Most people think that unless you have a modern, well insulated house, solar and batteries, heat pump systems are a gamble. I have to suggest that, while anecdotally there may be accounts that suggest this is the case, it is in my opinion a myth. We need to separate fact from fiction, and look at the real world and what happens in it. In practice, there is a bias towards old leaky buildings getting bad installations, but it doesn’t have to be that way, if you simply do the basics correctly.
Heat loss calculations
Heat loss calculations are not easy. They are tedious and boring and you need to concentrate and spend a lot of time to get things right. This is why installers cut corners and get things wrong.
Among the heat loss calculations done for me by installers, some were pure fantasy. One appeared to have been done by a blind surveyor, and this from a company with a full on ‘leave it to the professionals’ strap line. Others had careless mistakes (typos, entering 2.3 instead of 12.3) and some made wild assumptions such as reducing the house with rooms to a single giant shoe box and did the heat loss for that.
No wonder they all came out with vastly different results. But done correctly, they will give you a reasonable estimate of the heat loss. What you do is measure all the relevant areas (walls, windows, doors, floors, ceilings, etc.) which is very tedious and boring to do, and then apply (from standard tables) the heat loss in watts per degree centigrade per square metre to each area and add up all the losses to get a total loss, and then add in an allowance for air changes (a sealed room has less air changes than a kitchen with a back door that is frequently opened) and that’s pretty much it, apart from some minor refinements.
There are plenty of freely available spreadsheets available that are set up to do the calculations for you. You don’t have to set up the formulas – you just enter the measurements for each room, and the better ones even have drop down boxes with U values (the loss in watts per degree per area) for common building materials. You will nonetheless have lost several hours of your life that you will never get back, but you will know your heat loss. It’s worth doing.
A heat source is a heat source is a heat source
No heat source cares one iota about what it is heating. To any boiler or heat pump, there is nothing special about heat pumps in this. A huge, well insulated mansion with ten bedrooms on the south coast of England with a total heat loss of 10kW is no different to a battered, old leaky tiny crofter’s cottage in the highlands of Scotland with a total heat loss of 10kW.
In each case, under steady state conditions, which is what you are aiming for, each building loses 10kW. Match that with a 10kW input, and all will be well, however old and leaky the building.
Higher heat losses will always cost more (whatever the fuel)
The other unavoidable fact is that a building with a higher heat loss will always cost more to heat than a building with a lower heat loss. If you live in an old, attractive but leaky-as-hell cottage with a 12kW heat loss, as I do, it will always cost more to heat than a modern, super-insulated house with a heat loss of 6kW, whatever the fuel. I will always burn more oil or gas, or consume more electricity to heat my house than the owner of the aforementioned modern abode.
A pinch of the real world
Now we need to add the real world bit. By and large, most fossil fuel boilers are substantially oversized. In the past it didn’t matter because energy was relatively cheap, so just bang in a big one and be done with it. No need for all those tedious heat loss calculations. Just guess it, multiply by two and fit a boiler that size.
As a result, fossil fuel systems never get anywhere near being stretched beyond their capacity. They can almost always supply enough heat.
Heat pumps on the other hand tend to be matched much more to the predicted heat loss, and that’s where the problems start because there is far less margin for error.
Get the heat loss wrong, which as we know is very easy to do. My heat loss estimates from installers ranged from 9kW to 14kW. There is no way they are all correct. If the error is to underestimate, and then match the heat pump to that underestimate, then the system will fail.
A further pernicious problem is that heat pump suppliers deliberately mislead about their heat pumps output. The headline output (in my case 14kW) is for a sunny day in spring, when I don’t need 14kW. This very same heat pump, however, becomes progressively incapable as outside temperatures fall, and by the time it is zero degrees outside, it is only capable of putting out around 11.3kW. As my heat loss is 12.3kW, given a 14kW output branded heat pump, what could possibly go wrong? As you can see, quite a lot will go wrong once it gets cold outside. The heat pump will never supply enough heat.
Consider the entire heating system
The other real world bit is failing to consider the heating system as a whole. Because heat pumps run at lower temperatures, they need bigger emitters to deliver the same amount of heat.
A cooler-running heat pump delivers less heat than a hot fossil fuel boiler to radiators. Very roughly, the increase needed when moving from typical fossil fuel running temperatures to heat pump running temperatures is to double the size of the radiators.
Even if you match a heat pump’s output to the building’s heat loss correctly, taking into account the lower performance at cold temperatures, if you keep your old fossil fuel rated rads, they will be too small, and the system will not deliver enough heat in cooler weather, and your house will be cold.
Commissioning is critical
Lastly, heat pumps, because there is less margin for error, have to be commissioned correctly. Unlike a fossil fuel ‘fit-and-fire’ solution, heat pumps need to be tuned to match the building they are heating. This takes considerable time on the part of the installer.
Omit this step, as many installers do, and the system will either fail or your fuel bills will go through the roof because the heat pump has been left to run on its most inefficient settings. It will produce heat, but with appalling inefficiency.
Closing thoughts
My own personal experience is in line with the “underperforming heat pump in cold weather” problem. The rads are all upgraded to match heat pump running temperatures, but in cold weather my heat pump fails to deliver enough heat, and my house fails to reach, let alone stay, at design temperatures.
In warmer conditions, it is fine. The building is listed, so reducing heat loss options are limited, and there is no way solar panels are going to be installed. I am 100% reliant on the heat pump replacing my oil fired boiler.
The bottom line is there is no reason why within sensible limits a heat pump can’t heat any building (as long as you have somewhere to put the unit, and a supply of mains electricity). You just have to do the calculations carefully and thoroughly, because there is far less margin for error as I’ve stated several times.
The sad fact is that the heat pump industry in the UK has far too many jobbing workers who have done a half-day course on how to fit heat pumps, which they then install with their old fossil fuel fit and fire mentality. The forums on this site are littered with examples of this ‘workmanship’. Far too many have no idea how to do a valid heat loss calculation (how on earth did my prospective installers come up with heat losses ranging from 9kW to 14kW? They all had the same building to work with).
Result: failed systems. And this just happens more often on old leaky buildings because there is even less margin for error. It doesn’t have to be this way, so long as you design, install and commission the system correctly.
In closing, I had a number of installers tell me they couldn’t possibly fit a heat pump because my house didn’t fit the requirements as it was far too old and leaky. They tended to have a sanctimonious air about them, but they were just as incompetent as the installers who got other things horribly wrong.
To categorically say you can’t have a heat pump is just as wrong as fitting the wrong heat pump.
I have a Victorian house with a heat pump. The house is built of stone with no additional wall insulation. Totally agree this can be heated by an ASHP. The biggest error in the heat-loss calculation was a failure to recognise the effect of cold winds in Winter (the house is on a hill in Wales). So all the corner rooms are too cold while the centre of the house is fine.
Interesting point about the winds. Have you found a solution for the affected rooms or are you just living with it?
“The biggest error in the heat-loss calculation was a failure to recognise the effect of cold winds in Winter” Very true. The official MCS etc calculators take no account of ambient weather, only ambient temp. On solar gain days, my house (old stone cottage as well) gets a bit warmer than predicted, on cold wet windy days, the heat pumps struggles. The wind chill effect on such days, a cold wind blowing onto a wet surface, must be considerable, but is totally ignored by officialdom. But this doesn’t alter the main conclusion, ASHPs can heat old leaky buildings, you just need to size the heat pump correctly.
Were radiators changed and upsized?
Installers are guided by policy and available training, it should be noted.
If those two collude to a disaster then it’s not the installer’s fault!
Yes they were. Interestingly the quotes I had varied on the need to do this, some said yes (correct), others said no (back to school, dear boy). In reality I suspect they were just trying to undercut other quotes by leaving out the radiator upgrade. How they would have cooked the books for MCS (they were all MCS certified, not that that means a thing, as this example shows) is far from clear. Or it may be that given MCS don’t know their arse from the elbow, they can’t spot a rigged calculation.
The bizarre thing is it took me about 10 minutes to understand the need to upgrade the rads if the current ones are sized for a fossil fuel system. Cooler circulating water means you need a bigger surface area to transfer the same amount of heat. It is baked into every heat loss/transfer calculation, loss/transfer = U value x area x delta t. Since U remains constant, the only way to maintain the loss (from the rad to the room) when the delta t is less (because rads are cooler) is to increase the area. It’s not even O level maths. Yet installers chose to ignore it…
I do agree about policy (mostly bonkers) and available training (wholesaler half day courses on how to make fast bucks by fit and fire installing our heat pumps). But the bad installers also collude.
Apologies but I must again defend installers, about radiator changing!
To be clear there are benefits which you are enjoying to doing this.
In reality though with Compensation controllers the need is minimised.
This is because the compensation controller alllows heat going into the house ride the thermal inertia of the building, ensuring a peak load is reduced. But you do need “compensation controls” to achieve this
A heat pumps own controller is always a compensation controller, and it assumes there are no other controllers other than the heat pumps own brand which communicates seamlessly with the pump
If I was quoting I would have explained that and you could have made a value judgement on the cost and disruption associated
But be clear, you haven’t wasted money!
“A cooler-running heat pump delivers less heat than a hot fossil fuel boiler to radiators”
This is not quite correct. To maintain a room or house at say 19c requires the same amount of heat whatever the fuel
Compensated systems will do it at a lower temperature than on off systems and heat pumps will do this at lower temperatures and at higher flow rates.
You are right, I should have been clearer. Yes, to maintain something at a constant temp (assuming constant other conditions) requires the same amount of heat, a kWh lost and replaced is a kWh lost and replaced, wherever it came from, the same amount of heat is delivered, wherever it came from.
Perhaps what I should have said is “A cooler-running heat pump delivers less heat than a hot fossil fuel boiler to the room/house because of the lower temperature difference between the fossil fuel sized rads and the rooms (typically a difference of around 20 degrees, compared to say 50 degrees for fossil fuels) means less heat can be transferred. To compensate, you need to increase the size of the radiators.” I usually try to avoid equations when explaining things, preferring to use plain old words, but I might have added the equation I used in my earlier comment: “This can be seen form the heat transfer equation, heat transferred = transfer coefficient (U value, a constant) x transfer area (size of the rad) x temp diff between rad and room. Given the transfer coefficient remains the same, if the temp difference is less, then the area has to increase, to maintain the same amount of heat transferred”.
It is always a delicate balance, putting maths into words does risk over-simplifying things, but too much maths – and for some people any maths – is a real turn off for a lot of people – they just stop reading, which defeats the object of writing the thing in the first place.
The other thing I didn’t mention in the article is that if you have a small old leaky building, then you will start running out of available wall space to put the huge new rads. But there are solutions: K3 rads, tolerate a slightly higher leaving water temp. Again, it was a trade off between article length and information overload. The key, fundamental headline remains the same, ASHPs can work in old leaky buildings, you just need to make some adjustments. Furthermore, it has to be that way, if the government is to have even the slightest hope in hell of meeting its ASHP installation targets.
In terms of the room temperature you want, the heat requirement will remain the same, irrespective of the fuel, and for that matter the heat generator or emitters.
“In terms of the room temperature you want, the heat requirement will remain the same, irrespective of the fuel, and for that matter the heat generator or emitters.” I agree, providing you add “all other things being equal”, because heat loss and so heat requirement is very dependent on outside ambient temperature and in deed other things.
The key point that I was trying to get across with this piece is that an ASHP can work in an old leaky building – with some caveats, my installation proves that – but you have to take a systems approach, and realise that you are no longer running a high flow temp blast furnace system, but a Steady Eddy system, and there are two main consequences: firstly, your system will usually be on 24 hours a day (think tortoise and hare) and secondly, you are very likely to need to increase the size of your radiators. Given this is likely to be the most disruptive part of the installation, it is a good idea to manage expectations by saying something along the lines of you will need to increase the size of your rads, unless we can show the current ones are big enough.
I’m not sure I understand “In reality though with Compensation controllers the need is minimised.This is because the compensation controller alllows heat going into the house ride the thermal inertia of the building, ensuring a peak load is reduced. But you do need “compensation controls” to achieve this.” As I understand it, there are two types of compensation, weather compensation and load compensation. In the former, flow temp is set according to the outside ambient temp, in the latter flow temp is set according to the building’s heat demand (load). In both cases, the adjustment of output is achieved not by turning the boiler/heat pump on and off, but by varying the flow temp.
My system uses weather compensation, and, as I understand it, it has very limited feedback about the thermal inertia of the building. Almost all of the control is done based on the outside temperature, room temperature doesn’t even get a look in, and the only possible feedback about what is going on in the building is the RWT (returning water temperature) but I am not sure any use is made of this. That leaves me confused as to how it can ‘ride the thermal inertia of the building’ (which sounds rather too like management/sales speak to me…).
Mitsubishi, and maybe some other, heat pumps have something called Auto Adaption, whereby the controls ‘learn’ the heating characteristics of the building. If for example the building is slow to heat up in the mornings, the controller will ‘learn’ to add in a morning boost. My Midea unit can’t do this, and that is why it can’t recover from an overnight setback. It only knows the outside temp, and has no idea the building is colder than it should be, and needs a boost. The consequence for now, unless and until I can ever get Home Assistant to provide more finely tuned control, is that I need to have the system full on 24/7, ie no setback. That can work out rather expensive. Last night it was mostly -3 outside, and during the overnight period I burnt 4kWh per hour for most of the time. 4kWh x 10 hours = 40kWh. At today’s electricity prices that is expensive. A setback would have reduced some of that cost.
I am very aware there is a debate about whether always on or setback and recover uses the most energy. My current understanding is the latter does reduce consumption (or to put it the other way round, always on will always cost more). Maybe another article is called for, covering this debate in more detail, but for now, I would prefer to have a setback, but it doesn’t work, my house can’t recover from the setback in a reasonable time (it takes days not hours), because the ‘dumb’ heat pump only knows about the outside temp, not what is happening in the building.
“I would prefer to have a setback, but it doesn’t work, my house can’t recover from the setback in a reasonable time (it takes days not hours), because the ‘dumb’ heat pump only knows about the outside temp, not what is happening in the building.”
I live in a similar property to yours. Have you considered a setback between 10pm and 4am with higher flow temperatures between 4am and 7am by manually increasing weather compensation before you go to bed and reducing it again when you get up? This won’t work when you are near design temperatures, but might save you money with ambient temperatures above 4 deg C.
I have, but I think I need to try again with different timings, even earlier start time for the setback, the high thermal mass means I probably won’t notice any very slight cooling at bedtime (actually a good thing, it helps sleep) and perhaps end the setback earlier as well, maybe 2100 to 0300, and hope that gives enough time to recover. I could manually adjust the WC curve, but there is a risk I forget and it is a bit of a chore, eg going into the WC pages on the controller turns everything off, you have to remember to turn them back on again.
Longer term, if/when I get Home Assistant set up, I hope I may be able to have all this happen automatically, ie different curves for different times of day.
At last a sensible piece about fitting heat pumps to older buildings. This made a lot of sense to me. Of course a heat pump can be fitted to an older house (in my case 1850 or earlier) but will it do the job and will it be cost effective? These are crucial questions and the author emphasises the importance of getting the right answers. I am personally very wary of trying to find an installer who really knows what they are doing and how can you tell? ECO4 funding is potentially available but the armies of installers promising the earth is very off-putting.
A key point for me is that a building is a building, and heat is heat, and in that sense it doesn’t matter how or where the heat comes from, it can be from a fossil fuel boiler, a heat pump or even the sun through solar gain, it is all heat, however it got generated and delivered.
As long as the design calculations are sound, a heat pump will do the job just as well as any other man made heat source. The ‘as long as’ is a very important caveat, as the lower flow temps used by heat pumps means the rads may well need upgrading (bigger ones that can deliver enough heat even at lower flow temps), and pay particular attention to the heat pumps output at lower outside air temps. My nominal 14kW heat pump falls to a little over 11kW when the outside air temps fall to around 5 degrees, and to make matters worse the deaded defrost cycles start at around these temps. In these conditions, my heat pump fails to maintain the desired indoor air temp (by a few degrees). I would have done better with a 16kW heat pump, but then again my bills might have been higher.
As a side note, my 14kW Midea may be a generic 12/14/16kW unit that is just set to 14kW via dip switches (ie software limits the output of the base 16kW unit to 14kW), meaning I could very probably make it a 16kW unit by changing the dip switch settings, but I have never got round to doing that, because for the majority of the time, the heat pump does keep the house at design room temperature.
Will it be cost effective? It depends on what is meant by cost effective… If we consider just costs, it will be expensive, but an old leaky building will always be expensive to heat (it is a price we pay for living in old interesting and attractive buildings), precisely because it is old and leaky (though improvements can and should be made, the building will still perform badly compared to a modern well insulated building). Another key factor is you generally run heat pumps ‘low and steady’ ie on most or all of the time, whereas fossil fuel systems are often on ‘fast and furious’ timers, meaning in effect you move from timed heating to 24 hours heating, and that can have a significant effect on costs.
The forum has much information about the necessary calculations, all of which are doable, they are not complicated, though it has to be said some of the assumptions are sometimes not straightforward. They key ones are a sensible realistic building heat loss assessment, and then matching the heat pump and emitter output to the building’s heat loss, especially at lower outside air temperatures.
Knowing whether an installer is up to the job in advance is almost impossible, with one exception. I was ‘lucky’ in that the grant body ‘preferred’ installer demonstrated its incompetence at the design stage, before any installation took place, meaning I could refuse to have them anywhere near my house. Instead I was able to use an installer I had found. The best you can do is switch on all your cowboy detectors from the get go, ask questions questions and then more questions, and if you even get a whiff of horse manure, send then packing, and start looking elsewhere.
It is not a quick (or stress free) process. It took me over a year from deciding I would have a heat pump to getting one installed. Part of the delay was listed building and planning consent, which I guess will not infrequently apply to old leaky buildings.