The scores are in!
2 apr 2024
After a year living with my heat pump I’m pleasantly surprised by the economics of ownership
I’ve updated this blog after seeing a recent report from the National Audit Office, which suggests that average heat pump installation costs in the UK have fallen from £15,000 to around £13,500
Anyone who’s read my previous heat-pump blogs will be aware of the trials and the tribulations (some of them, admittedly, self-inflicted) of my installation ‘journey’. But how’s it performing? I’ve now got a year of fully comparable data on my heat pump versus gas boiler, so am able to give faithful readers the scores on the doors.
First a few headlines:
My home (and hot water) have been beautifully warm all year.
The system works perfectly all year round with zero adjustment or user intervention.
My electricity bill for heating and hot water in the first year was 20% lower than my comparable gas bill.
I’m saving around 5 Tonnes of carbon per annum.
Overall, having lived with a heat pump for a year I’m more convinced than I was about their long-term viability as a mass-market replacement for gas boilers. But we need to get the policy signals right and we need some market innovation in both installations and financing.
Living with a heat pump
I have a 1930s detached house in which half the external walls are the original solid brick construction. It is now double-glazed throughout and has modern-standard roof insulation following a loft conversion some years ago. Downstairs we have mainly suspended wood floorboards with an empty cavity underneath, which until very recently had no underfloor insulation. So the house is ok, but not brilliantly insulated and probably not massively unrepresentative of the efficiency of the UK’s housing stock.
For those who understand these things, the calculated heat loss is around 12kW leading to an estimated annual heat demand of 30,000kWh (kilowatt hours) for heat and hot water at a design temperature of 21C downstairs, 18C upstairs. This turns out to be a pretty accurate estimate, although our combined heating and hot water demand is actually slightly lower at about 25,000kWh, because we keep the house at 19C throughout.
To operate at maximum efficiency you want to set up your heat pump to be drip feeding heat into your house at precisely the rate it is losing it any point in time. This ensures that your heat pump runs at the lowest possible water temperature and minimises the number of times it turns off and on again, both of which improve efficiency. The slightly counterintuitive fact is that to do this, the heat pump doesn’t need to know what temperature it actually is in your house, it just needs to know what temperature you want it to be. Then, once you’ve set it up correctly, and balanced your radiator circuit, the heat pump responds only to the outside temperature, delivering hotter water to your radiators as the temperature outside drops, and vice versa.
Slightly to my surprise, this really works. We target a temperature of 19C in the daytime and 17C at night time. During the day, every room in our house is precisely at the target temperature +/-0.5C (although of course sometimes the temperature overshoots if we’ve had the oven on for a long period of time or have a dozen people round for drinks). The initial set up does need a little bit of work and experimentation to calibrate the system. A good heat pump engineer should do this; the one criticism I have of my installer is I needed to learn and do too much of this myself. But once set up, the system never needs adjustment and delivers a constantly comfortable home with zero input from the homeowner. Hot water is similarly simple. For efficiency reasons we heat our hot water to a perfectly adequate 48C and then do a weekly anti-legionella cycle which heats the water up to around 70C. Again, the heat pump, once configured, does all this automatically with no input required from the user.
The indoor temperature has remained locked onto the 19C target in conditions that have gone as low as -4C outside. So the idea that heat pumps don’t work in British houses or British weather is just nonsense. Low temperature heating is also extremely pleasant and comfortable, with no hot spots or cold spots, just a consistent Mediterranean climate throughout!
Oh, and they’re extremely quiet as well and cause no disturbance to the neighbours (we run ours on low power mode overnight, which restricts the compressor speed to 40% of maximum).
How efficient is it?
We now need to take a brief technical detour into heat pump efficiency. Heat pumps are typically around 300% efficient. No, this isn’t a violation of the first law of thermodynamics, but instead a function of how heat pumps operate. Heat pumps don’t create heat, they move it around. The genius is that instead of using electricity to produce heat directly through resistance like an emersion heater or electric bar fire, a heat pump instead uses electricity to gather heat energy from the air outside the house and pass it, via heated water, into the house. Like a fridge in reverse. It turns out to be much more efficient to move heat around than to create it, so a typical heat pump can put three units of heat energy into your house for every one unit of electrical energy consumed in gathering that heat. This contrasts with a gas boiler, which uses 1 unit of energy stored in the gas to put around 0.8 to 0.9 units of heat energy into your house when it’s burned.
This efficiency metric of a heat pump is catchily called its ‘Coefficient of Performance’ (COP). For heating and hot water combined, my heat pump efficiency across the year has been 3.4 (i.e. 340% efficiency). This isn’t too shabby, and ranks in the top 20% of UK system efficiencies according to the Electrification of Heat Demonstration Project. I’m reasonably pleased with this as there were reasons to think our installation may not be amongst the most efficient: the house has a relatively high heat demand; because of my planning travails we had to have a hybrid heat pump and gas boiler system, which required use of a buffer tank which can be a drag on efficiency; and finally our heat pump is located in an alley down the side of our house, which is closed in on two sides by house walls and a third by a fence, creating the risk of cold air collecting around the heat pump and getting progressively recycled. In truth, I wasn’t expecting such a good result.
It turns out that this statistic - the COP - is the most critical number when determining the viability of heat pumps, as we’ll come to later.
How much carbon are we saving?
Our main motivation for installing a heat pump was to reduce our carbon emissions while at the same time providing a demand signal to help develop the UK heat pump market. So what is the carbon saving?
Because so much less electrical than gas energy is needed to deliver the same heating outcome, this creates the potential for massive carbon savings, provided the electrical energy is not produced in a very carbon-intensive manner. To deliver a combined heat and hot water demand of 25,000kWh requires around 31,000kWh of gas to be burned assuming a gas boiler efficiency of around 80% (although theoretical condensing boiler efficiency is 90% most such boilers - including ours - operate at too high a flow temperature to condense, meaning they operate at lower efficiency). This gas usage is rather distressingly over 2.5x the UK average of 11,500 kWh for a three bedroomed house occupied by 2.4 people. Our higher usage arises from the size of our house, its age and construction, plus the number of people (4 to 5) living in it, and the amount of hot water used by young adults! Each kWh of gas burned produces 183g of CO2e (CO2 equivalent greenhouse gasses), meaning that we were responsible for around 5.7 Tonnes CO2e per annum through heating and hot water.
What about with our heat pump? Overall, on a like basis during 2023 our heat pump, like our gas boiler, delivered just under 25,000kWh of heat energy to our house and hot water. The heat pump consumed just under 7,300kWh of electricity in the process (based on our COP of 3.4). The mean UK temperature in 2023 was almost identical to 2022 and so the heat demand to maintain a similar temperature should be broadly the same so I wouldn’t expect major weather-based distortions between the two years.
The carbon emissions from generating 7,300kWh of electricity really depends on the fuel used to generate it, ranging from close to zero for renewables or nuclear, to just under 400g per kWh for gas, to nearly 1,000g per kWh for Coal. This equates to carbon emissions of zero, 2.9 Tonnes, or 6.8 Tonnes respectively. The figure for coal powered electricity is striking: heat pumps are a third more polluting than gas boilers if electricity is generated by coal - fortunately largely a thing of the past in the UK. However, also significant is the fact that even if every marginal unit of electricity to power a heat pump is generated by a gas-fired power station, there is still a saving in carbon emissions of nearly 50% compared with a gas boiler.
Which of these figures should we take? Fortunately, the Government publishes emissions factors based on current and projected grid carbon intensity, and recommends a grid average of 146g CO2e per kWh for 2023 for footprint purposes, leading to emissions of just over 1 Tonne or a reduction of 4.6 Tonnes per annum, around 80% lower than 2022. Note that given the planned decarbonisation of the grid, the Government’s projected values for this emissions intensity factor is below 50g CO2e per kWh by 2030, equating to emissions from our heat pump of about one third of a Tonne per annum: to all intents and purposes carbon free, and a reduction of over 5 Tonnes per annum.
Over the period up to and including 2050, installing a heat pump therefore saves around 150 Tonnes of CO2e, which is equivalent to about 17 UK person-years of (consumption) emissions at current rates.
How much lower are our bills?
Our total bill for heating and hot water was, in 2023, a remarkable 20% lower than in 2022 on a like for like basis (adjusting 2022 monthly consumption to 2023 monthly prices for a fair comparison, given that energy prices spiked in 2022). How did this come about? There are two components to it.
First, on average across 2023, on my tariff, a kWh of electricity cost 3.81x as much as a kWh of gas. Taking into account the fact my boiler was operating at 80% efficiency and my heat pump at a COP of 3.4, this results in a 10% saving from this relative efficiency and cost as shown below:
Second, the heat pump has enabled us to make much better use of our self-generated solar energy. Previously we consumed about 50% of the energy we generated. This has increased to around 75%, as the heat pump provides a consistent level of electricity demand, especially on sunny spring and autumn days when we used to generate excess solar energy. This amounts to around 700kWh, or 10% reduction in heating costs, which added to the pricing saving above amounts to 20%, or about £500 at my average electricity price of around 35p across the year .
Does it all stack up?
As mentioned above, our primary motivation for installing the heat pump was to save carbon. But we can’t save carbon at any cost. One of the principles of our middle class approach to decarbonisation is to act as if there were a £100 carbon tax today, reflecting the ball park cost of carbon required to put us on track to net zero in 2050 with a 1.5C temperature limit. But of course this carbon price has to rise through the century. NGFS scenarios suggest 2050 carbon prices of £225 to £750 (current prices) in order to achieve temperatures “well below” 2C. Taking a conservative 2050 carbon price of £300 (current prices) leads to a real terms increase of 4% pa from £100 in 2023, the first year of our heat pump.
How should we discount carbon savings? This is a controversial area. Some economists propose a low or no real discount rate, reflecting the ethical nature of climate change intergenerational trade-offs. Others suggest higher market-related rates. This is not the place to have this debate, so I’ll use a real discount rate of 2% per annum, which turns out to be a central view of surveyed experts.
So we have a projected carbon saving increasing from around 4.7 Tonnes in 2023 to 5.7 Tonnes by 2050, which has a net present value on the above assumptions of £20,000. This is quite a striking figure in the context of the debate on the “expense” of heat pumps. On reasonable assumptions about carbon prices, purely from a carbon perspective out to 2050, it is worth society paying £20,000 to install my heat pump, even if there is no saving in ongoing running cost.
In fact, we made savings, of around £500 in 2023, although this is declining as the relative pricing of electricity versus gas worsens.
My heat pump install was expensive. Installing a fully-flexible hybrid added components and design costs and meant I didn’t get the Government grant. I changed my mind half way through the project about what I wanted, requiring work to be undone and redone a different way. COVID-induced supply chain blockages led to spikes in some component prices. But it still cost less than £25,000. I had thought it was infeasibly expensive, but am encouraged to see that, on my assessment framework, and account for the ongoing savings I am making on bills, it is actually worth it for the carbon saved.
The financial case
The fact my five figure heat pump installation makes sense for society on a the basis of the net present value of carbon savings makes me very happy. But it’s not the basis for a mass market roll-out. What counts is whether it will make financial sense for consumers.
For this section I’m going to move away from my own case to a typical heat pump installation in an average home that, according to Ofgem, uses around 11,500 kWh of gas. I’ll assume that their heat pump installation costs £13,500, being the average cost of installations in 2023 (at current prices) identified from MCS data in the recent National Audit Office report on decarbonising home heating. It’s possible that installation costs will fall as competition and scale grow in the market, but on the other hand heat pumps are a mature technology and much of the price is in any case in the installation costs rather than the hardware. So I’ll be conservative and assume the £13,500 average cost remains reasonable. I’ll further assume that the heat pump achieves a COP of 3, in line with the average achieved in the decarbonisation of heat demonstration project. I’ll assume our consumer is paying April 2024 average energy price cap rates of 24.50p per kWh for electricity and 6.04p for gas. I’ll assume that the heat pump installation allows them to come off gas completely (so no gas cooking), saving the annual standing charge of £115 a year and that they qualify for the full Boiler Upgrade Scheme grant of £7,500. Finally, I’ll assume that they have succeeded in getting their existing gas boiler to condense, so it operates at 90% efficiency, but they now need to replace it, at a cost of £3,000 (including VAT).
Comparison of upfront costs and ongoing bills
The immediate cost of installing the heat pump is therefore £6,000 (being £13,500 less the Government grant of £7,500), which is £3,000 more than the £3,000 cost of installing a new gas boiler.
What about the impact on ongoing bills? Based on our core assumptions we can calculate the comparison of annual bills for heating and hot water as follows:
In other words, for the benefit of a £3,000 additional capital cost today, the heat pump owner has the privilege of receiving annual bills that are £36 a year higher! On the face of it not a great deal. This is slightly hard on the heat pump as most gas boilers are not operated at low enough flow temperatures to condense. If we had assumed 80% efficiency for the current gas boiler, this would have resulted in a £60 saving for the heat pump, keeping all other assumptions the same. But either way, the costs before and after the switch are broadly the same.
What if heat pumps are inevitable?
This simplistic comparison is unfair to heat pumps, if we assume that the decision to switch from a boiler to a heat pump is just a question of timing. Given a typical heat pump or boiler lifetime of, say, 15 years, let us imagine that the choice is really between getting a heat pump now or getting one in fifteen years’ time. A big part of the cost of installing the first heat pump is the costs of switching over to a different system and potentially installing some new radiators or pipework to facilitate this. Once a heat pump system is operating in a home, then if a new one is required in 15 years’ time it is a simple matter of just switching the old one out for a new one. It is probably still more expensive than a gas boiler, as the hardware cost of a heat pump is around £2,000 more than a gas boiler, although for a switch-over the installation costs should be similar. By contrast, the gas boiler owner will still need to install a heat pump from scratch in 15 years’ time.
So really the comparison is to compare the £6,000 immediate (net of subsidy) cost of a heat pump plus paying, say, £5,000 to install a replacement heat pump in 15 years’ time; with the £3,000 immediate cost of a boiler plus paying £13,500 for a heat pump to replace the gas boiler in 15 years’ time less any Government grant available at the time.
Assuming a 2% per annum real discount rate (which is a reasonable after-tax financial discount rate for a n individual), results in the following net present value of the options:
So for someone who believes the Government grant is a temporary inducement (which is a reasonable assumption given the precedents of incentives to install solar and buy electric vehicles) then even with no bill savings the heat pump installation is cheaper by over £3,000 on a net present value basis than deferring the switch for fifteen years. However, if you believe that the Government incentives will persist, then you may believe it is cheaper to wait. Although that is being optimistic, as the £7,500 grant multiplied across 26.4m dwellings in the UK would cost the exchequer a cool £200bn!
This analysis shows how heat pump adoption could be helped by clarity of policy (a) that the switch is going to happen and over what timescale and (b) that the government subsidy of £7,500 for every heat pump installation will have to be discontinued or reduced at some point.
However, while helpful, Government policy clarity is unlikely to be enough. First, the absolute net additional upfront capital cost of £3,000 involved in installing a heat pump is beyond the means of many, For all but the most well off in society, net present value analysis over 15 years is meaningless. Second, some form of future subsidy will likely remain in place, at least for the less well off. Third, the saving outlined, because it only materialises over a 15 year horizon, does not lend itself to market-based financing solutions.
Given the significant uncertainties in play, many people will consider the bird in the hand of saving £3,000 today (while avoiding the disruption of the switch) to be the winning factor. The reality is that heat pump adoption will only take off when running costs are reliably lower than current gas boiler systems.
Tipping the scales from boilers to heat pumps
The analysis above showed that on current average parameters, the ongoing bills from a heat pump and gas boiler are pretty comparable. But this need not be so. There are two crucial levers that can be pulled to tip the scales in favour of heat pumps:
Improving heat pump efficiency; and
Changing the relative price of gas and electricity
In my calculations I used an average heat pump efficiency of 3, being around the median in the latest Electrification of Heat Demonstration Project. However, even my installation, with the inherent efficiency impediments described earlier, has achieved 3.4 across a full year. Organisations like Heat Geek are now claiming COPs of over 4 and are offering guarantees on performance. Through a combination of improving heat pump technology (especially refrigerants) and, in particular, improved consistency of installation plus provider guarantees, COPs of 3.5 or even 4 could become normal.
Turning to the relative price of gas and electricity, for a brief period from October to June 2023, a kWh of electricity cost 3.4x or less than the cost of a kWh of gas. This followed the temporary removal of policy costs from electricity prices, which perversely penalise electricity in the transition. . Unfortunately, as gas prices have dropped, the Government has missed a strategic opportunity to maintain this relatively favourable ratio to electricity. Under the latest electricity price cap announcement for April 2024, the ratio stands at 4x. It’s worth noting that this is amongst the highest ratios in Europe, and well above a mean European ratio of 3.1x.
The table below recasts the average ongoing annual bill saving for different levels of heat pump COP and electricity to gas price ratio.
The top row and second column shows the £36 additional cost calculated on our core current assumptions. Moving left we can see where the nightmare stories of heat pump running costs come from. The Electrification of Heat project found that around one in five installations had a COP of 2.5 or below, leading to additional bills for our average customer of just over £200 a year, or more than 25% higher. There is really no excuse today for a heat pump installation with a COP of below 3, so driving out poor quality installations is important.
But moving in the other direction, we can see that increasing COP to 3.5 and reducing the electricity to gas price ratio to the same figure creates a saving of £184 per year, which is starting to become significant. Moreover, at this level and above we are starting to be within the range of innovative financing propositions, funded out of these savings.
These figures are not pie in the sky. It is conceivable to envisage installs with a guaranteed COP of close to 4 and, in conjunction Government support, the relative price paid for electricity getting towards 3x the cost of gas. At that point we are in business, with heat pumps looking an attractive proposition on purely financial terms, saving consumers over £300 (or around 40%) on their bills. Bundled installations with solar, batteries, and agile tariffs could drive even larger savings. At this level of saving it becomes easier to see how a £3,000 difference in upfront capital costs can be bridged.
Where next for heat pumps?
The UK Government is already consulting on low carbon standards for new build homes from 2025. This will almost certainly lead to lower running costs for homeowners, given that it should be no problem to install a heat pump operating at a COP of at least 3.5x in a new build home. Moreover, by designing the heat pump into the build of the house, the additional capital costs compared with gas boilers are minimised. So for new build houses, heat pumps are already a no-brainer.
However, there is no denying that the challenge remains much greater for retrofit applications. Here, heat pumps remain the preserve of the well-off climate-conscious consumer, prepared to deploy significant upfront capital in order to lower their carbon footprint. The good news is that, even for retrofits, the numbers seem to stack up from a carbon-cost perspective, based on an initial carbon price of £100 per Tonne, tripling to £300 per Tonne by 2050. Followers of a middle-class approach to decarbonisation should press ahead.
But the finances remain problematic. With good installation quality, running costs that are no higher than gas should be possible, or significantly lower if the Government follows through on its intention to rebalance electricity and gas costs. Combining with solar and batteries provides additional savings. On the assumption that heat pumps will become mandatory in future, and potentially with a less generous subsidy, the finances already stack up. But for the mass market, too much upfront capital is required.
The obvious answer is to attract private capital to bridge the gap. We are starting to see some innovative services emerge, looking at heating as a service models where the capital cost of the installation plus ongoing savings are bundled into a flat monthly fee combined with a performance guarantee (you can listen to this episode of the BetaTalk podcast with Nathan Gambling for one such example). But for any of these models really to drive demand, we need to be creating ongoing bills savings from moving to heat pumps.
So overall, we need four things:
Greater Government commitment to, and clarity on, the transition from gas boilers, with fewer pointless dalliances with, at best, niche solutions like hydrogen boilers, which just create market confusion.
Active policy to close the gap between electricity and gas pricing - even a 3.5x ratio would help a lot, in line with the ratio from October 2022 to June 2023. Better still a move to the European average of 3.1x.
Installer and supplier commitment to industry initiatives such as Heat Geek’s COP guarantee scheme, in order both to drive out poor implementations that end up being excessively costly, but also to push the industry towards consistently delivering COPs of 3.5 to 4, which coupled with even small changes in relative electricity and gas pricing could start saving homeowners hundreds of pounds a year on their bills.
Continued innovation from the private sector to bundle installations and financing, so as to make heat pumps affordable on a monthly basis for normal consumers.
Heat pumps already provide value for money for society in terms of the value of the carbon savings. But at the moment they are a tough sell financially for the retro-fitting home-owner for whom climate change is not a key priority. But that could change very quickly with some very achievable policy signals and market developments.
For those of us with the resources to do so, there’s an important role to play in helping this market get off the ground in the UK by installing heat pumps now. If you’re interested, I’d encourage you to roll your sleeves up and get stuck in!
The Transition Finance Market Review is an impressive piece of work. But I still don’t understand what a “transition finance” label will achieve.