Earth Notes: UK Homes Wet Heating Retrofits Needed for Net Zero (2023)

Updated 2023-05-24.
As of 2023 how many UK homes need wet space heat pump retrofit for Net Zero? #heating #retrofit #research


If the UK is to meet its Net Zero targets so as to have sufficient impact on reducing climate change, the ~10% to 20% of UK greenhouse gas emissions from home heating will need to be minimised or eliminated. This paper takes a quick snapshot of the UK in 2023 and estimates the number of already-standing homes that will need their current gas-fired wet (radiator-based) heating systems retrofitted by 2050, and some hard-to-treat cases. This snapshot also briefly drills down via England to the suburb of London (Kingston) where the author lives. In practice it appears that virtually all of these gas-boiler homes could have their space heating electrified with heat pumps to good climate effect.


UK, England, Kingston-upon-Thames, 2023, snapshot, homes, domestic, housing stock, housing, retrofit, heating, space heating, central heating, wet, hydronic, optimise, heatpump, heat-pump, gas, boiler, net zero, climate, climate change, emissions


If the UK is to meet its Net Zero targets, the ~10% to 20% of UK greenhouse gas emissions from home heating will need to be minimised or eliminated [reguis2021challenges] [govUK2022boiler].

Most of the dwellings that will be in use in the UK in 2050 are already built and are quite old compared to the UK's European neighbours [miller2020scaling] [fabbri2023insulation] [govUK2022dwellings] [piddington2020housing]. For example see from Table 2 below that a large fraction of English residential stock is pre-1945 and the majority pre-1965.

Partly because of that age profile, and partly because of the UK's cheap heating fuel sources in more recent times, eg abundant North Sea gas, UK homes are markedly less energy efficient than those of neighbouring countries too. ~80% of those existing homes have gas-fired wet (ie hydronic, with radiators) central heating [EHS2022energy] [henretty2020efficiency]. That range of housing archetypes and ages, and issues such as DHW storage and noise rules for external ASHP units mean that each inefficient home is inefficient in its own way. Customised solutions will be slower and more expensive than otherwise.

(For example, Kingston-upon-Thames has many pre- and inter- war leaky houses, and ~10% of the stock is in conservation areas which adds a ~10% heating efficiency penalty as outlined below [RBKdata] [govUK2022CTdataset] [fetzer2023conservation].)

What should those gas boiler wet systems be replaced with? They will need to be replaced if the UK is to hit Net Zero.

One alternative is district (aka community aka network) heating already popular in London (67% of new dwellings in London) [henretty2020efficiency] that can, for example, make use of the large amount of waste heat available [sorensen2023excess]. Though especially attractive in dense urban areas, this is unlikely to cover more than a small minority of UK dwellings' heat demand by 2050 [CCC2015district] [CCC2020sixth]. (A further nuance is that low-temperature district heat may be used in conjunction with in-home heat-pumps, so there will be overlap.)

Another oft-touted alternative is hydrogen-based space heating, ie converting the UK's natural-gas grid and infrastructure and home gas boilers to use hydrogen.

There are many problems with this use including safety and cost, but the most pertinent is the thermodynamic inefficiency of burning hydrogen made renewably, eg by hydrolysis from wind-generated electricity, rather than using that electricity to directly drive a heat pump: the latter is widely acknowledged to be better by a factor of ~5x, so only a few percent of 2050 home heating at most is likely from hydrogen [BEUC2022green] [rosenow2022hydrogen] [korberg2023hydrogen] [ryland2022economics] [CCC2020sixth].

The most obvious remaining contender is replacement of gas boilers with heat pumps (possibly hybrid, with natural gas or hydrogen or in conjunction with district heating [reiners2021district]), ideally with insulation and ventilation improvements also. Such a replacement can reduce household energy demand by a factor of 2x or 3x, difficult to match with any other intervention [ryland2022economics] [rosenow2023clean]. Using electricity from an increasingly green electricity grid to power heat pumps is a decarbonisation force multiplier.

(For another small proportion of low-demand dwellings, about the bottom decile according to [ryland2022economics], direct electric heating (DEH) may give the lowest total cost of ownership, and thus DEH may be used rather than heat pumps.)

This snapshot simply estimates from existing UK datasets the number of such gas boilers in existing dwellings that could (and likely should) be replaced with heat pumps, and also summarises some difficult cases.

It turns out that there are very few gas boilers that cannot be reasonably replaced with heat pumps, the UK's Climate Change Committee (CCC) and others make clear [CCC2019technical] [rivers2020retrofit] [rivers2020decarbonising] [ESC2022EoH] [ESC2021suitable].

The simple recalculated values here are sense-checked against the CCC's view.


This snapshot looks at authoritative housing stock data from the UK government, and makes a simple estimate of 2050 UK dwellings already built that currently have gas wet central heating and that are candidates for air-to-water Air-Source Heat Pumps (ASHPs) or Ground-Source Heat Pumps (GSHPs) in particular as low-carbon replacements.

This also looks at some nuances and impediments to replacing such gas boilers with heat-pumps to investigate whether there is any reason to believe that heat-pumps are not an available near-universal solution to decarbonise.

This paper then compares these first-order estimates against the views of the UK's The Climate Change Committee (CCC) whose purpose is to advise the UK and devolved governments on reducing greenhouse gas emissions, and adapting to climate change impacts.


To understand how much heat pump retrofit activity may be needed, it is useful to examine the current dwelling count (and what portion is still expected to be in use in 2050) and what fraction of those dwellings are currently heated by a system that is a candidate to be replaced with a heat pump.

For the purposes of this snapshot, data close to 2023 is generally acceptable, particularly from ~2019 onwards, as values are not changing very quickly. But there is not always the luxury of such recency.

("Retrofit" here includes the replacement of an existing heat source/system with a lower-carbon alternative, such as a gas combi with an air-source heat pump aka ASHP, as well as supporting measures such as improved insulation, if any.)

UK Dwelling Count and Profile

See Table 1 for a basic UK dwelling count, with breakdown into some smaller areas. These are as of the end of March 2020. English stock composed ~83% of the total (and rose ~0.9% the following year).

UK dwelling count at 2020-03-31
Constituent Parts
Great Britain28,740,000
Northern Ireland808,000
Table 1: count of dwellings in the UK (29.5M) and by nation/region as at 2020-03-31. (2021 data is also present for England and NI; eg England increased to 24,873,000 as of 2021-03-31, ~0.9% year-on-year.) Source: UK ONS (copy of data spreadsheet as of ) [ONShousing] [ONSdwelling]. Cf Valuation Office Agency (IndicatorID I21385) number of all properties (subject to Council Tax) in Kingston-upon-Thames local authority area at 2021: 68,550 (accessed ) [RBKdata].

Table 2 shows a basic profile of the English housing stock, representing over 80% of the entire UK stock.

(a) English 2020 dwelling stock profile: age
dwelling age range000s
post 19904,946

(b) English 2020 dwelling stock profile: archetype
dwelling type000s
all terrace6,417
semi-detached 5,810
converted flat1,028
purpose built flat, low rise3,764
purpose built flat, high rise625

(c) English 2020 dwelling stock profile: usable floor area m^2 (mean 96m^2)
m^2 range000s
less than 502,340
110 or more6,111
Table 2: elements from the 2020 English stock profile; age, archetype, usable floor area for 23,534,089 dwellings (11,152 samples). Source: table DA1101 (SST1.1), stock profile, 2020, English Housing Survey [EHSprofile].

While all new housing should be built not to require retrofit to achieve the UK's Net Zero goals, eg [BEIS2022building], in practice that is only a small fraction of the housing stock for 2050. The vast majority is already standing and needs to be retrofitted to reduce/eliminate carbon emissions; 80% according to [miller2020scaling].

For the EU as a whole the dwelling demolition rate is ~0.1% per year [fabbri2023insulation], or ~3.6% to 2050, with a renovation/retrofit rate of 1--2% per year (~30--60%). If the UK were to match the EU demolition rate, and given at most ~1% of UK homes with heat-pump space-heating now [henretty2020efficiency], that implies that ~95% of existing UK stock will still be standing in 2050 and potentially needs a heating-system retrofit of some sort to meet 2050 net zero goals. That would be ~28M homes or ~1M per year.

The English demolition rate in 2021/2022 was in fact 5,680 out of 24,873,320, ie 0.02%, or if extrapolated, only ~0.7% by 2050 [govUK2022dwellings].

England Dwellings Heating Fuel Type

In Figure 1 it can be seen that main heating fuel type or method for England is estimated for 2020 to be mains gas 79%, electricity 11%, community scheme %4, oil 3%, other 1%, heat-pump <1% [henretty2020efficiency].

Almost all those 79% mains gas systems will be wet (hydronic / radiator) based, but other schemes exist such as forced / warm air. Some sources claim a significantly higher share of gas wet heating, eg 86% according to [boilerguide2022central].

2020 EandW home heat source from homes with EPC source MHCLG
Figure 1: for England and for Wales, it is estimated that as of 2020, nearly eight out of every 10 dwellings with an EPC use mains gas to power central heating. Estimated distribution of dwellings with an EPC by main fuel type (or method of heating) used in central heating, as at financial year ending 2019, England, English regions and Wales. For England: mains gas 79%, electricity 11%, community scheme %4, oil 3%, other 1%, heat-pump <1%. London: higher on community schemes (13%); almost no oil or "other". London is the only region in England where the majority of new dwellings do not use mains gas directly in central heating; EPC data show 67% of new dwellings in London received heat from a community heating scheme in FYE 2019. Heat pumps, a low-carbon heating source, are not widely used for central heating in England and Wales, with 3% of new dwellings in both countries using them for central heating in financial year ending 2019. 18,171,912 EPC records covered dwellings in England and Wales; 17,544,671 (96.6%) records used after filtering. See copy of data download spreadsheet as of . Source: data and chart from section 5 [henretty2020efficiency]. (Contains public sector information licensed under the Open Government Licence v3.0.)

Dwelling Stock Turnover

As noted above, UK turnover of housing stock is fairly slow compared to the neighbouring EU. Thus UK housing stock is relatively old and inefficient within Europe. As of 2020 36% of the UK's 29.3M dwellings were built before World War II [piddington2020housing]. See also Table 2.

For the EU as a whole the demolition rate is ~0.1% per year or ~3.6% to 2050 [fabbri2023insulation], and the renovation/retrofit rate is 1--2% per year (~30--60%). If the UK were to match the EU demolition rate, and given at most ~1% of UK homes with heat-pump space-heating now [henretty2020efficiency], that would imply that ~95% of existing UK stock will still be standing in 2050, ie ~28M current homes will still be in use in 2050.

The English demolition rate in 2021/2022 was in fact 5,680 out of 24,873,320, ie 0.02%, or if extrapolated, only ~0.7% by 2050 [govUK2022dwellings], ie ~29M current homes will still be in use in 2050.

80% of 2050's UK homes are already standing according to [miller2020scaling], ie ~24M current homes will still be in use in 2050.

These extant homes and their implied required space-heat retrofit is the focus of this paper.

Technically Hard To Treat

Several studies and projects have shown that technical blockers to heat-pump installation are rare.

Will Rivers of the Carbon Trust notes from its work on Heat pump retrofit in London [rivers2020retrofit] that heat pumps can work in all building types and heat networks [rivers2020decarbonising].

Only 12% of properties surveyed were considered technically unfeasible for heat-pumps in the Energy Systems Catapult's Electrification of Heat work [ESC2022EoH].

The main non-participant barriers were:

  • Practical - external or internal space constraints
  • Technical - heating capacity constraints
  • Economic - cost of upgrades required

Some of those barriers could be overcome, eg by discussion with local planning authorities and DNOs; not all were "laws of physics" barriers.

Part of the same work [ESC2021suitable] found that There is no property type or architectural era that is unsuitable for a heat pump. That is, a building should not be excluded from consideration for a heat pump purely on the basis of its archetype. More practical case-by-case issues at a given expenditure and effort limit seem to be those barriers in practice.

The Carbon Trust found in 2020 that for London the main barriers to installation include upfront costs, the poor energy efficiency of existing buildings, the lack of space for equipment, and the noise levels of air-source heat pumps in urban areas [rivers2020retrofit] [rivers2020decarbonising].

Conservation Areas, Listed Buildings

"Heritage" building and areas are intended to preserve historic and notable buildings, and built and natural environments.

As [fetzer2023conservation] observes, by inhibiting the decarbonisation that we must undertake, such schemes may contribute to the destruction of what they aim to protect, especially if applied without pragmatism. For example, the "efficiency gap" from conservation-area status inhibiting efficiency retrofits is 5 to 15% in England. In Kingston, for example, ~10% of homes are in conservation areas; across England millions of buildings are in such areas.

According to Historic England there are around 500,000 listed buildings on the NHLE (National Heritage List for England) of all types in England [HElisted], ie at most a small fraction of the housing stock.

Not Binary: Hybrids, ULTDH

There is no silver-bullet heating decarbonisation tech, no one-size-fits-all. This is in part because of the wide range of building types and modifications and usage patterns and available infrastructure such as electricity and gas mains and capacity thereof.

This means that a significant fraction of dwellings may deploy a mix of energy vectors and heating technologies, including heat pumps. For example, hybrid heat pumps with natural gas (or hydrogen) to cover peak demand times or limited infeed ratings, and when grid carbon intensity is high. Another case is heat pumps lifting the temperature from Ultra Low Temperature District Heat (ULTDH) schemes.

The author mixes electrical and heat storage at home to dynamically minimise carbon footprint from DHW (Domestic Hot Water).

Gas to Heat-pump Candidate Count

The key premise of this paper, and of the CCC, is that essentially all UK home space heating has to be decarbonised by 2050 if the UK is to meet its Net Zero targets.

The calculations above indicate that between 24M and 29M of the current 29.5M UK dwellings will still be in use in 2050.

Extrapolating from [henretty2020efficiency] that ~80% of UK dwellings use mains gas heating, suggests that between 19M and 23M such homes are standing now and still will be in 2050.

(Other candidates amongst this retained stock are off-grid LPG and oil wet-heating systems, ~3% from [henretty2020efficiency].)

Various harder-to-treat cases (12% technically unfeasible [ESC2022EoH]), listed buildings (~500,000 in England [HElisted]), and conservation areas (~2M+ properties in England [fetzer2023conservation]), which will overlap, appear to encompass a limited fraction of the cohort. Those barriers are not absolute, but are based on current regulations and cost/effort acceptability levels, some of which may move as climate impacts become more evident for example.

(In dense urban areas for example, district heat may eat into in-home air-to-water heat-pump deployment, though in turn some district heat systems will use heat pumps in each home. Many of the same factors apply to making these low-temperature home systems work effectively, so are complementary and in need of very similar skills, controls and user education.)

Assuming a simple independent and uniform distribution of the above factors, final heat pump candidate numbers are computed from the estimated (range of) current dwelling stock in use in 2050 * 0.8 (currently mains gas) * 0.88 (excluding the 12% considered "technically unfeasible").

A range of 17M to 20M current domestic (mainly mains gas) wet-heat systems are likely candidates to be retrofitted with heat pumps.


The author lives with his family in Kingston-upon-Thames, a London borough at the south-west edge of London, and is involved in various green groups and their work with the council. Kingston is a famously "leafy" and wealthy borough, though also quite small in terms of population and dwellings relative to other London boroughs (68,550 properties subject to Council Tax in 2021, and 168,063 people as of the 2021-03-21 census date [RBKdata]).

The Royal Borough of Kingston (RBK) declared a climate emergency in June 2019, and aims through its Climate Action Plan to be carbon neutral in its own operations by 2030, and get the whole borough there by 2038. [RBK2022climate].

RBK's 2023 draft local plan includes district heating: RBK commissioned an energy masterplan in 2019 [RBK2019energy].

In private communication with RBK it seems that cost may be the main barrier to heat-pump uptake. Conservation areas and planning constraints could also add difficulty.

The RBK housing stock distribution seems closer to England as a whole than the wider London area, though like London has a much greater proportion of flats/maisonettes (including purpose-built flats) and a much smaller proportion of detached houses. Both may on the face of it be helpful in terms of energy demand and efficiency when it comes to upgrading to heat pumps [bouckaert2021net].

dwelling count by archetype Kingston London England 2021 RBK
Figure K1: 2021 VOA-derived RBK dwelling count percentages by archetype. In this sample RBK may have fewer of the hardest-to-heat detached homes than England as a whole, with noticeably more flats/maisonettes for example. Source: chart and CSV extracted 2023-02-09 [RBKdata].

RBK has a lot of pre-1950s solid-wall (brick) housing with uninsulated walls, thus with a high demand for space heat. (Gas wet heating systems seem to be close to typical, ie ~80%, suggesting ~48,000 gas boilers to replace with heat pumps in RBK, if the additional 12% "technically unfeasible" exclusion is applied.) ~10% of RBK homes are in conservation areas vs nearly double that for London.

For example, in the VOA CTSOP4.0 dataset for 2022, the build-period age profile for Kingston excerpted (with header row) indicates that ~14% was built before 1900, ~62% was built before 1955, and indeed ~29% was built between 1930 and 1939 inclusive (ie inter-war) [govUK2022CTdataset].

The combination of those space-heat-hungry large solid-wall houses in conservation areas (or listed) with the ~10% retrofit penalty [fetzer2023conservation] may prove to be a particularly hard nut for RBK to crack to reach Net Zero.

Climate Change Committee View

The CCC's 2019 Net Zero technical report states that the UK should install an estimated 19m heat pumps in homes ("Further Ambition"), alongside 5 million homes connected to low-carbon district heat, by 2050 to meet its emissions targets [CCC2019technical].

The CCC insists in 2019 that this retrofit of 29M existing UK homes must be treated as a national infrastructure priority [CCC2019technical]. In 2020 the CCC insists again that for heating buildings electrification is of primary strategic importance for Net Zero [CCC2020buildings].

In a comprehensive study specifically of emissions from space heating and hot water in existing UK dwellings, as is the subject of this snapshot, one of the key recommendations calls out the "Balanced Pathway" with 21M home heat-pumps (of which 5M are in a hybrid configuration) [CCC2020residential]. (This scenario also has 5.5M homes with district heat, and 2M with direct electric heat.)


Most current UK gas boilers (especially outside dense urban areas such as London with district heat) can and should be replaced with heat pumps before 2050.

There do not seem to be significant blockers to those replacements, though some will be more awkward and/or expensive due to space or conservation or other issues. Kingston will experience some of those issues more than most, but is probably wealthy enough to compensate somewhat.

The simple estimate of ~17M to ~20M gas boiler to heat-pump conversions required based on VOA and ONS data aligns with the CCC latest ~21M figure.


Datasets investigated during preparation of this snapshot include: [BEIS2016BEES] [EHSdataset] [EHSprofile] [EUeurostat] [govUK2022CTdataset] [govUK2022dwellings] [heatpumpMonitor] [henretty2020efficiency] [NGMdataset] [LAHSdataset] [NEEDdataset] [NHMdataset] [NIHSreport] [ONS2021census] [ONS2021heated] [ONSdwelling] [ONShousing] [RBKdata] [SHCScollection]

Key generated and cached third-party supporting data and images have been captured and are available.


Thanks for comments and suggestions from Peter M, Russell G, and my PhD supervisors, amongst others.

Notes on licensing for supporting data and images are available. Many thanks to those making data readily available.



A recent DESNZ policy document [govUK2023investment] (2023-04) notes that 90% of British homes have sufficient insulation and electrical capacity for a heat pump.


(Count: 50)