Explainer  ·  Issue 8  ·  Tuesday 15 September 2026

Watts Happening

 

The UK energy newsletter that doesn't work for the industry

The Spark

One real retrofit, surveyed twice. Worked the old way, the house needed a heat pump around 30% larger than the way the standard now requires.

Nothing about the building changed. What changed, in January, was how the industry counts the heat that leaves your house in the air rather than through the walls.

Around 93% of British homes were built before 2000. Those are the ones the old method treated most harshly, and the number it produces decides the size of the heat pump you get quoted for.

This Week in Watts

Two kinds of draught, only one of them is worth paying to stop

A house loses heat two ways. Through the fabric, conducting out through walls, roof, windows and floor. And through air, warm air leaving and cold air arriving to replace it. Everyone talks about the first. Almost nobody shows you the second.

Split the air half and it gets useful. Some is uncontrolled infiltration: draughts around frames, an open chimney, unsealed floorboards. Nobody designed those. The rest is deliberate ventilation through trickle vents and extract fans, and it exists because a house with no air exchange gets damp and unhealthy. The standard states a minimum of half an air change an hour in habitable rooms, because anyone sitting in a stuffy room opens a window and undoes the lot.

The leaky half is where the easy money sits. Failed sealant around window and door frames is the most common thing I find, and the fix is a tube of sealant rather than a project. Extract fans are next. The duct wants sealing where it passes through the wall, and there should be a cover outside with a flap that shuts when the fan is off. Without one you have left a hole in your wall with a breeze coming down it.

Until this year the calculation treated both with a guess. Each room got an air change rate from a table, by room type and roughly when the house was built. The CIBSE Domestic Heating Design Guide 2026, published in January, moved away from those defaults, and MCS took its figures from BS EN 12831-1:2017 instead. The replacement starts with the building: how leaky the structure is, then how much of each room has an external wall. Party walls are disconted, because there is a heated space next door. Chimneys and extract fans go back in, since a pressure test blocks them off.

Ventilation heat loss  ·  one room

36 m³ Room volume
Measured on site
0.5 ACH Air changes per hour
The only judgement call of the four
23 °C Temperature difference
20°C inside, −3°C design day
0.33 Constant
Heat capacity of air
0.33 × 0.5 × 36 × 23  =  137 watts

That is one room on the coldest day of the year, not an annual bill. Anyone turning a heat loss figure straight into yearly savings is selling something.

OpenEnergyMonitor, who publish monitored data from real installations rather than models, report the compliant calculation effectively halves ventilation losses in pre-2000 naturally ventilated buildings. One caveat: that halving applies when rooms are combined to size the heat pump. Radiators are still sized on each room's own figure.

Then the money. Oversize a heat pump and it cycles, switching on and off because it cannot modulate below what the house needs on a mild day. Spruce is blunt: shorter compressor life, lower SCOP, higher running costs. Size the radiators to match an overstated demand and you specify triple-panel radiators throughout, which Spruce says is where a lot of these projects die at proposal stage. A little oversizing, 10 to 15%, does beat undersizing. This is not that. This is a number nobody measured.

Every default in the old method erred the same way. None of them erred towards a smaller machine.

The Bill Buster

Three questions, and one thing not to touch

Ask which method the heat loss calculation used. A per-room rate from a table, by room type and house age, is the old approach. Starting from the air permeability of the whole building is current.

Ask to see the ventilation figure separately from the fabric figure, and whether that permeability came from a test on your house or an assumption from its construction. Both are legitimate. Only one is a measurement.

Ask whether an EPC was used anywhere in the sizing. It should not have been. A full room-by-room survey is required for MCS certification and for a Boiler Upgrade Scheme grant, and an EPC is not an acceptable substitute.

And leave the deliberate ventilation alone. If you have an open-flued gas fire, a stove or a live chimney, sealing around it changes how that appliance vents. That is a job for someone qualified.

→ Get the ventilation and fabric figures separately, before you agree a system size

Watts on the Market

Unsold this week  ·  Reader resource

A heat loss calculation tells you what your house needs on the coldest day. It says nothing about what your current setup costs you if you change nothing. The Cost of Inaction calculator runs your energy forward 25 years, on your own bill and property type, across two inflation scenarios. No installer at the end of it.

→ Run your 25-year number at calculator.wattshappening.energy

Quick Watts

Heat pump installs

20,895 heat pumps went into existing UK homes in the first half of 2026, down 22% on the 26,768 a year earlier. The government's ambition is 450,000 a year by 2030. DESNZ classes the series as still in development.

(DESNZ, 10 September 2026)

Capacity market

DESNZ is consulting on how Demand Side Response is de-rated, proposing either a duration-based or a technology-based method. It closes on 27 October. Worth a look if you have flexible load or storage earning capacity payments.

(GOV.UK, 11 September 2026)

Keep Reading

View more
caret-right