← Back to Coffee now

An air-source heat pump in an older UK terrace house

Coffee nowOpen original ↗

The gas boiler in our house has definitely been on its way out over the past few years; not only that, so have several of the radiators, and having lived all spring with them off to prevent leaking, the looming onset of summer’s end, meant it was time for decisions.

A combi boiler on a wall in a large cupboard. Next to it is a timeswitch

Before: The combi boiler

The cheapest option, would have been just to replace the gas boiler. But that’s no way to decarbonise your heating, so the other option is to take the plunge and spend a bit more (see conclusions for a summary) to install an air-source heat pump.

We’d been aware that heat pumps were a thing for a while but hadn’t seriously considered the possibility that one could be suitable for our home until relatively recently. We’d also heard of people getting heat pumps in the past, but it always seemed to be associated with modern houses, perhaps built to exacting insulation standards. But, in general, heat pumps are becoming more mainstream and visible, with big-brand energy companies promoting their offerings and giving the impression that they can be suitable for a lot of cases.

Indeed, we’d started to see articles about air-source heat pumps in regular, or even ‘heritage’, properties and seen examples on social media of installations in homes that weren’t totally dissimilar to ours. We certainly don’t have a modern well-insulated house. Ours is a solid wall end-of-terrace built around the 1910s (and each room also has a fireplace).

Although we’d taken care of the easy wins for improving its insulation, including improving the level of loft insulation, modernising the double glazing, applying chimney balloons to limit the draughts from the four (four!) unused chimneys, draft proofing round the front door and replacing the letterbox with one to modern standards, there are still areas where the heat loss is considerably higher than that for modern houses.

The house is also in a conservation area, which puts conditions on what you can and can’t do. Primarily, in this context, adding external insulation, as you’d often do in such properties, isn’t really possible. The wooden front door is also undoubtedly still a point of some heat loss, but is part of the heritage character of the building.

So, would a heat pump work for us and how much would it cost?

Would it work?

We won’t go into too much technical detail on the operation of heat pumps, but it’s worth mentioning some basics that affected the options open to us.

In short, today, you should be better off (modern houses) or at least shouldn’t be worse off (older houses) on the ongoing bill long-term for heating your house with a heat pump, because, although UK electricity costs several multiples (about four) more than does gas per kWh, a heat pump is several multiples as efficient.

A simplistic explanation of why is that you’re moving heat inside from heat that’s already there outside, not just creating heat from the electricity or gas.

This is similar in principle to how a fridge works; heat is transferred from a colder place (outside your house or the inside of the fridge) to a warmer place (the inside of the house or the outside of the fridge). For reasons related to thermodynamics, this requires some energy input (obtained from the electricity supply) to get the heat to move from the colder place to the warmer place.

One important point is that, the energy input required to pump a given amount of heat increases as the temperature of the source (the outside) decreases or the temperature of the output (the flow temperature of the liquid in your heating system in the case of air-to-water heat pumps) increases.

There’s not much you can do about the outside temperature, but the flow temperature of the heating system is an important parameter. The lower that temperature is, the more efficient the system can be to run, but the trade off is that a lower temperature will require larger radiators to transfer the heat. We should note that lower flow temperatures are more efficient on gas boilers too – just not as crucial.

To obtain a sensible flow temperature, we were definitely open to replacing some of the radiators, not least because several of them had been old when we bought the house and it felt we were already in a process of replacing them one by one as they started leaking.

How much would it cost?

From the above preliminary assessment, it was clear that the work would involve getting several of the radiators replaced with larger ones (mainly doubles where there were currently singles), and in some cases, having new radiators added, especially in our main living room (this was obvious to us because, even in the pre-heat pump setting, the single small radiator in the room wouldn’t quite cut it on a colder day).

To add these, as well a new water tank, means you’re looking at a moderate project, not one you can just get someone to do without quotes and without understanding what the work involves.

Government grants

In the UK, there is currently a Boiler Upgrade Scheme offering grants of £7500 towards the cost of upgrading from a gas boiler to a heat pump. As this is a significant fraction of the overall cost of such an upgrade, it almost certainly makes sense to take advantage of the grant.

From the consumer point of view, the main requirement to be eligible for the grant is to have the work done by a company that’s registered with the Microgeneration Certification Scheme (MCS). Unless you had some very unusual situation (maybe wanting to DIY a lot of the work), you’d almost certainly want to pick a company that is registered with the MCS in any case as the scheme will require the installation to conform to various standards.

It is also worth noting that MCS certification also means the installer has to have insurance in case they can’t complete the work. This was reassuring to learn when we had some last minute ‘what if…’ questions occur to us.

Getting quotes

The first step (or at least it’s essential for the MCS grant) is to get an energy performance certificate (EPC) for your house. Although we’d had one when we first bought the house, that had long since expired.

If you’re not sure if your property has an EPC certificate, you can search for EPCs online.

That wasn’t too hard to arrange (you also need one to sell your house, so there are serval providers around); we just went with someone local (Ballard Property). You can expect to pay in the region of £100, for which you get an inspection looking at details of construction, insulation levels, current heating system etc., as well as creation and registering of the certificate.

Compared to replacing the boiler like for like, there is more to consider and the overall scope of the work is bigger, so next we certainly wanted to get multiple views of how to proceed.

It is of course ideal to get recommendations from people you know who have had similar work done. Unfortunately, in our case that wasn’t really possible, as although we were able to track down a few colleagues and acquaintances that had heat pumps installed, these were all in different areas to us. That being the case, we used the installer search provided by the MCS to find heat pump installers based locally and then perused their websites to form a shortlist of possibilities.

There are also larger companies with more national coverage (including several of the large energy suppliers) to consider, and there are certainly pros and cons to those.

In our experience, you can expect to get some initial suggestions as to what type of system could be appropriate for you and a rough indication of the price based on either a short visit and /or phone call, or perhaps just providing details via the web. We went to that level of detail with a handful of companies, and it ruled out one or two where the cost seemed surprisingly high compared to others.

Overall, this initial process gave us a good impression that the recommended options were broadly similar in terms of heat pump specifications and that the cost would be something we could afford and justify.

The next stage was getting more detailed quotes for the work, including proposals that included the model of heat pump and radiators and recommended flow temperatures for operation.

This involves the following steps:

  • a more through survey of the house, measuring room dimensions and accounting for windows, doors and other details
  • thermal calculations based on the above to determine the heating requirements (the heat loss at the lowest outside temperature will determine the power the heat pump will need to supply)
  • production of a report proposing details including heat pump model, flow temperature and radiator locations, as well as the estimated power usage and cost compared to the existing heating system

Prediction of the heat requirement isn’t an exact science. There are likely to be uncertainties in variables, such as rate of ventilation, temperature of party walls in terraced houses, and outside and ground temperature, that can lead to significant variation in the results. It was notable that both the proposals we received (and also the EPC) suggested a fair bit more energy would be needed to heat our home than we’d historically seen through our gas usage.

Although we believe most installers will use tools provided by the MCS (which aren’t freely available to others) for the thermal calculations, in principle you could follow similar steps with other heat-loss calculation software to come to your own conclusion. [We purposely decided not to go down that path to avoid delay through over-analysing details.] We would expect most installers would err on the side of installing a slightly oversized heat pump (which would be only be slightly more costly) rather than risking the system being under powered.

There’s a fair bit of work to prepare these detailed quotes and proposals, and therefore most installers charge for this service (in our experience in the range £250–350). This tends to constrain the number of such quotes you can obtain, but given our lack of experience and the variation in the initial estimates, we felt we should get at least two to compare.

In both cases, the quote provided wasn’t the final word; rather it was expected that there is some follow up refining details. In hindsight, we feel that either of the installers we’d got quotes from could have provided a good system.

We picked one to go forward with based on a combination of factors: impressions of how easy communication had been, how many details in the proposal seemed they might need some more attention and the proposed heat pump specs, as well as (slight) differences in price. Our work was done by Brook View Plumbing.

On the choice of heat pump itself, we were more or less happy to go with the model recommended by our installer.

Installers typically tend to have a manufacturer or manufacturers that they have a relationship with, and then select the most appropriate model from that range. Although some manufacturers do make datasheets etc. available, it felt that over analysing that info might not be too helpful. The model we went with is from the Mitsubishi Ecodan range.

Heritage consent

Following recent changes to planning laws, our heat pump installation wouldn’t need planning permission, even if close to a property boundary. However, as our home is in a conservation area, we had to follow a local consent process to obtain permission, similar to other changes that affect the external appearance of the building.

This took a few weeks to complete, but as our application conformed to the relevant conservation area guidance, there were no major hiccups. We were asked about noise calculations, but as the relevant information was included in the heat pump proposal from the installer, it was easy to provide. Looking at the other visible applications, we could see that at least a couple of other properties had similar plans.

While waiting on this consent, we were able to look in more detail at other aspects of our plans.

Some more decisions

Radiator options

As mentioned above, there’s a trade off involved in sizing radiators for use with a heat pump system. In general, increasing the size and number of radiators allows a lower flow temperature to be used (so the heat pump will be more efficient, and offer lower bills in future), but is likely to be more expensive and disruptive to install, and there’s often limited available wall space.

One important point is that the input temperature is going to be the same for all your radiators, so effectively they all need to be sized in proportion to the heating requirements of the room. That is, if there are some rooms that have relatively small radiators, those will be the limiting factor when it comes to setting the flow temperature and there won’t be much benefit to having more oversized radiators in other rooms.

This meant that, contrary to our initial thoughts, it wouldn’t make sense to add underfloor heating in the living room without corresponding changes elsewhere. However, a major reason we’d been considering that option had been to remove an awkwardly positioned radiator that was already undersized for the room, and we realised that we could achieve that by replacing with larger vertical radiators in other positions.

That decision was made early in the process, but there were also some potentially problematic aspects that were resolved by the proposal:

  • it was difficult to increase the size of the radiators in the kitchen enough, so instead we opted for a fan-assisted plinth heater to meet the requirement
  • the bathroom ‘towel rail’ didn’t have a high enough output, so we decided to add an electric element to that for extra heating if necessary

So far we haven’t used either to a large extent.

A vertical panel designer radiator behind a door. A piece of abstract art hangs on the wall.

One of the new radiators in the living room

We also felt that the initial suggestion on a triple radiator in the hall would have been a bit impractical close to the door and in the end went with a double radiator there and another vertical radiator at the other end of the hall.

A vertical panel designer radiator in a hallway

An additional radiator in the hall

Radiator positions and styles is an area that can be more of a personal choice, so it also made sense to go through that fully.

When all was finalised, the radiator schedule was:

  • dining room: keep existing radiator
  • living room: remove existing undersized radiator and replace with two tall radiators in more convenient locations
  • hall: replace existing single radiator (old and leaky) with a double radiator and add one tall radiator
  • kitchen: keep one existing radiator, replace another with a slightly larger double radiator and add a plinth heater
  • bathroom: add an electric element to the existing towel radiator
  • front bedroom: keep existing radiator
  • back bedroom: replace existing radiator (old and leaky) with similarly sized radiator
  • study/bedroom: replace existing radiator (yep, also old and leaky) with a larger double radiator

Thermostats

Although the thermostat isn’t a very expensive part in the overall work, it will be something you’re directly interacting with, so it’s good to do some research on this.

We wanted something that would have potential to be controlled remotely, but without too much data being passed on, so we preferred something that should interact with the open-source software Home Assistant. We chose a Heatmiser neoStat WiFi.

More on using this with Home Assistant will follow in a separate blog post, but you can also control this remotely through a ‘normal’ account with the manufacturer!

A thermostat mounted on a wall

The thermostat

It’s worth noting that heat pumps (or indeed modern heating systems in general) tend to use thermostats that support multiple set temperatures rather than a simple on–off control and single temperature setting for when it’s on.

Typically, rather than turning off the heating completely, it is set to a lower ‘set back’ temperature at night or when briefly away to avoid the system taking a prolonged time to get back to temperature. For this reason, it wouldn’t have made sense to reuse either the time-switch or analogue thermostat that we had.

Hob

An induction hob heating a saucepan and frying pan

The new hob in use

We also decided to replace our gas hob with an induction hob, so that we could fully disconnect the gas and avoid paying a gas standing charge, as well as improve inside air quality.

Fortunately, we were able to choose an induction hob that seemed it would be a simple replacement, fitting the existing worktop hole. This is actually also much easier to clean than a gas hob.

Hot water tank

Heat pumps aren’t able to produce instant hot water as our combi boiler had been doing, so we’d be reinstating a water cylinder.

We initially had some doubts about this aspect as the cylinder we’d been used to from the heating system we inherited with the house had been vented, and we associated it with low hot-water pressure (which at the time had meant it was used in conjunction with an electric power shower).

However, this turned out to be nothing to worry about, as the proposed water cylinder is unvented, as is now common/standard for use with the heat pumps, so the water is at a good pressure for showers etc.

We’d also wondered about the efficiency of maintaining a large tank of hot water, especially as we’re relatively light users, preferring showers to baths.

In hindsight, the previous tank was probably pretty inefficient being a large, warm-to-the-touch copper cylinder with some unconvincing lagging. Modern tanks with integrated insulation seem a different proposition altogether!

The spec suggests a heat loss of 55 W, which works out as ~500 kWh per year. That seems quite OK when considering that it’s heated by the heat pump, so the electricity used will be reduced by the heat pump COP. Probably, the cost due to losses from the tank is ~£50/year, and for half the year those losses are actually heating the house to some extent anyway.

A hot water cylinder in a large cupboard. Also visible is a control box, pump, expansion vessel, pipework

After: hot water cylinder and controls

Installation week

The installation itself was done over 5 days (a Monday to a Friday) scheduled in late September, so just before the heating would be required, and also before what must be a busy period for plumbers as heating is turned on for the first time in the year!

We were at work (in the office rather than trying to work from home) during the week, so didn’t see all the plumbing occurring, just occasional updates and queries.

However, each night when we returned home, there was a good amount of progress.

The boiler was removed early on (no turning back after that), and the new cylinder installed in its place. Radiators were removed and then new radiators appeared. The heat pump itself was put in position before connecting the electricity supply and pipework. Also on the first day, we had the gas meter removed. This had been relatively painless to arrange after only a small amount of negotiation with a chat bot and seemed to only need a few week’s notice.

Most of the electrical work, including replacing the hob (which did indeed fit smoothly into the worktop) was done on the Wednesday. Last but not least, there was the installation of the various control electronics (an indoor box with flow temperature controller, as well as the thermostat) and a fancy-looking pump and expansion vessels.

As promised, there was only one evening without any sort of hot water and one more evening where the strong smell of jointing compound (?) meant we decided eating out would be better!

The rear of a terraced house with a heat pump visible

The outdoor unit and pipework

We had wondered about the practicalities of changing the radiators en mass, worrying that our furniture would need a lot of moving around to accommodate access to the radiators and also to lift floorboards where pipes needed to be adjusted under the floors. That turned out not to be too bad, as the radiators in our dining room and front bedroom were suitable to keep unchanged, so we were able to move items out the way in those rooms.

By the end of the week, everything was done, almost perfectly to schedule.

We had one small hiccup/minor panic when we woke up on Saturday morning with no power to half the house! It turned out that the new immersion heater (used for a scheduled Legionella-killing cycle, which started in the early hours) had chosen that moment to fail and trip the RCD. That was quickly diagnosed and replaced by the installer, and everything has been running smoothly since.

Conclusions

We’re planning another blog post (or posts) to go into some more details on our post-installation experiences so far, namely:

  • control of the heat pump and automation with Home Assistant (disclaimer, we feel like this is a slightly nerdy topic, and not really something that you actually need to worry about as a prospective heat pump owner unless you wanted to!)
  • smart electricity tariffs and carbon intensity

In the meantime, we think we’re pretty happy so far.

The upfront cost, even once the grant was taken into account, was slightly higher than that for a gas boiler, but it seems the running costs should be at worst similar and possibly a bit cheaper. That part will become a bit clearer when we have a bit more data on the electricity usage to assess whether a time-of-use or agile tariff could make sense for us.

We might also hope that the lifespan of a typical heat pump might exceed that of a typical modern gas boiler.

To put it in perspective though, the overall costs are pretty small compared to the cost for even a relatively modest Electric Vehicle (EV), another ‘big ticket’ item that can help decarbonise.

For us, being fortunate to be able to commute by train and therefore being relatively light car users, the carbon reduction from no longer using a gas boiler definitely exceeds what we could hope for by replacing the car.

Also, to answer some worries we had or had heard:

  • no we’re not cold – if anything it’s a more pleasant, uniform warmth
  • it’s not broken the bank – possibly it will work out a bit cheaper long term
  • no, it’s not noisy – unlike the gas boiler, we can’t notice it at all when indoors and it seems similar volume to a fridge in the garden (when it’s on, which it mostly won’t be in summer when we’d be in the garden)

So all in all, so far so good!

Overall, we’d say we found this change a bit easier than we’d feared, having been mulling it over for a little while but not quite having the impetus to actually do it, possibly assuming that it would need more work on the house than turned out to be the case.

Part of the reason for this post is to encourage anyone in a similar position that a heat pump could work for them too.

However, we can also see that there were a lot of factors that made it possible for us that might not be the case (we own the property, there were obvious places to put the pump and the cylinder, we had the funds to not need to go for the absolute immediately cheapest option and perhaps cushion us if the running costs were more than expected, we’re probably relatively able to cope with a week of slight disruption, and we had the time and energy to at least understand the basics of what we were buying and go through the various choices we needed to make).

Ultimately, although we can probably all make some changes, minimising the worst effects of climate change is clearly a global challenge that will need a collective determination. That will certainly need a continued political will to encourage take up of green technologies and ongoing de-carbonisation of power making it easier for people to adapt.


Extra shots

Comments (0)

Comments (0)

No comments yet.