EPC ratings are the single most quoted number in the UK housing market, yet the government’s own technical annex shows they measure something few homeowners realise. The rating is not about how much energy your home actually uses. It is about the modelled cost of heating a standardised space using standard assumptions.
As reported by GOV.UK, the technical annex for chapter 2 of the EPC methodology explains that the rating is derived from ‘the cost of heating, lighting and hot water in a standardised way’, not from your actual meter readings or even your thermostat setting.
What the EPC actually measures, and what it misses
The rating is built on a model called RdSAP (Reduced Data Standard Assessment Procedure). It uses a handful of inputs: wall type, loft insulation thickness, window glazing, boiler age, floor area. It then calculates an energy cost per square metre under a notional occupancy pattern. The result is a letter grade from A to G.
But the model makes assumptions that can mislead. It assumes every household heats the whole home to 21°C in the living room and 18°C elsewhere, regardless of how you actually live. It assumes a standard number of occupants, typically 2.3 people, even if you live alone. It assumes all rooms are heated equally, even if you keep the spare bedroom cold.
The catch is that a home with thick loft insulation and double glazing can score a B even if it has an old gas boiler that costs £1,200 a year to run. Meanwhile, a home with a modern heat pump and solar panels but solid brick walls might scrape a D, because the model penalises the higher unit cost of electricity over gas.
Why this matters for your bill, and your EPC target
For a typical 3-bed semi using 12,000 kWh of gas and 3,000 kWh of electricity, the difference between a D and a C rating can mean £200–£300 a year in modelled costs. But the real-world difference depends entirely on how you use energy. A family that heats to 19°C and showers at off-peak times will spend far less than the model predicts, and a household that runs the heating 14 hours a day will spend far more.
The government is consulting on raising the minimum EPC rating for rented homes to C by 2028. Landlords are scrambling to meet that target, often by installing cheap LED bulbs and topping up loft insulation rather than addressing the real fabric issues that cause heat loss. The technical annex confirms that these quick wins do improve the score, but they may not reduce the tenant’s bills by much if the walls are single-skin and the windows are draughty.
What homeowners should actually do
If you are planning eco upgrades, do not let the EPC target dictate your decisions. The rating is a useful benchmark but it is not a measure of comfort or actual energy use. The Energy Saving Trust recommends starting with a home energy audit that includes a thermographic survey and a draught-proofing check, not just the RdSAP model.
Heat pumps, for instance, are penalised by the current methodology because electricity costs more per kWh than gas. Yet they can cut a home’s carbon emissions by 60–70% and, when paired with solar panels, can reduce bills by £400–£600 a year. The EPC does not capture that. Similarly, triple glazing improves the score modestly, but its real benefit is comfort and noise reduction, not just the letter grade.
Ofgem’s data shows that homes with heat pumps and solar panels typically have lower actual bills than the EPC predicts, because the model overestimates electricity use. The government’s own consultation on EPC reform, due in 2025, is expected to address this distortion, but for now, the rating remains a lagging indicator of low-carbon living.
Households on standard variable tariffs can check their actual energy use via their smart meter or supplier portal. Compare that to the EPC’s modelled figure. If the gap is large, more than 20%, the rating is likely misleading you. Invest in measures that reduce your actual consumption, not just the score. By 2026, the government plans to introduce a ‘real-world EPC’ that uses smart meter data. Until then, take the letter with a pinch of salt.
Frequently Asked Questions
Yes, because it is based on a standardised model (RdSAP), not your actual energy use. A home that scores a C may still have high bills if it is draughty or poorly heated. The rating is a guide, not a guarantee.
If you are selling, a better EPC can help with marketing, but buyers are increasingly asking about actual bills and running costs. For rentals, the minimum C target by 2028 means landlords will need to act, but focus on cost-effective fabric upgrades first, not just the cheapest fix.
