Heat pumps in the UK in 2026: an honest state-of-play

I've now spent enough years living with a heat pump, talking to other owners, and reading performance data to be reasonably sure of two things. First, the technology works. Second, the public conversation about it is broken in both directions — too triumphalist on one side, too cynical on the other.

The triumphalists tell you heat pumps will save you money on day one and reduce your emissions by half. The cynics tell you they're a Westminster fantasy that doesn't work in British weather. Neither matches the data. The reality is more interesting and considerably more practical: heat pumps work very well in well-prepared homes, work poorly in unprepared homes, and the quality of the install is the single biggest variable in whether you end up happy or angry. The BUS grant of £7,500 changes the economics enough to make the conversation worth having for a much wider range of homes than was true two years ago — but it doesn't make the building physics easier.

So, with the qualifications out of the way, here's where I think we actually are in 2026, what the typical install really costs, what the running cost really looks like, and the honest tradeoffs for the homes that shouldn't go heat-pump first.

What the data is actually showing

The clearest signal we have is from the multi-thousand-home monitoring exercises that have been running on UK installs since the late 2010s. The honest summary is that well-installed air-source heat pumps in the UK consistently achieve seasonal performance factors (SPF) between 2.8 and 3.6 in real homes, with the median sitting around 3.0–3.2. Top-decile installs reach 3.8–4.2. Bottom-decile installs — usually because the radiators are wrong, the flow temperature is too high, or the system was sized off a quick back-of-the-envelope rather than a heat-loss survey — drop to 2.2–2.6, which is where the horror stories come from.

Translated into plain English: every kilowatt-hour of electricity you put into a well-installed heat pump becomes about three kilowatt-hours of heat in your home. A badly installed one gets you closer to 2.2. That difference is the difference between cheaper than gas and considerably more expensive than gas at current tariffs.

What the data does not show is any meaningful correlation with house age once you control for fabric performance and radiator size. Victorian terraces with appropriate fabric upgrades and properly sized emitters perform within a few percent of modern builds with the same setup. The narrative that 'heat pumps don't work in old houses' is false; the narrative that 'heat pumps work in any house regardless of fabric' is also false. Fabric matters. Radiator sizing matters. The age stamp on the deeds doesn't.

What an install really costs in 2026

Here are the numbers I see consistently across the UK in 2026:

• Air-source heat pump (ASHP), straightforward retrofit: £8,000–£14,000 before the BUS grant, depending on output, brand tier, and how much radiator work is needed. After the £7,500 BUS payment, your out-of-pocket cost typically lands at £500–£6,500.
• ASHP with significant radiator upgrades or hot-water cylinder install: £12,000–£18,000 before grant. Plan for £5,000–£11,000 net.
• Ground-source heat pump (GSHP): £18,000–£28,000 typical, the spread driven mostly by ground-loop installation cost. After the £7,500 grant, £10,500–£20,500 net. Often only viable if you already have the land and the disruption tolerance.
• Cylinder + controls only (where you already have a recent ASHP): £1,500–£3,500.

The number that varies most is the radiator upgrade. We've seen households quoted £400 for a single rad swap and others quoted £4,000 for a comprehensive emitter upgrade. The honest answer is that proper sizing for the lower flow temperatures heat pumps prefer (35–45°C rather than 70°C) often means at least one or two rooms need a bigger emitter. If a quote leaves the radiators alone entirely, ask why — sometimes the existing rads are already oversized for a fabric-improved house, but more often the installer is hoping for the best.

Running cost: the calculation that actually matters

This is where the conversation gets interesting and where most published analyses go wrong. A heat pump's running cost depends on three numbers and only three numbers: your annual heat demand, the SPF you actually achieve, and the price you pay per kWh of electricity. Multiply, divide, done.

For a typical 3-bed semi with reasonable fabric (EPC C), annual space-heating demand lands somewhere around 9,000–13,000 kWh of heat. With an SPF of 3.0, that's 3,000–4,300 kWh of electricity. At a flat 28p/kWh standard tariff, that's £840–£1,200/year for space heating, plus another £300–£500 for hot water.

For comparison, the same home on a gas boiler at 6.5p/kWh and 88% efficiency uses about 10,200–14,800 kWh of gas for space heating, costing £660–£960. Add £200–£350 for hot water.

On a flat tariff, gas is cheaper. The narrative that heat pumps save you money automatically — it's not true at flat-rate electricity prices. The arithmetic only flips when you move to a heat-pump-specific time-of-use tariff, where overnight rates can drop to 12–15p/kWh and the system charges hot water and pre-heats the house cheaply. That's the £200–£500/year saving that gets quoted in marketing literature, and it's real, but it requires both the tariff and a system that can actually shift load.

The homes that should not go heat-pump-first

I'll be unfashionable about this. There are UK homes where a heat pump is not the right first move, and pretending otherwise just produces unhappy customers and reputational damage for the technology.

1. Solid-wall homes with no fabric upgrades planned. The flow temperature you'd need to keep these warm in January is high enough that the SPF collapses. Insulate first, heat pump second.
2. Microbore-plumbed homes where the entire system needs replumbing. The additional £4,000–£8,000 of pipework can dwarf the heat-pump cost itself. Sometimes worth doing as part of a bigger renovation; rarely worth doing alone.
3. Properties on cheap LPG or oil with limited mains-electric capacity. An electrical supply upgrade adds £1,500–£3,500 and weeks of waiting.
4. Homes due for sale within 2–3 years where the SAP-rating uplift won't fully land in the buyer's mind. The economic case is built on 10+ year ownership.
5. Listed buildings or conservation-area properties where external unit placement triggers planning friction. Solvable, but rarely fast.

This isn't an argument against heat pumps. It's an argument for sequencing. The right answer for a poorly insulated 1930s semi is fabric-first retrofit, then heat pump, in that order, with a year between to let the heat-loss numbers re-stabilise.

The BUS grant: what it actually changes

The Boiler Upgrade Scheme grants £7,500 for an air-source or ground-source heat pump. It's available to owner-occupiers of existing homes (not new builds), through MCS-certified installers using RECC-registered processes. Self-builds and second homes are eligible. Social housing and rented properties go through different routes.

What the grant does, in practical terms, is take the median net cost of a well-specified 8kW ASHP install down from around £11,000 to £3,500. That's not free. But it's enough that the payback against a high-running-cost gas boiler — particularly if you're on a heat-pump tariff with overnight rates — comes in around 7–11 years for a typical household. Without the grant, payback was 14–18 years and only made sense for owners with strong climate motivation. With it, the calculation is closer to a normal capital-investment decision.

The grant does not cover insulation, hot-water cylinders that aren't part of the heat-pump install, controls retrofits, or radiator upgrades unless they're bundled into the MCS quote. It also can't be combined with ECO4 funding for the same measure. Read the small print.

Choosing an installer: where most of the variance lives

If I had to give one piece of advice to a household considering a heat pump, it would be this: spend more time choosing the installer than choosing the equipment. Brand matters less than people think. Install quality matters more than people think.

What to look for:

• A proper room-by-room heat-loss calculation. If they quote you off a single number 'kW per square metre' rule of thumb, walk away. Real heat-loss surveys take 2–4 hours and include U-value estimates, ventilation rates, and per-room emitter sizing.
• A flow-temperature design target written down. A good design hits the design heat-loss day at 45°C or lower. If the installer is targeting 55°C+, you're paying for a heat pump that will run like a bad gas boiler.
• Radiator-by-radiator emitter sizing. They should tell you which radiators are staying, which need swapping, and what the new outputs will be at the design flow temperature.
• Clear separation between MCS-required documents and your own records. Ask for the heat-loss workbook, the emitter schedule, and the controls strategy as separate deliverables.
• Twelve months of post-install support included. Heat pumps need tuning over their first heating season. An installer who walks away on day one is a warning sign.

What I'd do in 2026 if I were starting from scratch

If I were starting again on a typical mid-quality UK home — say, a 1970s semi at EPC D — here's the order I'd run it in. This is genuinely how I think about it for friends and family who ask.

1. Get a proper heat-loss survey done first, ideally by an independent assessor not selling the install. £400–£800. Tells you what the building physics actually looks like.
2. Address the cheapest fabric wins. Loft insulation to 300mm, draught sealing, hot-water cylinder insulation, suspended-floor underfloor work where accessible. Total: £1,500–£4,000.
3. Run the existing system for one heating season post-fabric. Confirm the new heat-loss profile against your gas or oil bills.
4. Re-run the heat-pump quote against the new heat-loss figures. The system is often a kW or two smaller than before, which saves money.
5. Install heat pump + cylinder + radiator upgrades together. Don't drip-feed; the engineering is interconnected.
6. Move to a heat-pump-friendly tariff at the same time. Set up overnight charging and pre-heating strategies in the controls.
7. Tune through the first heating season. Adjust weather compensation, schedules, and zone settings. This is where SPF moves from 2.8 to 3.4.

Heat pumps in 2026 are a genuinely good piece of technology with a difficult installation ecosystem. The technology delivers what it promises when fitted properly and tuned thoughtfully. The installer market is uneven, the public conversation is polarised, and the building stock is varied enough that no single recommendation fits every house.

The honest answer for most UK households at EPC C or better is: yes, a heat pump is probably the right next step, take the BUS grant, but pick the installer with care and budget for radiator and controls work. The honest answer for many EPC D and worse households is: fabric first, then heat pump, with a 12–18 month gap. The technology isn't the bottleneck. The sequencing is.