Our editorial method: how we research and rank heat pumps

Heat Pump Report is free to read, and today no manufacturer or installer pays us. We plan to earn money from lead generation once homeowners can ask installers for a price through the site; nobody pays for a place in our ranking or for what a brand review says. Every figure on our pages comes from a source we check before publication or is labelled as our own assumption, and our worked examples use two example homes whose inputs are set out below.

At a glance

• Funding. Free to read; no manufacturer or installer pays us. Later, installers will pay us for price requests made through the site. Placement in our ranking and brand reviews is never for sale.
• Example homes. A 1970s gas semi (97 m², 9,500 kWh of gas a year) and an off-gas oil detached house (149 m², 1,480 litres a year), both in England, with 85% boiler efficiency as our assumption.
• Sources. Law and official bodies first, then manufacturers' UK documents, then field trials and statistics.
• Ranking. Efficiency at 55 °C counts 40%, outdoor sound 25%, warranty at no extra charge 25% and maximum flow temperature 10%. We test nothing ourselves.
• Corrections. Tell us through the contact page; we check the primary source and correct every sum affected.

Who we are and how we are funded

Our guides are published under the name of the Heat Pump Report editorial team. We don't invent authors or reviewers, and we don't present anyone as a heating engineer or installer who has checked a page for us.

Drafts are written with software, including AI language models, from a brief and a record of every fact the page uses, with its source and the full arithmetic of every sum. Two separate review passes follow, also software-assisted. One re-checks every statement and sum against its source and against UK advertising and consumer rules, including the CAP Code and the Digital Markets, Competition and Consumers Act 2024; the other checks that the page answers your question clearly. Their findings are fixed in one pass, and nothing goes live until the publisher approves it.

How we plan to earn. Quote requests to installers aren't open on the site, and today no manufacturer or installer pays us. Later, homeowners will be able to ask installers for a price here, and the installers will pay us for those requests. Only the installer or installers you choose will then receive your details, the request will be free and without obligation for you, and we can't promise whether, when or how many installers reply. Before the first paid request goes through, we will say so on this page.

What is never paid for. Neither manufacturers nor installers pay to appear in our ranking, to move up it or to change a brand review. A model's place comes only from the criteria further down this page.

Why we err on the cautious side. Because we plan to earn when homeowners ask installers for a price, we benefit if you choose a heat pump, so our running-cost examples show a less favourable case beside the central one.

Our two example homes

Every worked example uses one of two homes, so figures stay comparable from page to page.

Where the inputs come from. The floor areas are the average usable floor area of semi-detached and detached homes in the English Housing Survey 2018-19. The 9,500 kWh is Ofgem's typical gas use for a medium user, the figure behind its "typical bill" from 1 July 2026. For the oil house, with no official typical figure, we apply the 2024 median gas use per square metre of detached houses in England and Wales from the government's National Energy Efficiency Data-Framework: 149 m² × 102 kWh/m² ≈ 15,200 kWh. At about 10.3 kWh per litre of heating oil (government energy statistics), that is 1,480 litres. Using a gas-house figure for an oil house is our choice; off-gas homes are often older and larger, so real use may be higher.

Useful heat is the heat that reaches your rooms and hot water. We assume the boiler turns 85% of its fuel into useful heat: 9,500 kWh × 0.85 = 8,075 kWh, and 15,200 kWh × 0.85 = 12,920 kWh. The 85% is our assumption for the existing boiler, not a measurement; where a page compares a heat pump with a new gas boiler, we use 93%, the factor MCS's pre-sale calculation method gives a new boiler.

We gave the semi a combi boiler because replacing one raises the hot-water cylinder question, not because most homes have one.

The assumptions behind every worked example

SCOP says how many kilowatt-hours of heat a heat pump delivers over a year for each kilowatt-hour of electricity. In the government-funded Electrification of Heat trial, the median for 428 air source heat pumps was 2.78, with the middle half between 2.55 and 3.05. We work with 2.8 and always show 2.5 and 3.1 beside it (more on efficiency).

Energy prices. Both rates are Ofgem price-cap averages for Great Britain, paying by direct debit. Ofgem resets the cap every three months, so every rate carries its period, and GB cap rates are never used for Northern Ireland. Heating oil is outside the cap and its price moves quickly, so we show at least two periods from the government's oil price statistics, VAT included.

Prices and grants. Our installed prices are medians from the government's Boiler Upgrade Scheme statistics for England and Wales. MCS installers report the installed price on the installation certificate. It covers equipment, labour and VAT and is meant to include the part the grant pays, and the top and bottom 10% of reports are left out. So a median is a price before the grant is taken off, not a quote (more on costs). We don't collect or analyse quotes ourselves. A grant is always its own line with its nation: the Boiler Upgrade Scheme grant (£7,500 for an air source heat pump, if the home qualifies) applies in England and Wales, Scotland has its own scheme, and Northern Ireland has no general homeowner grant (more on grants).

How we calculate. Useful heat ÷ SCOP = electricity; electricity × unit rate = cost. For the semi at SCOP 2.8: 8,075 kWh ÷ 2.8 = 2,884 kWh, and 2,884 kWh × 26.32p = £759 a year.

At SCOP 2.5, 2.8 and 3.1 the semi's electricity costs £850, £759 and £686 a year, against £757 of gas units at 7.97p/kWh. Electricity is at the 1 October to 31 December 2026 cap rate with no VAT in Great Britain until 31 March 2027, gas includes 5% VAT, and standing charges, servicing and tariff changes are left out. At SCOP 2.8 the heat pump costs roughly the same to run as the gas (break-even is SCOP 2.81: 0.85 × 26.32 ÷ 7.97), and at 2.5 it costs more. Running costs and the oil-boiler case are taken further on running costs and replacing an oil boiler.

Where our facts come from

When two sources disagree, the higher tier wins:

1. Law on legislation.gov.uk, named by instrument and section, with amendments in force.
2. Official bodies and schemes: government departments, Ofgem, MCS and its standards, Home Energy Scotland, the Welsh Government, nidirect, the Planning Portal and the RECC and HIES consumer codes.
3. Energy suppliers, for their own tariffs only.
4. Products: the manufacturer's UK datasheet, installation manual and UK warranty terms, and the MCS product directory.
5. Public field trials and statistics, such as the Electrification of Heat trial and the Boiler Upgrade Scheme figures: we take averages and ranges from them, never a rule or a grant amount.
6. Context only: news and trade press can alert us to a change but are never the source of a figure.

Comparison sites, other advice publishers, installer blogs, shops and discussion boards are never a source. We don't link to other websites; we name the source in the text instead. For every page we keep an internal record of each checkable statement and its source.

Keeping figures current

Every page is checked against its sources before it is published and whenever it is revised. We don't promise a fixed review cycle.

• Figures that change, such as cap rates, oil prices, grant windows and VAT end dates, sit in a table on each page with their period. We re-check them when Ofgem sets a new cap, when new Boiler Upgrade Scheme statistics come out and when a scheme changes.
• The ranking is re-scored whenever we revise it, from a fresh reading of the MCS product directory, datasheets and warranty terms.

How our heat pump ranking is scored

Our comparison of air source heat pumps scores models only from published data: the MCS product directory, manufacturers' datasheets and their UK warranty terms, on the same basis for every model. We don't repeat anyone else's test results.

Who can be ranked. A model must be an air-to-water heat pump and hold a current MCS product certificate (MCS certifies products as well as installers). Refrigerant is not a condition, though every model we compare uses propane (R290). A brand whose UK homeowner pages give no warranty length or terms, such as Samsung, is not ranked; the ranking page says why.

One model per brand. From each current UK air-to-water range we take the single-phase size whose heat output at −7 °C outside and 35 °C flow, as the manufacturer states it, is nearest 8 kW (on a tie, the smaller; without that figure, the nominal size). If a maker has several ranges, the one with the most points counts; if a model comes with several indoor units, we use the MCS combination with the lowest SCOP at 55 °C. Why 8 kW: it was the median capacity of the air source heat pumps installed under the Boiler Upgrade Scheme in England and Wales from July to September 2026, with the middle half between 6 and 10.2 kW.

One register for efficiency. Every SCOP comes from the MCS product directory, which lists one for each flow temperature from 35 to 65 °C; we never mix in datasheet values.

Each criterion becomes 0 to 100 points, then counts with its weight:

A worked score for an imaginary model. MCS SCOP 3.70 at 55 °C gives (3.70 − 3.20) × 80 = 40 points; 50 dB(A) gives (60 − 50) × 5 = 50; a seven-year warranty at no extra charge gives 100; a 70 °C maximum flow gives (70 − 55) × 5 = 75. Weighted: 0.40 × 40 + 0.25 × 50 + 0.25 × 100 + 0.10 × 75 = 16 + 12.5 + 25 + 7.5 = 61 points.

Why these weights. Efficiency at 55 °C drives the electricity bill, above all in an older home with radiators. Sound decides where the outdoor unit can go; what reaches a neighbour's window also depends on distance and walls (more on noise). How long a heat pump lasts can't be compared from published data, so a written warranty is the closest promise we can score. Maximum flow temperature is headroom for cold days, not a target, so it weighs least.

The UK warranty rule. Most UK heat pump warranties depend on the installer: the longer cover applies when an installer in the maker's own scheme registers the job, often with an annual service as a condition. We count that cover where the maker says it costs nothing extra or names no fee, and the ranking page lists each maker's conditions. Cover granted only to the installer scores 0; paid extensions are named but not scored. The ranking page also shows the scores counting only cover without an installer tier or scheme membership.

What the points don't tell you. Other weights could change the order, and for your home the heat loss, radiators and installation quality matter more than a few points. Our brand reviews show what raises or lowers a range's score by our criteria.

Scores for brands. For the brand ranking on our home page and brand reviews we score each brand out of 5: quality from the points for efficiency, sound and flow temperature (40:25:10; 0 points give 2.5, 100 give 5), warranty from 3.5 for the usual two-year manufacturer warranty plus 0.25 for each further year at no extra charge (0.25 less where it needs registration, servicing or an installer tier; 3 for unclear terms, 2.5 for no cover at no extra charge), plus our estimate of the brand's usual price level (5 = most affordable), weighted 50:25:25.

What we don't do

We test nothing: no laboratory runs, home visits, noise measurements or inspections of installers' work. We award no titles or seals and host no customer reviews. MCS certification is a requirement, not proof of a good installation (see how to choose an MCS installer), so we never present an installer as checked or recommended by us. We give no individual tax, legal or financial advice: whether a grant, loan or rule applies to you is for your installer, Ofgem, Home Energy Scotland, your local planning authority or an independent adviser to confirm. Carbon figures appear only as an estimate with the emission factor, its source and its year.

Corrections

If something looks wrong or out of date, tell us through our contact page: the page's address, the sentence and your source if you have one. We check it against the primary source; if we got it wrong, we correct the text and every example and table that depends on it. Wrong rules, grant amounts and prices come first.

We change a brand review or the ranking at a manufacturer's or installer's request only where a fact is shown to be wrong or out of date. Criteria, weights and scores don't change on request.

Figures that change

Frequently asked questions

How does Heat Pump Report make money?

The site is free to read, and today no manufacturer or installer pays us; quote requests to installers aren't open here. We plan to earn from lead generation: once homeowners can request a price through the site, the installers will pay us for those requests, and only the installers you choose will receive your details. Nobody pays for a place in our ranking or for what a brand review says.

Who writes your guides, and do you use AI?

Our guides appear under the Heat Pump Report editorial team's name; we don't invent authors or experts. Drafts are written with software, including AI language models, from a brief and a record of every fact and sum. Separate software-assisted passes then check each figure against its source and check that the page answers your question, and the publisher approves it before it goes live.

Do you test heat pumps yourselves?

No. We score models from the MCS product directory, manufacturers' datasheets and their UK warranty terms, and we don't repeat other organisations' test results. The ranking compares what makers publish in writing, on the same basis for every model.

Why do you assume an 85% efficient boiler?

It is our assumption for an older working boiler, not a measured figure, and every page that uses it says so. If your boiler is less efficient, the same fuel bill means less heat reaches your rooms, and a heat pump would need less electricity than our example shows.

Why do you compare heat pumps of about 8 kW?

Because 8 kW was the median size of air source heat pumps installed under the Boiler Upgrade Scheme in England and Wales from July to September 2026, so the model nearest 8 kW from each brand compares like with like. The right size for your home comes from the installer's room-by-room heat-loss calculation, not from our ranking.

How do I report a mistake?

Use our contact page and give the page's address, the sentence and, if you have it, your source. We check it against the primary source and, if it is wrong, correct the text and every sum that depends on it.