Heating a listed Cotswold stone building with an air-source heat pump: our journey

Update a: A survey by Nesta UK of Heat Pump users published 23 May 2023 finds high levels of satisfaction with heat pumps, see Note [7]

Update b: Following requests from some readers for more information on the data, an Appendix has been added to provide details of the data collection, analysis of the COP and SCOP achieved, and a comparison of running costs pre and post the heat pump installation.

Here is a plot spoiler: my wife and I are delighted with the results of living in our listed Cotswold stone home heated by an air-source heat pump since December 2021. I want to share our story as a corrective to the belief, widely expressed in the media, that it would be impossible to do what we did: self-evidently, this is untrue.

In case you want answers to the burning questions I often hear, I’ve collected a few (see Note [1]).

Here’s our story.

Beginnings

Twenty five years ago my wife and I bought an old Grade 2 Listed property with friends and split it in two.  It was a bargain; ignoring the subsequent years of work! Renovations of sash windows and shutters, valley gutter lead work, lime mortar of the end terrace, and so much more, followed over the years.

It turned out the party wall used to be an external wall when the property was first built in 1805, but about a decade later the mill-owner who acquired it wanted something much more substantial. He extended the small cottage outwards and upwards into what was advertised later as a “capital messuage”. Not a property I’d ever imagined owning, but somehow we found ourselves as custodians of this beautiful property.

When we arrived there was a higgledy piggledy array of partitioned rooms created for the care home it had been used for prior to its closure. We set about restoring it to its former Georgian glory. But there is always something to do on an old house, and the journey continues. That’s why it always amuses me when people talk about ‘retrofit’ as if it is some fast and easy project, because from my point of view, maintaining the fabric of a property is always a process not an event; a process that never really ends.

We couldn’t find any way to split the existing water, gas and electricity utilities in two, externally or internally. So we had to start again with new services, that were routed in at the back of the property, so as not to impact on the Georgian front elevation. Both households were essentially starting with a blank canvas and put in completely new plumbing and gas boilers. We called our plumber ‘Danny the ferret’ because of his ability to get into impossible spaces and never leave a mess. The small new gas boiler could now heat our now separate property in less than an hour; like a Ferrari, able to go from 0 to 60 in 5 seconds. 

Little did we know then, but Danny did such a great job on the pipework and radiators that when last year we had an air-source heat pump fitted, the plumbing through the house turned out to be fit for purpose (big enough pipes in the right places). No new copper piping required except to connect the new equipment, and only one third of the double-panel (with fins) radiators he fitted needed fattening a little to be replaced by triple-panel radiators.

And those that were upsized (like the one illustrated) were a bit fatter but the same height and width, so fitted into their positions without any need to change the pumbing. This in effect increases the surface area of the radiator, allowing for a lower flow temperature to still heat the room to its target temperature.

Exploring opinions on heat pumps

While I am no laggard, I’m also generally not an ‘early adopter’ of anything, even when I believe its the right thing to do. It took me a decade longer than most to switch from a film to a digital camera because I was unconvinced they were good enough. I tend to prefer for others to learn the lessons and pass them on. I like it also when there is an inevitable reduction in costs as a market matures a little. So I did get a digital camera finally, 20 years ago now, and it’s been great, but I was hardly an advocate for ‘new tech’.

My old friend Chris runs a successful business, Yorkshire Energy Systems (YES), installing solar and heat pump systems (amongst others), and had already been giving me an education on heat pumps.  I thought we should consider one when we had to replace our existing system, but again, I was in no rush, I needed time to explore the subject.

I invited Chris down to Nailsworth to give a talk to the local climate group I helped run (Nailsworth Climate Action Network), and also invited leaders of local political parties of all persuasions from the Stroud Valleys, including the then MP for Stroud, David Drew for Labour, and the prospective Conservative candidate, Siobhan Baillie (who in a subsequent election became our MP). 

The talk went down extremely well – Chris is a brilliant speaker.

Chris later gave another talk for us online during Covid. This also went down well and was recorded, then posted on our website: What if we could all heat our buildings with renewable technologies? It addresses most of the myths you hear about heat pumps, and a number of people have told me how it helped make up their mind, and go for a heat pump.

I have a good natural science background, so I was not in any doubt about the soundness of the 19th century physics that underpin the workings of a heat pump, which we all blithely rely on every day (most people have a fridge, which pumps heat from inside the fridge to its outside). But understanding the science is rarely enough, even for ex-scientists!

While in one ear Chris was telling me that any building that can be heated with a gas boiler can be heated with a heat pump, in the other ear, other friends, including experts in insulating old buildings, told me that ‘deep retrofit’ was an essential precursor to heating an old building with a heat pump, especially an air-source heat pump. Doubts crept in.

I also cannot have been totally immune to the drip drip effect of reports on the BBC and elsewhere. Heat pumps were first demonstrated in 1830 and the theory was in place in 1852 [Note 2], yet this is what Roger Harrabin said in a 2014 report on a water-source heat pump installed to heat the historic house of Plas Newydd in Wales (science facepalm warning):

It’s barely believable that this sea water has enough heat to warm anything, it’s pretty chilly at this time of year, but yet, thanks to an extraordinary technology called a heat exchanger, it’s the sea that’s going to heat this house.

But cold water contains enormous amounts of energy. Thermal energy is just the jostling of molecules. So the water Roger Harrabin was feeling, say at 7°C, is a sweltering 280 Kelvin (on the absolute temperature scale), with no shortage of jostling molecules from which energy can be extracted. Similarly for air. If you stand in front of an air-source heat pump the air blown out is colder than the surrounding air by a few degrees, because the heat pump has extracted thermal energy from it (using the same tech that a fridge uses to extract thermal energy from anything you put in the fridge). This is more than enough to heat a home.

Yet journalists and commentators have continued to assert that heat pumps can’t heat older buildings without substantial insulation work. Our national broadcaster (the BBC) is culpable, despite its duty to “inform, educate and entertain”. Most recently, during the 8th March 2023 episode of Jeremy Vine’s BBC Radio 2 programme, the host expressed dismay that a radiator was lukewarm. What hope is there when a key programme is failing to “inform and educate” in this way. A hint: it is the thermometer in the room that tells how warm the room is, not any preconceptions on how hot a radiator should be.

Roger Harrabin, Jeremy Vine and many others in the media do a great disservice by sharing their incredulity, and repeating unfounded beliefs about heat pumps.

Why did we decide to go for it?

For us there was a stark choice we faced that forced us to stop vacillating, and certainly stop listening to the naysayers:

• Firstly, the 25 year old gas boiler was creaking and behaving a bit oddly even after a service. There was no ‘ripping out’ as the tabloids and others emotively campaign against, merely the natural end of life that could no longer be serviced to stay in operation much longer.

• Secondly, I had seen the data from the Energy Saving Trust, and it was quite clear that getting off using gas to heat our home and replacing it with a heat pump would massively reduce our annual carbon footprint (as I wrote about in Are Air Source Heat Pumps (ASHPs) a Silver Bullet?). This is true even after allowing for an electricity grid that was not yet fully powered by wind, solar and other alternatives to fossil fuels. As a family who wanted to do their bit to reduce the impact of our household on global warming, this was a pivotal moment. If we didn’t take the plunge now, we’d be locking in a high carbon footprint for another 20 or so years with a new gas boiler.

• Thirdly, the idea that we must carry out extensive insulation (as the The Retrofit Academy claims) before even considering a heat pump, was simply impractical.  Even if we could have afforded the eye-wateringly expensive costs of ‘deep retrofit’, Listed Building approval would not have been forthcoming. Imagine wrapping this Georgian splendour in external wall insulation, or removing coving and panelling to somehow fit internal wall insulation.

• Finally, at the time, the domestic Renewable Heat Incentive (RHI) was still available but due to close at the end of March 2022, so we needed to get a heat pump system in by then if we were to benefit from the scheme.

Going for it – plans and preparations

So we made the decision in July 2021 to ‘go for it!’. I prepared an outline plan, checklist, questions, etc. We wanted to do it right first time.

I prepared a plan of the house showing the floor area and volume of each room, existing radiators, etc. We bought seven digital thermometers to distribute around the house to ‘benchmark’ how it behaved with the existing gas boiler. I did calculations using current bills and ‘what if’ estimates of future gas and electricity prices to determine whether we would break even on running costs. I remember saying to my wife “gas prices are likely to go up faster than electricity prices sometime in the future. So while our running costs may struggle to compete with cheap gas today, with the current ratio of units costs, that’ll change sooner or later, as the grid gets greener”. Little did I know then that a combination of Putin and a super efficient heat pump would make it sooner, not later.

The biggest challenge was our living room with its large bay window – a thing of beauty but also a significant challenge in terms of heat loss. When the sun is up, even in winter, the room receives a lot of extra warmth (so-called ‘solar gain’), but when the sun goes down, we have to use shutters to reduce heat loss. Not ideal, but sometimes compromises have to be made.

The project forced us to finally clear out the loft. About 100 boxes from our two daughters and us needed sorting, many unsorted from when we had moved in 1998! After several dusty visits to the loft we cleared it, cleaned it, then beefed up the insulation with 300mm depth of Knauf insulation rolls.

The first thing to be grateful to the heat pump for is that it forced us to clear the loft when in our late 60s, rather than having to face it in a decade or two when our knees will have gone. It was such a relief, I can’t tell you.

We had already done what we could to reduce draughts. Our beautiful sash windows could not be replaced even if we’d wanted to. When we had a major servicing job done on them (thanks to a local firm Simply Sash Windows with expertise in historic buildings). We had discreet brushes fitted that paid back immediately with a significant reduction in draughts (when our local climate group did a local ‘retrofit fair’ we included short video snippets on our website to explore options, such as for draught proofing).

For other homes and householders there may be good reasons for carrying out more extensive insulation and other ‘retrofit’ measures. However, bear in mind that if the objective is to get off gas, and thereby reduce your carbon footprint significantly, you really should leave enough in your budget to finance a heat pump system. I discussed this in an essay Insulate Britain: Yes, but by how much? that has attracted a fair amount of mostly positive attention.

For building regulation reasons when we took on the property, we’d had fire strips with brushes fitted around existing internal doors, and it turns out this was very helpful in reducing the turnover of air in the house, and so reducing heat loss.

We contacted two companies specialising in heat pumps. My old friend Chris’s company YES in Yorkshire, and a local company CEG (Cotswold Energy Group) which was recommended to me by a customer of theirs in Nailsworth. Both firms have a proven track record of installing high quality systems (my research assured me). 

We had approval from the Listed Building Officer in August 2021 to proceed, based on outline plans I submitted for locating the external unit well out of sight at the rear of the property. The new services we got installed 25 years ago ran up the back of the building and into a boiler room that could be repurposed for the internal units required for the new system. We had a few lucky breaks like that. The new hot water tank would just fit. Another lucky break.

The surveying of the house is a crucial stage in the process and ought to be whatever heating system is installed. Overall, these were the checkpoints I ticked off on the project prior to installation:

• Listed Building approval ✔︎
• Electrical ‘load survey’ and mains fuse; existing 80 Amps would be fine ✔︎
• Water pressure checks ✔︎
• Sizing of pipework in the house ✔︎
• Cleared loft and installed new beefed up insulation ✔︎
• Sorted out draughts in windows and doors ✔︎
• EPC certificate ✔︎
• RHI requirements met ✔︎
• Check by installers of suitability of platform outside for external units ✔︎
• Check space inside for internal units ✔︎

We were now ready to proceed and made our requirements very clear. For example, I said I was not interested in a hybrid system with gas as a backup; that seemed a bit like buying a petrol car in 1910 and having towing arms for a horse, just in case. A few people tried that, but it didn’t catch on.

The instructions was clear: size the system properly to meet our requirements and those of any subsequent owner (including one who might have 4 teenage children who shower a lot).

The two quotes were very similar in terms of proposed design and costs. In the end, we felt that everything else being equal, going local swung it, in case we needed any call outs to sort out teething issues and future servicing. Chris felt this was an important factor to be taken into consideration.

System design

With a big enough heat pump you can heat any building that a gas boiler can heat. Geneva City Hall has been heated by a water-source heat pump since 1928, and a large air-source heat pump at Hillpark Drive, Glasgow has been heating 350 homes since 2017.

The system design followed MCS standards (Microgeneration Certification Scheme), which stipulates that the system should be able to achieve at least 21°C in living rooms, 18°C in hallways and bedrooms, and 22°C in bathrooms; and be able to do this on the nominal coldest day of the year for your location. For our system and location, that meant that the heat pump itself, and radiators, were sized to achieve these targets with a flow temperature of 50°C, when it was -1.6°C outside. The ‘flow temperature’ refers to the temperature of the heated water pumped to the radiators.

It is crucial that the assessment is done room by room. If just one large room has an undersized radiator, the room may not reach its desired temperature at the expected flow temperature, undermining expectations, for no fault of the heat source itself. In our case, only a third of the radiators

The heat pump itself must be able to achieve the peak heating demand for the whole house.

Because the largest heat pump available at the time was just short of the peak demand estimated for our home, we ended up with a so-called ‘cascade’ system, with two smaller heat pumps working in parallel, in conjunction with a buffer tank. In total, it had more peak power than we needed, but not so much as to cause an issue.

In a way, the so-called ‘cascade’ system (actually not in series but two in parallel) turned out to have the benefit that when it was comparatively warm outside, only one of these units needs to be in operation at any time. The clever electronics made sure that each one shared the work equally over the year.

The only issue was that both firms were maxed out with work because others like us were trying to get things done before the end of the Renewable Heat Incentive (RHI). In addition the Covid pandemic had disrupted supply chains for heat pumps, like everything else, so kit was also scarce.

The installation

Consequently, the installation was finally done at the start of December 2021, taking just over a week. It was a larger than normal system for a larger than normal house, but the principles are the same, whatever the property.

The twin Mitsubishi Ecodans were placed on a platform we created outside the old boiler room:

The new internal setup looked as follows:

You may think that this looks complicated, but it is not a complexity that you need to deal with or even to understand, any more than you need to understand how a modern car works. In many ways, the increase in sophistication of systems (be they cars or heat pumps) reflects their ability to be ‘easier to drive’, extremely efficient, environmentally aware, and with very little in the way of maintenance to worry about. Hardware and software combining forces!

The installation process went as follows:

• Introduction to installation team, and logistics agreed
• Old gas boiler and tank removed
• Existing pipes and radiators flushed
• Heat pump and other kit moved into position
• Plumbing in of kit (heat pump, hot water tank, etc.)
• Plumbing in the few new radiators required 
• Heat pump connected to electrical power
• Control system system installed 
• One wireless thermostat placed in living room, set to 21°C
• Various control setups competed
• Setback thermostat by 3°C to 18°C between 10pm and 6am
• Weather compensation setup
• System put into operation
• Radiators ‘balanced’ to ensure optimal heat distribution
• Metering installed for flow/return heat and electricity usage
• Certificates produced for MCS compliance

In addition to the assessor/ designer who did the design we had two plumbers who did the physical work and ‘plumbing in’ of the new kit, followed by an electrician who did the setting up of the heat pump (see ‘The Team’, Note [3]).

As it turned out, only on a few very cold days (in practice, a minimum of -5°C) did the flow temperature ever get as high as 50°C (the maximum design ‘flow temperature’ for our system). The flow temperature is best kept as low as possible, while still doing its job (see Note [4]).

The results

So how did things turn out?

We get the space heating we need and plenty of hot water. Our hot showers have never been better. We have bills that of course have gone up due to the energy crisis, but less than they would have done if we’d stayed on gas.

A key measure of the success of a heat pump installation is the performance it actually achieves in practice – as opposed to some published figure that assumes an idealised situation.

For any heating system one needs to think of the performance of the whole system. In addition to the heating system, there are the radiators and fabric of the building, because it is all these together that determine how efficiently rooms are maintained at a desired temperature.

For our heat pump, performance was assessed using the ‘seasonal coefficient of performance’ (or SCOP), which is the heat energy delivered during the year (in kilowatt-hours (kWh)) divided by the electrical energy used by the heat pump (in kWh).

For our system, the SCOP achieved in the year to March 2023 was 3.3, or 330% in percentage terms; pretty impressive I feel. That means that for every 1 kWh of electricity we put in, we get 3.3 kWh of heat out of the heat pump (2.3 kWh of this is harvested from the ambient air). By comparison, the old gas boiler was only 72% efficient, meaning that for every 1 kWh of primary energy in the gas put in we got 0.72 kWh of heat out.

This means that even though the unit electricity price is greater than the unit gas price by a factor of over 3, this is more than offset by the relative performance of the new heat pump compared to the boiler it replaced. So our running costs are less by comparison.

This result is totally at odds with the naysayers we hear incessantly in the media that they are difficult to install, won’t work on older buildings without substantial insulation measures, or will cost a fortune to run. None of this received wisdom has been true in our case.

Our experience is in line with the Catapult study (2021/22) that concluded that All housing types are suitable for heat pumps, finds Electrification of Heat project

The project has not identified any particular type or age of property that cannot have a successful heat pump installation. The suggestion that there are particular home archetypes in Britain that are “unsuitable” for heat pumps is not supported by project experience and data.

So I don’t believe our experience is in any sense exceptional or to be treated as anecdotal. Rather, it demonstrates that even in a building that is ‘hard to treat’, as the experts would call it, there is no reason an air-source heat pump cannot be successfully installed and operated, with a little care and preparation.

I hope that our story provides an illuminating corrective to the media naysayers, and others who should know better.

If it gives encouragement to those wishing to get off gas for whatever reason, it will have done its job.

If you are doing it to lower your carbon footprint, then whatever your lifestyle, it will be one of the most impactful decisions you will ever make in your lifetime.

© Richard W. Erskine, 2023

If you liked this essay, you might also be interested in Insulate Britain: Yes, but by how much?

.

NOTES

[1] Burning Questions (and Answers)

Q. So can you really heat a large old house with an air-source heat pump, without ‘deep’ retrofit?

A. Yes. 

Q. But does it cost more to run than than if you’d stuck with a gas boiler?

A. No,  comparatively it costs less. Of course, gas and electricity unit prices have both gone up, but gas proportionally more and the efficiency of the heat pump trumps the differential in electricity-to-gas unit price. See Will my heating bill increase if I get a heat pump? another essay I have written, where I show this in detail.

Q. But it’s really expensive to install surely?

A. It costs less than a new kitchen that is mostly MDF and air, and yet we all seem happy to pay for a new kitchen that the next owner of the house will quite possibly “rip out”. The heat pump will probably last for 25 years and will over its lifetime save you more in carbon emissions than anything else, for a typical householder at least 3 tonnes of carbon dioxide per year compared to gas with the electricity grid as at present (and even more as the grid gets greener). Octopus Energy are creating a market for a low cost heat pump installation, that may meet the needs of a majority of householders, if not outlier cases such as mine.

Q. But can we go green using biogas?

A. Burning stuff is so last century! And as Prof. David Mackay said, if you have some gas it is much more efficient to send it to a gas turbine to create electricity to power a heat pump in your home than to burn the gas in your home. The same would be true of biogas, an idea that has been heavily criticised by independent expert Dr Richard Lowes , and flies in the face of the clear recommendations of the Climate Change Committee, see Note [6].

Q. Should we hold on for hydrogen boilers?

A. No. Using ‘green hydrogen’ from excess renewable electricity to make hydrogen to burn in our homes would require 6 times as many wind turbines than if we simply sent the electricity direct to home to power heat pumps. Hydrogen will be in demand in hard to decarbonise sectors including fertiliser production, etc. Anyone who gets a hydrogen boiler will be locked into expensive hydrogen. The Climate Change Committee expect hydrogen to play only a niche role in heating (see Note [6]).

Q. Did you have to change all your radiators, and pipework?

A. No. The pipework did not need changing and only 1/3rd of the radiators needed fattening a little (from 2-panel to 3-panel); their width and height were unchanged.

Q. Is it noisy?

A. Not at all, even standing close by. If birds are singing you can’t hear it at all.

Q. Can it heat hot (tap) water?

A. Yes. It does that in shortish bursts, and because the hot water tank is under mains pressure, the showers are now much better.

Q. Does it use an immersion heater some of the time?

A. Not for day-to-day water heating. It does use an immersion heater periodically to boost the tank’s temperature from 50°C (the target temperature, achieved with the heat pump) to 60°C (required to kill the organisms); no point flogging the heat pump for this little job. This deals with the risk of Legionella, which is a requirement under current regulations (even though the sealed nature of the system makes it an extremely low risk). In any typical month, the legionella ‘purge’ occurs only for a few hours overnight every two weeks. It has very little impact on the measured performance or the heating bills.

Q. But my plumber said you can’t get the water hot enough?

A. Untrue. Because who wants the hot water from a tap to be more than the new building regulation of 48°C, which a heat pump can easily achieve. Why would we want young or older relatives and visitors to scold themselves?

Q. My plumber also said that radiators never get hot enough?

A. Your plumber is mistaken. Again the new building regulations (whatever the source of heat) are that the ‘flow’ temperature should not exceed 55°C. If your room needs to get to 21°C say, then (depending on the external temperature) you may only need 35°C, 40°C or 50°C (in the most cold days), to get the room to 21°C. Not as fast as if the flow temperature was 75°C, but just as assuredly, and more cost effectively. Your skin temperature is say 34°C, so a radiator at 40°C or even 50°C might feel lukewarm, but that is irrelevant. It’s what the thermometer on the wall says that counts. Don’t touch, look!

Q. Was the project disruptive?

A. Not the fitting of the heat pump, which took just over a week. For us, the most challenging thing was clearing the loft, in order to increase the loft insulation to modern standards, and get the grant.

Q. Do you have smart controls around the house?

A. No. A well designed system (whatever the source), which includes properly sized radiators and weather compensation, will always ensure that if the single thermostat in your living room is reading the target temperature there, the other rooms will also be at their target temperature. No fiddling with controls in rooms or zones. By keeping it simple in this way, we actually can forget we have a heat pump at all. It works reliably, without us ever having to fiddle with any dials or thermostats. We just let it do its thing.

You will have many more questions no doubt, but these are typical of the ones people raise with me, so I hope the answers have proved illuminating.

[2] History of heat and heat pumps

19th Century scientists developed the ‘kinetic theory of heat’.

They established that ‘heat’ in substances was no more than the jostling of molecules, and the higher the temperature, the greater the average speed of their jostling. Not only that, but that this jostling only ceases at ‘absolute zero’, which is -273.15 degrees on the Celsius scale. So the water Roger Harrabin was feeling, say at 7°C, was a sweltering 280 Kelvin (the name subsequently given to this new scale). No shortage of jostling molecules from which energy can be extracted. A balmy -5°C for winter air passing through an air source heat pump – no problem – it being a balmy 268 Kelvin.

As has been noted:

The first working compression–expansion refrigerator (or heat pump) was built by Jacob Perkins in the 1830s. William Thomson (Lord Kelvin) first proposed using heat pumps for space heating in 1852.

For those interested in learning more about the history, Boltzmann’s Atom: The Great Debate That Launched a Revolution in Physics is a great biography of one the principal scientists who shaped this theory, and fought a long battle with naysayers such as Mach who couldn’t bring himself to believe in the existence of atoms.

[3] The Team, for our project overall were:

• Senior manager, who provided oversight at every stage to ensure a quality delivery: during assessment, proposal, installation and hand-over.

• The assessor/ designer. She was in her 20s, a graduate with a 1st Class degree in Geophysical Sciences who’d decided to not go into the oil & gas industry, and instead become a heat pump consultant. She did the surveying of the property, estimating heat loss for every room, and preparing a detailed design and costing for the system.

• There were two plumbers who did the physical work, both young (also in their 20s). They first removed the existing boiler and hot water tank. They then got the units in place on the day, including a heat pump (actually two in our case) that had to be lumbered up a steep path to reach the back of the house (built into the side of a hill); a hot water tank; water pumps etc. They then connected all the pipework to the units. Because they were part of a team, these plumbers didn’t need a deep knowledge of heat pumps, just the principles of low flow temperature systems. They drained and flushed the existing plumbing in the house, and found no issues with it holding the slightly higher water pressure needed for a heat pump. Only a third of the radiators needed up-sizing to triple panel ones – and given all the dire warnings I’d heard (false as it turned out) I was pleasantly surprised by this. Once the system was operating, they ‘balanced’ the system, so heat was as evenly distributed as possible.

• There was one electrician, who not only did the electrical connections, but also setup the controls for the heat pump(s). The so-called ‘weather compensation’ is a crucial part of the setup, as it optimises the efficiency of the system. The temperature of the water being pumped around the radiators is kept as low as needed to deliver the space heating. We also decided to have the heat pump on continuously but set back a few degrees at night. This meant that the heat pump was not having to work too hard in the morning to reach the target temperature in rooms. With our thick walls, it would not be a good idea for these to cool down too much. Paradoxically, we use less energy this way than if the system went off completely over night.

I explore the question of how we scale up capacity in heat pumps in How hard will it be to scale up heat pump capacity in the UK?, another essay of mine.

[4] Flow temperature is best kept as low as possible, that does the job

The larger the surface area of the radiator, the lower the flow temperature needed to deliver the required amount of heat. This is why underfloor heating is often able to operate at a mere 35°C. But even for a system like ours, the flow temperature can be relatively much lower than expected. For example, in the last 7 days, the external temperature has ranged from 8°C to 11°C, and the flow temperature has never exceeded 36°C.

Which is why the talk of ‘high temperature heat pumps’ to deal with difficult buildings strikes me as a non sequiter.

In any case, building regulations that came into force in June 2022 in the UK mean that all new systems must have a flow temperature no higher than 55°C, whatever the heat source, so those trying to flog high temperature heat pumps are barking up the wrong tree in my view. 

Just design the low flow temperature system properly. Job done.

[5] Performance of air-source heat pump based system

The performance for the system relates to the heat pump, radiators and the fabric of the building itself. Any one of these can influence how good the measured performance is. The performance over a given period (say a day) is calculated as the ratio of the heat energy that the system delivers (in kilowatt hours, kWh), and the electrical energy used by the system (in kWh). Because a heat pump harvests energy from the ambient environment (air, ground or water) warmed by the sun, the performamce will be greater than one (or 100%).

For space heating the performance gets less the greater the difference is between the outside temperature and the required internal temperature. We follow a common practice of calling the performance measure the ‘coefficient of performance’ (or COP) and over a whole year, calling it the ‘seasonal coefficient of performance’ (or SCOP).

A 2013 Study by Energy Saving Trust – https://www.gov.uk/government/publications/analysis-from-the-first-phase-of-the-energy-saving-trust-s-heat-pump-field-trial – reckoned that the SCOP for an air-source heat pump is typically 2.5 (or 250%), but as we have seen, it is more common to get a SCOP of over 3 (300%) in 2023, using a modern heat pump and practices.

But what about the daily performance? Well, for our system, even when it was -5°C (colder than the nominal coldest day -1.6°C) the daily COP was 2.1 (210%) and over the year, the daily COP got as high as 5.4 (540%).

The SCOP we achieved of 3.3 (330%) was extremely good, in our view.

Another way to measure performance is the kilowatt hours required to heat one square metre per annum (kWh/m².a). The average British home needs over 130 kWh/m².a, but a new house built to Passivhaus standards would need just 15 kWh/m².a. The Association of Environmentally Conscious Builders aim for 50 kWh/m².a when doing retrofit on existing buildings, with a “certifier approved exemption” 100 kWh/m².a for difficult older buildings. Their standards for doing retrofit can be found here: https://aecb.net/introduction-to-the-aecb-carbonlite-building-and-retrofit-standards/

Our building was using delivered energy of 123 kWh/m².a to heat the house when using the gas boiler. After modest retrofit ‘fabric’ measures and moving to the air-source heat pump, the demand has reduced to about 114 kWh/m².a (not far off the AECB higher “exemption” of 110 kWh/m².a, although I should stress we have not asked an AECB member to review our fabric measures). The reduction in heat loss due to the loft insulation is possibly offset by the 24/7 operations of the heat pump (with overnight setback). See the Appendix for more details.

Nevertheless, I am convinced that the gentler heating with the heat pump (rather than the wild swings we had with the gas boiler) gives rise to improved comfort and energy management overall.

[6] Heat pumps will be primary tool in decarbonising heating

The Committee on Climate Change (CCC) in their 6th Carbon Budget stated (based on very detailed modelling of scenarios, costs and risks):

‘By 2030 37% of public and commercial heat demand is met by low-carbon sources. Of this low-carbon heat demand 65% is met by heat pumps, 32% district heating and 3% biomass. By 2050 all heat demand is met by low-carbon sources of which 52% is heat pumps, 42% is district heat, 5% is hydrogen boilers and around 1% is new direct electric heating.’

Other experts agree. LETI conclude that heat pumps are a far superior option for home heating than hydrogen; see their report Hydrogen: A decarbonisation route for heat in buildings?”, LETI, February 2021

And the district heating itself can be provided by commercial scale heat pumps. Some can have a heat power rating of 48,000 kW (compared to a typical 3 bedroom home needing 6kW) – see https://www.bbc.co.uk/news/business-65321487. Since district heating will often be needed in towns which often have a river flowing through them, or are by the sea, water-sourced heat pumps can be used.

So in practice, combining the domestic and commercial/ district heat pump provision, heat pumps would eventually provide the great majority of the heat ti dwellings of all kinds.

[7] Nesta UK, ‘Heat pumps: a user survey’, 23 May 2023 is a large and comprehensive survey of user satisfaction with heat pumps. Find it here: https://www.nesta.org.uk/report/heat-pumps-a-user-survey/ and a Guardian piece about it here: https://www.theguardian.com/business/2023/may/30/heat-pumps-more-than-80-per-cent-of-households-in-great-britain-satisfied-with-system

Deputy Director of Nesta, Katy King, provided some summary points from survey on Twitter:

• Satisfaction with heat pumps is high and, overall, satisfaction levels between heat pump and gas boiler users are very similar.

• But heat pumps users were MORE satisfied with their running costs than boiler users.

• People who installed a heat pump into their own home were the most satisfied (81% as or more satisfied than previous heating system)

• When you include homeowners who didn’t commission the heat pump themselves in the sample, 73% of heat pump owners are as satisfied or more satisfied with their heat pump compared to their previous heating system.

• Satisfaction with heat pumps is just as high in older properties – People living in Victorian houses were just as satisfied with their heat pump as people in mid-century properties or modern homes

• As the overwhelming majority of heat pump users are satisfied with space and hot water heating, safety, reliability and noise – it’s time to put to rest outdated concerns about the heat pumps.

ACKNOWLEDGEMENTS

I’d like to thank my old friend Chris Wilde, who is Managing Director of Yorkshire Energy Systems (YES), for educating me on heat pumps over the last few years, and showing me the art of the possible. Always witty and wise, I have learned so much from Chris.

Also, thanks to the team from Cotswold Energy Group (CEG), who were very professional at all stages in the project. I’d like in particular to thank Zoe Phillips for her advice and support.

Thanks to Marilyn, my wife, for reviewing the essay and providing improvements, but most of all, for being my companion on this journey.

Any errors in the essay are all mine.

———————————————————————————————————————————————————————————

APPENDIX – A Closer Look at the Data

A number of people have asked me to provide additional data on the performance of the system. I was keen to keep the essay accessible and non-technical but I appreciate that some readers may want more. So here it is – for those that want more!

Data prior to heat pump when house was heated by a gas boiler

Prior to the Air Source Heat Pump (ASHP), the house was heated with a gas boiler dating from 1998 (both space and water heating).

The house is a large semi-detached dwelling with a floor area of 251 m² over three floors.

Prior to 2020 we had done what we could to the house to help reduce heat loss, including:

• fixing guttering
• fixing lime mortar on external walls
• adding brushes to sash windows to reduce draughts
• in a back area fixing insulation of the pitched room of a small extension room

But as to the main building, it is Grade 2 Listed and so certain measures (such as wall insulation). As part of the move to a heat pump (and as a condition for getting the RHI grant), we did add beefed up insulation to the main building’s loft in 2021.

In 2020 the gas usage was 45,567 kWh and cost £1,409 (with an average unit cost of 3p per kWh; including standing charges).

I estimated that 94% of this was for space heating, which gives  42,833 kWh/yr for space heating. But allowing for the old gas boiler being only 72% efficient (see Appendix Note [ii]), the actual house heat demand for 2020 was less, at 30,840 kWh.

This yields a unit heating demand of (30,840 kWh/a)/ (251 m²) = 123 kWh/m².a – when comparing homes this is a very useful measure to use because one can compare two houses of different sizes if one uses this measure of energy per unit area.

Now, given one estimate for the average heat demand for UK is 133 kWh/ m².a, it seems that my old house was already doing quite well – being slightly better than the UK national average per unit area.

Data collection

Most heat pumps come with consoles and / or Apps to enable a householder to monitor energy usage and performance. The Mitsubishi Ecodan does have this capability.

So normally, finding out the Coefficient Of Performance for a heat pump over a day, week, month, or any other period, should be easy and out of the box. Our situation was unusual and more complex than most people would have to experience for reasons I will explain.

In our setup we had two 11.2 kW Ecodans, that work in parallel (what Mitsubishi confusingly term a ‘cascade’ system), so that the peak heat output is 22.4 kW, although the peak heat loss estimated was a bit less than this (at 18.6 kW when the external temperature was at the nominal coldest day of the year, at -1.6°C)

To cut a long story short, it turns out that Mitsubishi’s marketing blurb was wrong and the standard metering features didn’t work with a ‘cascade’ setup as they do for a single unit (which is the most common situation). However, Cotswold Energy Group were brilliant and remedied the situation by installing 3rd party meters to enable performance to be measured and recorded. It did mean we ended up with a slightly more complex and bespoke setup. The good news is that it all works fine and now I don’t have to worry about it – I just read the numbers off a table or graph.

The heat meters get rate-of-flow data from the pumps and temperature data from gauges that were fitted to the ‘flow’ and ‘return’ pipes. The temperature drop between the warm flow and cooler return, multiplied by the flow rate, gives an instantaneous measurement of the heat being delivered to the house (e.g. to the radiators, but when in hot water heating mode, to the hot water tank).

These are the two heat monitoring meters used:

In addition, there was a meter for each heat pump measuring how much electricity they were using:

These 4 meters were then wired up to a data aggregation unit (Teracomsystems TCW260) which does the mathematics to work out the COP for each heat pump (and for the system overall). This local wifi connected unit stores the data over time, and can be accessed by a browser to produce reports, which can be configured in different ways.

I am able to download the data into a spreadsheet to then do further analysis if the raw reports are not quite enough to answer all my questions.

I also manually read the data for a while and did the calculations manually, just to satisfy myself that the system was setup correctly to give me the answers I needed. 

In addition to the heat pump data, I also bought some digital thermometers to place around the house to assure myself that each room was reaching its design target temperature.

When doing visits of people thinking of getting a heat pump, I will get them to place their hand on a radiator and ask them “is that on?”. Oh, it doesn’t feel very hot? They often reply. I then get a thermal gun (cheap to buy – its not a camera – and a useful tool) and check the temperature of the radiator. Say, it is a typical 40°C. I then check the temperature of the palm of their hand – I typically get 34°C (a bit less than a human’s core temperature). “You see, not much different, so not surprising that it doesn’t feel ‘hot’ to touch!”. I then direct their gaze to the digital thermometer on the wall – “what does that say?”. Oh, it’s 21°C! “Indeed, just as it was designed to achieve, 40°C in this radiator is enough to heat this room to 21°C – so don’t be misled by touching the radiator and assume that tells you anything useful, it doesn’t”

By the way, there is just one thermostat in the house (in the living room). The design of the system ensures that when the sitting room is meeting its target temperature (21°C), then the other rooms will meet their temperature. So no need for thermostats or fancy controls on each radiator or in each room. I don’t even touch the TRVs. The system was ‘balanced’ by the installers, and we then just leave it alone.

The heat pump will adjust the flow temperature to deliver the right amount of heat to this and other rooms. The colder it is outside, the higher the flow temperature required and delivered. This is called ‘weather compensation’. In our system, the flow temperature never needs to get higher than 50°C. Most of the time it is a lot less than this. This ensures that the efficiency of the system is maximised (which also means the running costs are correspondingly reduced).

Needless to say, the system has performed brilliantly, thanks to the quality work done by Cotswold Energy Group in the design, installation and commissioning of the system.

We basically now do very little. No twiddling of dials. We just let the system ‘do its thing’.

Performance data for heat pump

In the year from when data was available for the new Air Source Heat Pump – from 1st April 2022 and 31st March 2023 – the total heat demand in the house was 29,689 kWh/a (‘a’ here standing for ‘annum’).

Only slightly (about 4%) less than the inferred heat demand for the pre-heat pump period (despite additional loft insulation).

Due to reduced occupancy, we estimated a slightly lower level of hot water usage, and so space heating was estimated at 96.6% of the total demand, that is. 28,681 kWh/a (see Appendix Note [i]).

In terms of unit area heat demand, that now comes to 114 kWh/ m².a

The Association of Environmentally Conscious Builders (AECB) aim during retrofit to reduce heat demand to 50 kWh/ m².a, but will allow exceptions for difficult to treat older dwellings like mine, with relaxed target of 100 kWh/ m².a. So, our figure of 114 kWh/ m².a is not too far off that relaxed target.

SCOP (Seasonal Coefficient Of Performance)

The total electricity used by the ASHP during the year in question was 8,843 kWh.

The SCOP can be calculated as ‘total heat demand’ divided by the ‘total electricity used’, which in our case was

= (29,689 kWh / 8,843 kWh) = 3.36

This is an astonishly good result for our old house. I had been told by some people to expect much worse. But Zoe had estimated accurately what to expect, and so it was gratifying to see her estimates confirmed in practice.

There is a slight error in this calculation as it does not separate out the direct hot water components (for taps/ showers). But it is only a small error as the great majority of the energy used was for space heating.

Winter COP

Ah, yes, but was it OK in winter? People will be wondering.

I have daily data on heat demand and electricity usage. Obviously this was greatest during the winter months. But interestingly the average COP (Coefficient Of Performance) held up remarkable well; remembering it is not an absolute measure of heat demand but the ratio of heat demand and electricity used.

This is what my analysis found (and is illustrated in the scatter diagram):

• Average COP over winter months 22/23 (Dec’22, Jan’23, Feb’23), was 3.4  [note that the average external temperature was 6.2°C]
• Design MCS coldest day for location was -1.6°C and on actual data fit line this gives a day-COP of 2.3
• Actual coldest day was -5°C and data point gives COP of 2.

It is interesting that in summer, the COP can actually be worse in summer than in winter because, while much less energy is being used, the proportion of that energy being used for hot water rather than space heat is higher and that tends to be less efficient, so the ratio can be worse.

Overall, the winter COP was on average close to the overall annual COP (or SCOP).

Running costs comparison

My current (5th Nov 2023) flexible Octopus tariff charges 26.85p/kWh for electricity and a 52.32p/day standing charge. The gas unit pricing is 6.82p/kWh and 27.47p/day.

If this was the charge for the full 12 months, the cost would be:

(8,843 kWh x 26.85 p/kWh) + (365 day x 52.32 p/day)

= £2,565

Whereas for the same period, if we’d stayed with the gas boiler, we’d need to deliver 29,689 kWh of heat, but to do this with a 72% efficient boiler that would required burning gas with a calorific value of (29,689/0.72), 41,234 kWh, and the per annum costs of this would have been:

(41,234 kWh x 6.82 p/kWh) + (365 day x 27.47 p/day)

= £2,912

So, despite the fact that the unit price of electricity is almost 4 times that of gas (26.85/6.82), the efficiency of the heat pump combined with the inefficiency of the retired boiler more than compensates for the difference.

Of course the unit prices of both electricity and gas have risen since 2020, and even the differential in price has not improved, but the heat pump ensures we are better off than we would have been if we’d stuck with the old gas boiler.

Appendix Notes

[i] How much energy does it take to heat the 300 litre hot water tank?

The temperature in a modern unvented hot water cylinder is stratified with cold water entering from the base of the tank under mains pressure and hot water being delivered also under mains pressure from the top of the tank. The thermostat is in the upper part of the tank.

However, for simplicity, let’s work out how much energy would be needed to heat 300 litres uniformally at 40°C, up to the target temperature of 50°C.

Being careful to do housekeeping on the dimensions.

It takes 4180 joule to heat 1 litre of water by 1°C

So to raising 300 litre by 10°C (from 40°C to 50°C)

requires 4180 (joule/litre.°C) x 300 (litre) x 10 (°C) = 1.254 x 10⁷ joule

1 joule = 2.778 x 10⁻⁷ kWh

So the energy required can be converted to kWh units as follows:

(1.254 x 10⁷ joule) x (2.778 x 10⁻⁷ kWh/joule) 

= 1.254 x 2.778 kWh

= 3.48 kWh

or 3.5 kWh approx.

I used this figure as an estimate of hot water daily energy usage during the period post installation of the heat pump.

[ii] Retired gas boiler efficiency

The retired gas boiler was a Glowworm Hideaway 120B (a Balanced Flue Boiler).

I used the BRE/SAP products database https://www.ncm-pcdb.org.uk/sap/ to get the efficiency figure for this boiler, which stated a ‘SAP seasonal winter efficiency of 72.9%’ and for seasonal summer efficiency 62.8%. Given that winter was when the boiler worked hardest, I used the former figure, rounded down to 72%. Given it was 25 years old, this may have been slightly optimistic.

THE END