple: a high HTC means the home is leaking heat quickly, while a low HTC means it’s retaining heat well. Crucially, this is a measured, in-use figure rather than a value derived from construction assumptions.
To be clear about what it is not: a SMETER assessment does not replace an EPC or a detailed survey. It sits alongside modelled assessments, giving real-world evidence of how a building performs against how it was predicted to perform.
Why this matters: the performance gap
For years, decisions about energy efficiency have leaned on modelled assessments and assumptions about how a building was built. Those models remain essential — but hidden defects, unknown or missing insulation and air leakage all mean the real building can behave differently from the one on paper. The difference between predicted and measured performance is the performance gap, and it’s exactly the thing measured HTC can expose.
The guide’s evidence base is the interesting part. It highlights the MEASURED project — a field trial run by Build Test Solutions, Veritherm UK and Elmhurst Energy across 56 homes — which tested whether measured heat loss could improve heat-pump surveys and design. The headline finding is striking: the traditional heat-loss calculation matched the measured result in only 30% of homes. It overestimated heat loss in 59% of cases and underestimated it in 11%. In other words, for roughly 70% of the homes, relying on the standard BS EN 12831 calculation alone could have led to a heat pump being sized incorrectly.
That’s a useful reminder for anyone specifying low-carbon heating under the Future Homes Standard: oversized kit costs more and can run inefficiently, while undersized kit leaves homes cold. Measured data offers a way to sense-check the assumptions.
An important caveat
Measured HTC tells you how much heat a building is losing overall — not where it’s being lost. So it’s best used to calibrate and strengthen detailed, room-by-room calculations, not to replace the wider heating-design process. If the number looks off, the follow-up is the familiar toolkit: airtightness testing, thermal imaging and in-situ U-value measurement to pin down where the losses actually are.
Where it’s heading — EPCs and policy
This is where assessors should pay attention. The government has consulted on a voluntary option to record validated, quality-assured SMETER HTC values on future EPCs, alongside an HTC produced by the Home Energy Model, and is analysing the responses. DESNZ is also building the validation and quality-assurance systems that measured performance would need, and has signalled interest in using SMETER within future policy, including the Warm Homes Plan.
None of that makes measured HTC a general requirement today. But taken together with CIBSE TM71 — the new framework for measuring and reporting HTC values across different methods — it points to a future where calculated and measured figures are used side by side: the model for a consistent assessment framework, the measurement for evidence of how a specific home really performs. With smart meters now in most homes, the raw data is already there.
What to do with this now
There’s nothing to comply with yet, so treat this as horizon-scanning rather than a deadline. The practical takeaways: expect measured performance to feature more heavily in retrofit funding, quality assurance and eventually EPCs; be ready for clients to ask how a modelled result compares with real-world data; and remember that a strong measured HTC still comes down to the fundamentals — fabric, airtightness and well-detailed junctions.
That last point is the one we’d flag hardest. Measured performance will increasingly hold designs to account, and the homes that measure well are the ones where the thermal bridging and fabric detailing were right on paper first. If you’d like your PSI-values and fabric performance modelled properly so the as-built numbers stand up to real-world measurement, get in touch and we’ll tell you how we’d approach it.
Sources: Smart-meter-enabled thermal efficiency ratings (SMETER): Strategic Guide, GOV.UK · SMETER case studies and evidence base annex, GOV.UK · CIBSE TM71: Measuring Heat Transfer Coefficients in Buildings (2026) · Elmhurst Energy: New SMETER guidance — from measurement to action