Report: Mind The Gap

The costs of moving from boilers to heat pumps


Centre for Net Zero is building an agent-based model (ABM) for the energy transition to identify faster, fairer and more affordable paths to net zero. The first release of our simulation will inform a consolidated research paper to be published in March 2022. Ahead of this paper’s release, we are publishing a series of interim research findings to share our learnings as we leverage our access to Octopus Energy data, and other public datasets and research, and develop the simulation in full.

These research findings provide a unique insight into current domestic energy behaviours and help improve understandings of expected future behaviours. In creating our ABM, we want to discover how a diversity of actors within the energy landscape make autonomous decisions that affect one another and influence the behaviour of the overall system. These insights will enable decision makers to take faster, bolder and more equitable climate action.

Key research findings

1. Air source heat pumps (ASHPs) are expected to be one of the dominant technologies to replace gas boilers, but their higher up-front cost is a commonly cited barrier to adoption.

2. While variety in Great Britain’s (GB) housing type, size and efficiency is broadly acknowledged, debate often centres on the ‘average’ household — comparing a £10,000 investment in a heat pump system to £2,000 for a replacement boiler.

3. We looked across 30m GB properties to simply estimate the distribution of costs to retrofit an air source heat pump.

4. We found that a £4,000 subsidy, as proposed under the Clean Heat Grant consulted on by government, could make air source heat pumps cheaper than gas boilers for 13% of GB homes. However, three-quarters of these homes are flats, for which there are currently practical challenges in adopting heat pumps. Excluding flats, the same subsidy will bring 3% of homes to cost parity.

5. Around 50,000 ASHPs have been deployed in GB to date. As with other energy technologies such as solar and wind, we would expect heat pump costs to fall as more are installed. In this way, a subsidy that affects 3% of homes in the first instance could deliver cost-parity for 16% of homes after two years. This could lead to uptake of around 230,000 heat pumps per year.

6. Subsidies could be reduced over time, but the government will need to balance fiscal responsibility against industry and household confidence, which will be essential in underpinning sustained innovation and demand growth. There may also be a role for the Government in enabling wider adoption within flats, through supporting technological innovation and review of regulation.

7. Up-front cost is just one factor in household decisions about heating systems. Our next step is to integrate this analysis into our agent-based model to consider a wider range of technologies and factors including running costs, technical suitability and awareness. Our model will consider both heat and transport, as two critical areas where we need fair and affordable options for households to reduce greenhouse gas emissions.

*Following the Government’s Heat & Buildings Strategy announcement in October 2021, we updated our modelling to reflect the subsidy confirmed in the £450m Boiler Upgrade Scheme. Please see the final part of this article for details.

Context — why focus on heat pumps?

Homes account for 21% of total greenhouse gas (GHG) emissions (1) and 35% of total energy consumption in the UK in 2020 (2). 75% of the total energy demand in the UK housing stock comes from heating (3).

Natural gas is the most common heating fuel in domestic buildings, covering 80% of domestic heat demand in the UK (3). There are currently about 22 million households in the UK with gas boilers (4), with 1.7 million units sold in 2019.

The UK government supports the deployment of heat pumps: in 2020, it pledged to install 600,000 heat pumps per year by 2028. Currently fewer than 30,000 heat pumps are installed yearly (5). One of the major reasons for this is the high up-front costs of heat pump systems relative to gas boilers.

In this analysis we have focused on the difference in up-front cost between ASHPs and gas boilers, ignoring other technologies and factors in decision-making. ASHPs have made up more than 60% of accreditations under the UK government’s Renewable Heat Incentive (RHI) scheme — and this trend is evident across property types, from detached homes to flats and maisonettes.

Figure 1: Percentage of low carbon heating technologies installed in different property types in the UK (6)

What drives the cost of a heat pump?

Heat pumps are sized according to forecast heating demand. Desired temperature, weather conditions, insulation and dwelling size all play a part. Proper sizing typically requires a site visit from a trained heating engineer.

For the purpose of this analysis, we use the following rule of thumb:

  • kW capacity = 0.1 * Floor Area/m2 (rounded down to the nearest integer)
  • Multiply capacity by 70% if property is ‘well insulated’ (EPC A-C). Note that for EPC D-G, we did not explicitly consider the trade-off between heat pump capacity and remedial energy efficiency measures
  • Limit capacity to 20kW. Cost data from the domestic RHI is limited for larger ASHPs (those between 20–45kW make up less than 2% of current accreditations) so we decided to cap at a level where the sample size is more robust.

Once the capacities of the heat pumps for each household have been estimated, the Ofgem RHI database (6) was used to estimate the range of up-front costs. While this is an imperfect estimation, we believe it is a significant improvement on reliance on average properties and heating systems.

For gas boilers we split homes into three different categories based on their floor area: small (including all flats); medium; and large. We then allocated boiler costs according to an article on the Boiler Guide website.

Variety in the GB housing stock

A central foundation of this analysis has been a commercial dataset derived from the Energy Performance Certificate (EPC) database — taking the data from EPCs and inferring equivalent values for properties in England, Scotland and Wales which do not have a certificate. An EPC gives a property an energy efficiency rating from A (most efficient) to G (least efficient). The distribution of ratings in Great Britain can be found in Figure 1. Along with these ratings, the dataset provides information about the number of bedrooms and the floor area of the properties.

As illustrated in Figure 2, the most common EPC in the GB is EPC D. Remember, an A rating is the most energy efficient, a G rating is the least.

Figure 2: EPC distribution of 30 million GB households

Figure 3 illustrates that, as expected, houses are normally bigger than flats and bungalows. However, there is a wide variety of floor sizes per category which we have taken into account in estimating the costs of ASHP installations.

Figure 3: Floor areas for different household types

In Figure 4 you can see that flats and maisonettes are more likely to be well insulated, with EPC C being the most common one, while houses and bungalows are mostly EPC D rated.

Figure 4: Distribution of EPC by household type

What does this mean for the distribution of costs per household?

On average, the poor insulation and larger floor space of houses will result in higher capital costs. On the other hand, flats and maisonettes are in a much better position, since their EPC ratings indicate that these types of households are better insulated, have less floor space and therefore would require smaller heat pumps. Our modelling of ASHP capacity per property type can be seen in Figure 5. In general, houses need larger heat pumps, while bungalows and flats require more modest devices.

Figure 5: Distribution of recommended ASHP kW Capacity by household type

Across all property types, the ASHP capacity required for households is generally between 5 and 10 kW (as seen in Figure 6). On the right hand side you see the effect of our decision to limit capacity to 20kW — for these 3.5% of households the modelled costs may be underestimated.

Figure 6: Distribution of ASHP capacities to decarbonise heating in the UK household stock

Using these capacities and the costs estimated using RHI data (6), we reviewed the estimated ASHP costs depending on the EPC rating and the number of rooms.

Figure 7: Costs to purchase a heat pump per EPC and number of rooms

Figure 7 shows that poorly insulated households will need to spend considerably more (between £12k and £15k) than households that are well insulated of the same size (less than £10k). Gas boiler up-front costs are much more consistent and do not vary as much depending on the capacity needed. In the UK, the average capital and installation cost of these boilers is around £2,000 (7).

A £4,000 subsidy would make an ASHP the cheaper option for 13% of homes

The Clean Heat Grant is expected to be announced shortly by the UK government, offering households £4,000 to reduce the up-front cost of low-carbon technologies like ASHPs.

Figure 8 shows the difference between ASHP costs and gas boilers across GB households. Looking at the turquoise line (which reflects median ASHP costs reported by the RHI), 13% of households have a pre-subsidy ASHP premium less than £4,000. If average ASHP costs are reduced to lower quartile levels, the proportion of properties for which an ASHP is cheaper post-subsidy rises to 40%, while it falls to 7% at upper quartile cost levels.

Figure 8: Proportion of households that can benefit from government subsidy to bring ASHP to cost parity with gas boilers. The vertical line represents the expected Clean Heat Grant of £4,000 (8)

As identified previously, flats will generally require smaller ASHPs. However there are practical challenges in installing them in flats, particularly around space and being more likely to be rented. To date, ASHPs have been installed in around 3,000 flats, an adoption rate five times lower than for houses and bungalows (9). Figure 9 below splits the turquoise line from Figure 8, to show the impact of a £4,000 subsidy on different types of home. Here we can see while 13% of all homes may benefit, the proportion is far higher for flats (>40%) than for houses and bungalows (2–7%).

Figure 9: Proportion of households (split by property type) that can benefit from government subsidy to bring ASHP to cost parity with gas boilers. The vertical line represents the proposed Clean Heat Grant of £4,000 (8)

If we cannot make it easy for flats to adopt heat pumps, either through innovation or changes in regulation, then we may want to consider them separately within our analysis. In the table below we can see how bringing a certain number of homes to cost parity costs significantly more if we assume flats cannot adopt heat pumps.

Subsidy should cause the cost of ASHPs to fall, as more are installed

Over the past two decades we have witnessed huge falls in the unit costs of solar and wind generation, widely documented as ‘learning curves’ which express how technologies get cheaper as they come to scale. The learning curves for solar and wind show a 20% reduction in costs for every doubling of cumulative production. Studies in Germany and Switzerland have shown even higher learning rates for domestic heat pumps of 30–35% (10), though these date from analysis up to 2004.

While we would expect global rather than GB penetration of heat pumps to shape the cost of the equipment itself, currently around 50% of cost relates to labour. We can expect this labour cost to reduce in line with GB adoption.

In this way, stimulating demand through a subsidy can ultimately drive down costs, making heat pumps affordable for more households and/or reducing the level of subsidy required on an ongoing basis.

Of course, even where heat pumps are the cheaper option, it is still not guaranteed that households will choose them. Awareness is low, inertia is high. If we assume that one in five households will need to replace their boiler over the two-year life of the Clean Heat Grant and that half of those at cost-parity would choose an ASHP, then a subsidy of £4,100 would achieve a deployment rate of 50,000 heat pumps per year, trebling the installed base within GB. If we apply a 30% learning rate to the heat pump’s labour costs then we would expect the overall cost of an ASHP to fall by 20%. The same subsidy that brought 3% of homes to cost-parity now achieves the same for 16%. On the same assumptions as above, that could mean deployment of 230,000 heat pumps per year.

Government must decide on the level & length of investment

The government wants to jump start the mass deployment of heat pumps in a way that will not burden households with exorbitant costs, whilst also demonstrating fiscal responsibility after months of Covid-19 related expenditure. The analysis above has shown that its proposed £4,000 subsidy could make air source heat pumps cheaper than gas boilers for a significant minority of properties and that this could lead to significant cost reductions in industry. However, subsidies are likely to be required for longer than the proposed two-year life of the Clean Heat Grant, and it is important for households and industry alike to have confidence as soon as possible in the plan for 2022–24 and beyond in order for us to see sustained innovation and demand growth for heat pumps.

In theory a larger grant could see more properties transition away from fossil fuels, reduce ASHP costs faster and accelerate to a time where subsidies are no longer required. Recent media reports suggest that the Government is considering a £7,000 grant (11). Given the extended period of waiting for the Heat and Buildings Strategy, it is unclear whether industry would be ready to deliver the volumes implied by a substantially increased grant both in terms of capacity and capability. At this point in time a smaller grant, while it would still support some households to transition to low carbon heat, risks undermining confidence in the government’s commitment to the heating transition.

Limitations of this analysis

Significantly, our analysis is based purely on the up-front cost of heat pumps. Our estimates of these costs depend only on the floor area of the household and the current or estimated EPC of the property.

This analysis does not consider other factors that may influence household heating choices, such as ongoing costs, public acceptance and awareness, space constraints, alternative heating technologies and property tenure.

Taking the example of property tenure, in Figure 11 we can see that while almost 80% of UK houses are owner-occupied, less than half of flats are. Tenanted properties are likely to be subject to different, and potentially more complex, decision-making.

Figure 11: Percentages of Tenure vs Property Type


The electric heat revolution should garner optimism: heat pumps are proven and available now. They provide a useful tool in the decarbonisation of heat and while they won’t be the only low-carbon technology for households, they are likely to represent a large and meaningful proportion of future installations (2).

However, in order to enable consumer uptake of ASHPs at scale, we need to address the up-front cost premium, through combinations of cost reductions and focused government subsidies. We need to improve our understanding of how factors other than capital cost combine with cost to impact adoption rates.

Our agent-based model will consider a diversity of low-carbon heating alternatives with agent behaviours dependent on a greater range of factors than cost alone such as income, tenure and social factors, as well as industry considerations. Understanding how these factors interact will enable policy makers to understand the relationship between different intervention strategies and adoption and the rates at which they can best be deployed.

Update: Heat & Buildings Strategy Announcement — October 2021

Following the Government’s Heat & Buildings Strategy announcement in October 2021, we updated our modelling to reflect the subsidy confirmed in the £450m Boiler Upgrade Scheme. Grants of £5,000 will be offered to households for a duration of three years, starting in April 2022, to encourage homeowners to install more efficient, low carbon heating systems like heat pumps. Capped at £450 million, the grant scheme will only be available to 90,000 households in total.

We inputted this level of grant and duration into our model and updated assumptions that we applied to our learning rates — applying a 32% rate in line with the literature, increased from 30%. We also updated a previous assumption where we assumed that 50% of households who had cost parity between ASHP and gas boilers would choose an ASHP. We now assume that households are on average 30% aware of heat pumps (from the literature), and it is 70% convenient to change from gas to electricity as the main source of heating fuel. This means that we are assuming 21% of households who have cost parity between ASHPs and gas boilers choose an ASHP.

We assume that households only change their heating system when their current one breaks. Assuming that boilers break every 10 years, we calculate that per year 10% of households will consider switching to an ASHP. In practice, uptake may be higher because people might choose to get an ASHP even if they have a working boiler. We assume that only households with cost-parity (after the £5k grant) will seek to choose between ASHP and gas boiler.

Note that many of these assumptions make the findings a conservative lower bound — uptake is likely to be higher than the results presented.


Our research shows that actual cumulative demand for heat pump installations over the three year duration of the grant could be closer to 560,000 households — over six times the number that would be covered by the £450m allocated to the new Boiler Upgrade Scheme (Y1: 80k; Y2: 100k; Y3: 380k).

Meeting that unmet demand with further grant funding would bring costs down for the majority more quickly by supporting learning effects (which express how technologies become cheaper as they come to scale) — taking society rapidly through the innovators and early adopters and into the majority. This is vital if we are to meet whole economy decarbonisation targets.


1. 2020 UK Provisional Greenhouse Gas Emissions. GOV.UK

2. National Grid. Future Energy Scenarios. (2021)

3. Watson, S. D., Lomas, K. J. & Buswell, R. A. Decarbonising domestic heating: What is the peak GB demand? Energy Policy 126, 533–544 (2019)

4. 2019 was a record year for gas boiler sales. Installer Online (2020)

5. Energy white paper: Powering our net zero future. GOV.UK

6. Public reports and data: Domestic RHI. Ofgem

7. Boiler Guide. What Size Boiler Do I Need?

8. Domestic Renewable Heat Incentive (RHI). GOV.UK

9. RHI Statistics June 2021, Table S2.4
RHI monthly deployment data: June 2021 (Quarterly edition)

10. Weiss et al, A review of experience curve analyses for energy demand technologies
A Review of Experience Curve for Energy Demand Technologies | Request PDF

11. The Times. Green Grants of £7,000 to help households replace gas boilers




An open research lab that uses data-driven modelling and simulation to provide the critical insights that policy makers need to take bold climate action.

Recommended from Medium

Concept: The Green Cocoon: A Pergola-like Trellis to Shade Hot Baking Roads with Lush Green Foliage

Keep your BS out of our Knowledge Pool

The City of Palo Alto’s Ongoing Efforts to Address Public Safety and Resiliency

Getting closer to your mega project problem

An Interview with Vanessa Nakate, Uganda’s first #FridaysForFuture Climate Striker

MythBusters: Solar Energy Edition

These Are Not The Masks You Are Looking For

It’s time for each of us to expand our horizons.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Centre for Net Zero

Centre for Net Zero

Open Research Lab, realising faster, fairer and more affordable paths to Net Zero. Powered by Octopus Energy

More from Medium

On Nests and Metaphors: Honoring Robert Bly

Ingroup Colossus: @visakanv

Thesis Research — Online Counseling & Therapy

Challenges In Implementing Sustainable Fashion