Skip to main content

What Size Heat Pump? (and lots more besides!)

“What size heat pump do I need?” seems to end up in confusion and arguments, especially on social media.

So we go deep here trying to help you answer that question. Grab a brew, this is going to be a long one. I make no apologies for that, there’s a lot to cover.

But don’t fret, as well as lots of reading there’s some “Do It Yourself” calculations you can take part in. It’s proper a Blue Peter job this one.

Right from the outset though I want to make this clear. I am NOT a heating engineer, I’m an IT Geek by trade.

I’m just a regular punter that likes to know how things work. And I have picked all this info up on my journey into low temperature heating with our combi boiler and now onto sizing our system for our soon to be installed air source heat pump.

The info has been gleaned from a variety of sources over the past year, many of whom are mentioned directly within the article. I thank them for their help.

So this article is presented from the view point of an informed punter and I hope other folk find the content useful.

There is no paid affiliation here. I’m telling you honestly about the tools I’ve found value in. There are of course, other fine resources out there.

If you are in any doubt or have any questions about your own heating system, always consult and hire an expert. Your heating system is likely the most complex thing in your house.

With that said, on with the article.

Things you need to get right for a well performing heat pump system

  • Correctly size heat pump unit (the heat source)
  • Correctly sized radiators / under floor (the emitters)
  • Correctly sized primary pipework that runs around your house

If you get any or all of these wrong and you are not going to have the most efficient setup.

By “Correctly Sized” we mean matching the 3 different elements of the heating system (listed above) to the heat demand of the property .

When folks have already got combi boilers sized at 24kW, 32kW and even 40kW hung on their walls they are rightly sceptical when they see Heat Pump sizes starting at just 3.5kW and going up to a maximum of around 12kW in a single unit.

Combi Boilers are primarily sized for the instantaneous hot water demand of the property and in many cases are oversized for both heat and hot water.

This is usually because there has been no formal calculations done to determine the correct size boiler required to match the demand of the property.

Remember, the average heat loss for UK properties is around 6kW to 8kW. (source: Heat Geek)

Almost ANY house can be heated by a heat pump, leaky or not!

The key is to match the heat loss of the property with the heat output of the heat pump.

In this guide we will talk about the following topics:

  • Room by room heat loss survey (we detail the survey of my house)
  • Limitations of a room by room survey
  • What is Design Outdoor Temperature (DOT)
  • Flow and Return Temperatures
  • Radiator Sizing
  • How Weather Compensation works
  • Do it yourself heat loss calculation tools (a chance to try the process yourself)
  • Heat Geek Watts Per Square Metre (a DIY for you to try)
  • (HTC) Heat Transfer Coefficient (You can try this one too)
  • Michael De Podesta formulas and calculations (more DIY for you to try)
  • A summary of all the above calculations
  • In house, time based thermal performance testing
  • Ensure pipe sizing is correct (both heat pump and heating)
  • What size hot water cylinder
  • Final thoughts and disclaimer

The majority of the information provided here is all real world.  This article is me putting my notes down that I have been taken during my journey from an oversized combi boiler (32kW) towards installing a heat pump.

So I will be sharing the calculations and the outcomes of surveys on our 1930’s 3 bed semi detached house with a hope they help you and allow you to think critically about your own property.

Then long term, once we have had the heat pump installed (hopefully late 2022) we can look back and see which method of sizing turned out to be closest to reality.

When the heat pump goes in I’m going to be installing a full monitoring system based around a heat meter and the fantastic Open Energy Monitor platform.

So I should be able to share lots of performance data and see how the actual data compares to the sizing calculations. Here is a screenshot of what information the Open Energy Monitor platform can show.

Room by Room Heat Loss Survey

The standard method of sizing a heat pump is a room by room heat loss survey of your home as mandated by the Microgeneration Certificate Scheme (MCS).

Every heat pump installed via the MCS process should have a full home room by room heat loss survey done. And to claim any government grant, like Boiler Upgrade Scheme (BUS), the MCS process must be followed.

As mentioned in my Low Temp Heating document, the heat loss survey involves a heating engineer coming to your house and surveying the whole property.

But you don’t have to wait until you want to install a heat pump to get your property surveyed.

You can use the Heat Geek “find a heat geek” installer map to find someone capable of doing this in your area.

Heat Engineer also has a ‘search engineers’ page with links to engineers who can do heat loss calculations for you.

Even if you’re still running a gas boiler, LPG or Oil.  Knowing the heating requirement of your property and heat loss of each of the room is vital information if you want to try and get ‘heat pump ready’.

The survey process starts with each room being measured and all elements in each room are noted. 

Room dimensions, window types, their size, what is the room next door/above/below, what are the walls made out of, how thick are they, what insulation do they have in them. 

Note: It would be helpful if you knew some of this or had information to hand. Maybe you have building plans/drawings about the house or any extension.

Also what property type (semi, terrace etc), whether the property is in an exposed location and it’s elevation above sea level.

The survey would also record details about the current emitters; ie underfloor heating or radiators sizes and types (P+, K1, K2 etc).

Emitters can get a bit confusing as there are a few different types of standard looking radiators available.

  • P1 also known as Type 10. (1 radiator panel and no convection fins)
  • K1 also known as Type 11. (1 radiator panel and 1 set of convection fins)
  • P+ also known as Type 21. (2 radiator panels and 1 set of convection fins)
  • K2 also known as Type 22. (2 radiator panels and 2 sets of convection fins)
  • K3 also known as Type 33. (3 radiator panels and 3 sets of convection fins)

Here is a diagram showing the different radiator types.

Note: this whole room by room process could take a few hours depending on how big your house is.

If you’re having a heat pump installed and this process has not taken a few hours on site, then it probably won’t be detailed enough.

Once the engineer has all the raw data they can go away and create the report for you. Again, this can take them some time.

When finished the engineer will then send you over a completed report.

Note: no engineer will do the heat loss calculation for free.

They will certainly charge for the work if undertaken as a one off piece of work. As mentioned above, it takes many hours of work.

Alternatively you may be charged, then refunded as part of a full heat pump installation.

Ask the engineer what their preferred method is as they will likely all have different processes.

Limitations of a room by room survey

The obvious limitation to the room by room method is that you or the engineer will probably never truly know the exact materials that have been used to build your house.

Assumptions can and will be made based on the property age.  

  • How thick are the building materials that you can’t see?  
  • What is and how thick is the insulation behind the plasterboard?
  • Is there any thermal/cold bridging?  
  • Has all the insulation been taped?  (ie, how well is it installed)
  • How well has the cavity insulation been installed and is the insulation uniform within the whole cavity?  
  • How many air changes per hour? (ie, how draughty the house is)

The thermal performance of materials can differ greatly depending on how well or badly things have been installed.  Which in turn, could affect the results of the survey.

Air changes per hour (ACH) can differ greatly between houses, even those built at the same time. 

Take two houses in a semi-detached pair. One homeowner may have been meticulous sealing up draughts around skirting boards, under floors and around windows.  Whereas the next door neighbour may have done no work.  One house may have double glazing, the next single glazing.

But because the properties are the same age, both could be given the same ACH rating.

The only way to get an accurate ACH reading is to undertake a blower door test.

But this would give a ‘best case’ reading. So with all doors, windows and extractors closed.  So not quite real life.  People come and go in houses, open doors and windows, use extractors etc.  So the ACH setting used in your heat loss calculation needs to be based on the real world.

So it is questionable what value a full blower test brings if you then can’t use the result?

At the other end of the spectrum, the default ACH settings in the MCS room by room calculator spreadsheet are very much at the worst case, assuming zero work on sealing up draughts.

The engineer completing your heat loss survey needs to make a judgement call on your property based on the information you can provide. Again, using their experience.

So as you can imagine, with lots of guesstimates around building materials and air changes, there are challenges with the room by room method.

But these challenges don’t mean don’t do it, far from it.  It is still vitally important to do the survey to get the heat loss for each room so that you can choose the correct sized radiators, underfloor and ultimately the heat pump.

Remember, this process is mandatory for any MCS installation.

What is Design Outdoor Temperature (DOT)

So there are two aims of the home survey

  1. Room by room heatloss (to size the emitters – radiators or underfloor)
  2. Whole house heatloss (to size the heat pump)

The whole house heat loss is obtained by totalling up the heat loss from each individual room.

A room’s heat loss (in watts) is how much heat each room will lose at your ‘design outdoor temperature’ when the room is at your ‘target indoor temperature’.  

On my heat loss survey we went with a target indoor temperature of 20.5 ℃ for the downstairs “lived in” rooms and 18  ℃ for bedrooms.

Once you know both target indoor and design outdoor temperatures you can size radiators to match the heat loss of the room.

Design Outdoor Temperature (DOT) varies where you are in the country.  Here are a few examples taken from the MCS Spreadsheet:

  • Belfast (-1.2 ℃)
  • Birmingham (-3.4 ℃)
  • Cardiff (-1.6 ℃)
  • Edinburgh (-3.4 ℃)
  • Glasgow (-3.9 ℃)
  • London (-1.8 ℃)
  • Manchester (-2.2 ℃)
  • Plymouth (-1.2 ℃)

Note: you can put your postcode and elevation into the MCS Spreadsheet and see your own location Design Outdoor Temperature.

When designing a heating system, engineers will ensure that there is enough heat being provided at Design Outdoor Temperature.  So if you’re in Manchester, the DOT is -2.2 ℃.

Obviously if you are further north than Glasgow, you may amend the DOT to be lower than the listed -3.9 ℃ and design the system accordingly.

Similarly, if you want your target indoor temperature to be 23 ℃, then the calculations used would reflect that.

Elevation also affects DOT.  The higher up you are, the colder it can become compared to being at sea level.

In my case, our area of Sheffield is 100m above sea level, so my DOT comes out at -2.5 ℃.

Simply put, the bigger the difference between outside design temperature and target indoor temperature, the bigger the calculated heat loss will be. Makes sense right?

Once you have the heat loss for all the rooms you add all those up and this will give you the total heat loss for the property at design outdoor temperature and your target indoor temperature.

The total of all the rooms’ heat loss is then the figure that would help you choose the heat pump size.  Remember, you size the heat pump for this worst case design temperature scenario.

The size of hot water cylinder and water use in the house may also influence the size of heat pump. Hot water cylinders are discussed later in the document.

As a real world example, here are the results of the room by room survey undertaken for me by Damon Blakemore Plumbing Heating and Renewables (Heating Installer of the Year for Yorkshire and National runner-up 2022).

Our house is a 1930’s semi with 15 year old cavity wall insulation and mainly 10-20 year old double glazing.  There is a loft conversion which has PIR insulation, but I know this insulation has been poorly installed.

The chart below shows the different target indoor temperatures and the calculated heat loss in watts for each room when -2.5 ℃ outside.

Room Target Indoor Temp Heat Loss (watts) at -2.5 ℃
Kitchen / Dining Room 20.5 ℃ 1702W
Lounge 20.5 ℃ 886W
Hall 18 ℃ 315W
Landing 18 ℃ 838W
Bedroom 1 18 ℃ 517W
Bedroom 2 18 ℃ 327W
Bedroom 3 18 ℃ 573W
Bathroom 18 ℃ 293W
Total   5451W (5.45kW)

So the survey suggests I would need a heat pump capable of generating 5.5kW of heat when it is -2.5 ℃ outside.

Note: not all heat pumps live up to the badge on the front of them.  Some 6kW models may not hit 6kW at -3 ℃.  Some 5kW models may provide 7kW at -3 ℃. 

It will be a task for your heating engineer to ensure the correct model is chosen based on the results of your heat loss calculations.  They will choose the appropriate make and model by analysing the heat output charts of the different models available and match those to your specific heat loss and flow rate requirements.

So picking the right tool for the job.

Flow and Return Temperatures

Flow is the temperature of the water (or glycol) leaving the heat source (heat pump) and return is the temperature of the water when it comes back to the heat pump having gone all around your radiators and/or underfloor heating.

In most cases the engineer sizing for a heat pump will aim for there to be a 5 degree difference between flow and return temperatures in the system.  Which is known as DT5.

So if you’re running at a flow temperature of 45 ℃, the return will come back the heat pump unit at 40 ℃.

Running DT5 means the heat pump isn’t having to work as hard to turn the returning 40 ℃ water back into 45 ℃ before sending it back out to the emitters again.

With a heat pump, the lower you can get the flow temperature the higher the performance you will get.

The same is true of a combi boiler, hence the recent drive to get people to lower their flow temperatures as mentioned in my Low Temp Heating blog post.

Take a look at this snapshot summary from the Vaillant Arotherm Plus heat pump brochure.

As you can see, you will get better performance the lower the flow temperature that you can run through the whole heating system.

Look across the row for the 5kW model as indicated by the red boxes.  

This shows the SCOP at each flow temperature.

  • 35 ℃ flow – 4.48
  • 40 ℃ flow – 4.13 
  • 45 ℃ flow – 3.77 
  • 50 ℃ flow – 3.41 
  • 55 ℃ flow – 3.06
  • SCOP = seasonal coefficient of performance (average across the whole year)
  • COP = ‘instant’ coefficient of performance (like a point in time reading, say last 5 mins)

The SCOP/COP number signifies that for each unit of electricity you put in, you will get the SCOP figure back out as heat.  

At 40 ℃ flow, a SCOP of 4.13 means 1kW of electricity in, 4.13kW of heat out.
Whereas at 55 ℃ flow, a SCOP of 3.06 means 1kW of electricity in and 3.06kW of heat out.

Over the whole heating season and over the years, this could add up.

Remember though, a combi boiler is normally only around 85% efficient. 1 unit of gas in and just 0.85 units of heat out.

So a Vaillant heat pump running at 55 ℃ flow is still way more efficient than a combi boiler. (306% versus 85%).

Although if talking pounds and pence, as of August 2022, electricity is still 3 times more expensive than gas. So the efficiency of the heat pump is almost cancelled out by the price difference.

Obviously, the carbon savings of the heat pump blows the gas boiler away. No comparison.

But you would hope that as electricity becomes ‘greener’ with more renewables coming onto the grid, this will swing the pricing in favour of the heat pump.

If you already have solar PV and perhaps battery storage that can take advantage of time of use tariffs where electricity is cheaper overnight, then you’re already swinging the cost per unit in your favour.

A heat pump, solar PV and battery storage is heralded as the dream setup.

You could even call it the Holy Trinity of Home Renewables!

I am very lucky that I will have that setup very soon.


Radiator Sizing

The key to radiator sizing is to ensure the heat output of the radiator is capable of matching or exceeding the heat loss of the room.

The challenge with heat pumps is that they work best at much lower flow temperatures than boilers do. The lower the flow temperature chosen, the lower the heat output of the radiator, but the higher the efficiency of the heat pump.

So we need to find the lowest flow temperature that matches the heat loss of the room when it is -2.5 ℃ outside (or your own design outside temperature).

Let’s take my lounge as an example.  If you remember from the previous table, the heat loss of that room when -2.5 ℃ outside and 20.5 ℃ inside was 886 watts.

So to satisfy that heat demand we would need a radiator big enough to match or exceed that heat loss.  

In the chart below you can see the original radiator we had in the lounge, a 1400 wide x 600 high P+ (type 21) type under the bay window.

Then below it we have also listed the outputs from ‘bigger’ K2 (type 22) and K3 (type 33) variant radiators of the same dimensions.

By ‘ bigger radiator’, we mean it has a higher surface area, i.e. usually more panels and more fins, so can output more heat (watts).

Note: the heat output listed below is in watts, taken from the Heat Calculation software, working to a target indoor temperature of 20.5 ℃ and design temperature of -2.5 ℃.

Type Height Length Room Heatloss at -2.5 ℃ Output at 35 ℃ Output at 40 ℃ Output at 45 ℃ Output at 50 ℃ Output at 55 ℃ Output at 60 ℃
P+ (21) 600 1400 886 388 577 782 1000 1230 1470
K2 (22) 600 1400 886 471 700 949 1214 1492 1784
K3 (33) 600 1400 886 637 947 1283 1641 2018 2412

You can instantly see that ‘bigger’ radiators output more heat across all the flow temperatures.

To achieve 886W of heat output that the lounge needs, we’d need to run 50 ℃ flow on the original P+ radiator (it is capable of 1000W at  50 ℃).

The K2 would be able to run at 45 ℃ (outputting 949W) and the K3 could output 947W at just 40 ℃ flow.

This is also the reason why underfloor heating is recommended as it can run very low temperatures, ie 35 ℃ and below because there is such a large surface area.

Here is the 1400 x 600 K3 (type 33) radiator we put in the lounge.

Using the heat loss survey tool you would be able to go through each and every room looking at the current emitters and finding a common flow temperature that all rooms could all satisfy running at -2.5 ℃ (or your DOT temperature).

Luckily for the engineer completing the survey, the Heat Engineer software does most of this for them and is why it is the go-to heat loss calculation software out there right now.

Be sure to check it out over on their website.

My engineer used this for my survey and I have also signed up and had a play with it.

Once you have that baseline of emitters and flow temperatures you can start to think about where you could upgrade radiators to bring the flow temperature down (and increase performance).

For some folk it might be just one or two radiators, for others it might be every radiator in the house.

It’s obviously a balancing act of cost, aesthetics, upheaval and reward (higher SCOP, higher performance).

Remember: a higher SCOP means it will cost less to run the heat pump as you’re getting more heat out for the same energy you put in.

But with the Vaillant Arotherm Plus capable of 300% efficiency even with 55 ℃ flow, many homes may find that many of their existing radiators are already okay. Although it would be advisable to drive down the flow temperature wherever possible and achieve even better SCOP.

The following diagram shows how most of my old set of radiators were capable of supporting 55 ℃ flow when -2.5 ℃ outside.  

If I’d planned for 55 ℃ flow I would have only had to replace the Landing and Bathroom radiators with slightly bigger ones as they just fell short.

But I took the decision that as part of the upgrade to the heat pump I wanted to try and get as low a flow temperature as possible by replacing my existing radiators with ‘bigger’ versions.  This meant we mainly had to upgrade to K2 variants and one being a K3 radiator.  

Almost all my replacement radiators are the same height and width as the previous versions, which meant for an easy swap with little disruption.

Although I did have to install a new second radiator in the kitchen/dining room space to be able to run lower flow temperatures. It was going to be impossible to run really low flow with just a single radiator in there.

Note: I did most of the radiator changes over a period of around 12-18 months.  Slowly upgrading them when funds allowed or we were redecorating a room for example. This was all part of my ‘get heat pump ready’ plan whilst running the combi boiler through last winter.

This was the advantage of having the heat loss survey done even when we were running the combi boiler. It gave us the information to plan the radiators upgrades. So we could choose the right sizes to ensure we could hit our common target flow temperature.

By making the radiator changes that we have, we should now be able to support a flow temperature very close to 40 ℃ when -2.5 ℃ outside as shown on this updated diagram.


As we saw from the Vaillant datasheet, this drop from 55 ℃ to 40 ℃ flow will make a difference to the seasonal coefficient of performance (SCOP), 3.06 versus 4.13.

This will save us money in the long run as we won’t need to buy as much electricity to output the same amount of heat. So money invested now to save money later.

But let’s be honest, we’ve done a deep retrofit here by replacing almost every radiator close to get to 40 ℃ flow. Your home and situation may not require such upheaval.

How Weather Compensation works

Once you have worked out the flow temperature that your system is capable of outputting at design outside temperature you can move onto weather compensation.

So let’s say the Heat Engineer software has come to the conclusion that when it’s -3 ℃ outside you need 43 ℃ flow.

When it’s -2 ℃, you might only need 42 ℃ flow.  Then when it’s -1 ℃, that becomes 41 ℃ and at zero  ℃ outside, perhaps 40 ℃ and so on.

Because it isn’t as cold outside, the heat loss (in watts) of the room reduces as the outside temperature rises. The difference between the outdoor and indoor temperatures reduces.

Most heat pumps (and combi boilers for that matter) will or can be fitted with an external temperature sensor that reports that reading back to the heat pump (or combi).

The external sensor would normally be fixed on a north facing wall out of direct sunlight and wired back to your heat pump or combi boiler.

Take a look at this mock-up where we have chosen 43 ℃ flow at design temperature -3 ℃ and the calculated heat loss of 5.4kW and target indoor temperature of 21 ℃

So each degree of temperature change is 225W (5400W / 24)

24 being the difference between -3 ℃ outdoor and 21 ℃ indoors.

Outside Temperature  ℃ Flow Temperature ℃ Heat Loss (watts)
-3 43 5400
-2 42 5175
-1 41 4950
0 40 4725
1 39 4500
2 38 4275
3 37 4050
4 36 3825
5 35 3600
6 34 3375
7 33 3150
8 32 2925
9 31 2700
10 30 2475
11 29 2250
12 28 2025
13 27 1800
14 26 1575

A simplified example perhaps, but one to illustrate how the weather compensation sensor can influence and modulate the flow temperature of the system.  And how the heat loss of the property decreases as the outside temperature rises.

In this example, when it’s 12 ℃ outside you will only need a flow temperature of 28 ℃ and 2kW of heat compared to 43 ℃ water in the radiators for a heat loss of 5.4kW when -3 ℃ outside.

And remember, the lower the flow temperature the higher the COP we can achieve.  So why not automatically get lower flows without lifting a finger?

Here is a sample heating curve. If you find where your DOT and target flow temperature meet you can see how the curve slopes off.

I highly recommend you consider using weather compensation for both heat pumps and combi boilers. 

This can be a simple add-on for many combi boilers that could save you a lot of gas. If you’re still running gas I would absolutely recommended speaking to a heating engineer about bolting on weather compensation.

Do it yourself Heat Loss Calculation Tools

If you know the building fabric components of your house and are wanting to give the room by room heat loss survey a go yourself, there are a number of free and cheap ways you can go about it yourself and have a play.

Heat Punk provided by Midsummer Energy:

MCS Spreadsheet:

Heat Calculation:

Trystan Lea:

Messing around with these tools and calculators they really helped me understand parts of the heat loss process.

Heat Geek Watts Per Square Metre

If you know the floor space of your house you can use the following Heatgeek cheat sheet to get an estimated heat pump size.

So find your property on the list and make the calculation. Obviously this is guidance only as homes come in all shapes and sizes with wildly different heat loss profiles.

M2 (metres squared) relates to the floor area of your whole property. So measure the width and length of the floor area in every room in the house. 

Length x Width = area squared.

In my case our total floor space is 98 square metres and I class my 1930’s Semi Detached in the “Renovated with cavity wall insulation” bracket.

So I took worst case, middle case and best case from the figures with came out as:

Best case: 40W x 98m2 = 3920W (3.9kW)
Mid case: 52W x 98m2 = 5096W (5.1kW)
Worst case: 65W x 98m2 = 6370W (6.4kW)

You can find this cheat sheet, full explanation and accompanying video via this Heat Geek article on their website.

Heat Transfer Coefficient (HTC)

Warning: If you’re not a fan of complex calculations, I’d suggest skipping this section as it goes deep into the weeds.

TL;DR, I put the result of the calculation in the summary section later on.

Heat Transfer Coefficient (HTC) is something that was first brought to my attention by Nathan Gambling, host of the excellent (and award winning) Beta Talk renewable heating podcast and his Beta Teach website.

HTC is a way of determining the maximum heat loss of your property by using a combination of fuel used to heat the home and the average daily outside temperature.

You can take the measurements for single days, but ideally you’d do this for between 2 to 4 whole weeks across the heating season. So stack up many days of data. The more data the better.

I will explain how it works using my Gas usage data from January 2022.

Requirements for the single day calculation

  • Daily gas usage (for heating only) across the whole 24 hour period
  • Average daily outside temperature for the whole 24 hour period

How to get daily gas usage

The old skool way is to manually record the figure each day from your gas meter.

If you have a smart meter then you should be able to get this from your supplier.  Maybe via their app or their website.

Personally, I’ve had great success using the Carbon Coop Powershaper website.

For a small annual fee you can interrogate your smart meter data regardless of which supplier you have been with.

The data is historical too, so if you join today, you’ll still be able to get any historical smart meter data stored.

This screenshot shows how my January 2022 daily gas usage is presented.

If you scroll across each day you can see the daily totals, which is a very simple way to pull out each day’s usage. They also allow downloading 30 min usage data in CSV format.

But what if you are cooking on gas and/or have a combi boiler for hot water?  Won’t this contaminate your data?

Yes! HTC works best by using pure heating usage data, so we have to eliminate cooking and hot water usage from the daily figure.

The best way I found was to look at your daily gas usage in July.  As it is very unlikely your heating is going to be on in July, so the data you see will only be cooking and/or hot water depending on your setup.

Here is my data from July 2021.

So simply take the total figure (448kWh) and divide that by the number of days in July (31).

This gives me an average of around 14kWh per day.

We can then use that as our cooking and water baseline and remove that 14kWh from each total day usage through the winter season to leave just heating.

So the 83kWh I highlighted on 8th January 2022 becomes 69kWh (83kWh total – 14kWh hot water/cooking).

Note: If you go away on holiday in July you may have blank weeks, then maybe choose days in June or August to achieve the same result and find average daily gas usage from summer days.

Average Daily Outside Temperature

The second element we need to be able to calculate HTC is outside temperature data.

We need the average temperature across the whole 24 hours for each day we are going to calculate.

Weather Underground is a good place to get that from.

You can see here that the average Sheffield temperature on January 8th was 4.7 ℃.

Note: Weather Underground will use the closest weather station to your chosen city.  Hence it has chosen the one at Doncaster Airport for me.

The HTC Calculation

Let’s stick with January 8th as we have that data to hand.

First off, take the 69kWh heating only gas figure (total from whole day gas minus the 14 kWh baseline cooking/water July average) and divide that by 24 hours.  

So in our case, (83kWh – 14kWh) = 69kWh / 24 = 2.875kWh

This gives us a figure showing how much energy is required for each hour of the day, in this case 2.875kWh

Average outside temperature was 4.7 ℃

We will use 21 ℃ as a target inside temperature.

You can obviously adapt the calculations to suit if you have different internal temperature targets.

21 ℃ outside minus 4.7 ℃ leaves 16.3 degrees.

Next calculation is taking the per hour gas figure of 2.875kWh and divide that by the 16.3 degrees difference.

So 2.875kWh / 16.3 = 0.176

Multiply this by 1000 to give us a value in watts; 176 watts.

Next we take that 176W and apply a combi boiler efficiency calculation.

If you remember from the Low Temp Heating document, a combi boiler will operate at different efficiencies depending on the return water temperature coming back into the boiler.

My Combi return was coming back around 40 ℃ all through this winter, so my efficiency should be around 94%.

So 1 unit of gas in and 0.94 of a unit back out in heat.

If you are at 55 ℃ return, then choose 85% for this calculation. Or choose your own point from the diagram.

So we take the 176W output and apply the 94% efficiency to that.

176 x 0.94 = 165.4W

This is now the amount of energy required to raise the internal temperature by 1 ℃, so ‘watts per degree’.

The final calculation we make is to find out the maximum amount of heat output required when it is -3 ℃ outside (or whatever your DOT temperature is).

We look to get the difference between our outside DOT and the inside target temperature.

So in my example, this will be 21 ℃ minus -3 ℃ which leaves us with 24 degrees difference.

Once we have this figure we multiply this by the ‘watts per degree’ figure.

So 24 x 165.4W = 3970W (4.0kW)

Having done all the whole calculation, for this single day, the calculation suggests that the peak amount of heat required to heat the house when it’s -3 ℃ outside and you want 21 ℃ inside would be 4.0kW

Summary of calculation

  • Gas used for only heating 24 hours: 69kWh
  • Gas used for heating per hour: 2.875kWh
  • Target inside temperature: 21 ℃
  • Target ‘design outside temperature’: -3 ℃
  • Average outside temperature 24 hours: 4.7 ℃
  • Difference between average outside and target inside: 16.3 ℃
  • Combi boiler efficiency: 94%

And for those that like diagrams:

BUT, this is just one day.  As stated above, it is recommended that you run the same calculation over at least a 2 to 4 week period through the heating season.

Here is a summary of 22 days in January 2022 highlighting the variance of results from day to day with rounded up/down figures.

Note: By putting your daily gas and average outside temp in Excel columns (or other spreadsheet software), you can quickly whip up the rest of data like this using the above calculations and a bit of copy/paste.

Date Avg Outside Temp Heating Watts Combi 94% kW Required
05/01/2022 2.6 124 116 2.8
06/01/2022 1.5 167 157 3.8
07/01/2022 2.2 156 146 3.5
08/01/2022 4.7 174 165 4.0
09/01/2022 4.7 149 140 3.4
10/01/2022 6.6 170 160 3.8
11/01/2022 6.6 121 114 2.7
12/01/2022 5.3 138 129 3.1
13/01/2022 5.1 134 126 3.0
14/01/2022 3.3 131 124 3.0
15/01/2022 2.2 141 132 3.2
16/01/2022 5.2 148 139 3.3
17/01/2022 4.7 144 135 3.2
18/01/2022 3.8 141 133 3.2
19/01/2022 5.6 115 108 2.6
20/01/2022 2.2 122 115 2.7
21/01/2022 3.5 143 134 3.2
22/01/2022 5.6 184 173 4.2
23/01/2022 4.5 146 137 3.3
24/01/2022 2.5 150 141 3.4
25/01/2022 1.9 132 124 3.0
26/01/2022 6.1 151 142 3.4
Average 4.1 145 136 3.3

Note: Yes, there’s some variance in the figures. Even on the days where it was the same average temperature the kW output is different. Was it colder in the day rather than night? did someone bump the thermostat up? did we use more water/cooking that day than our 14kWh average?

Lots of variables, which is why it is recommended to use as big a data sample period as possible, to iron out the extremes and outlying days.

So taking an average across those 22 days in the above table, the heat loss per degree was 136W and the calculated full maximum heat output required would be 3.3kW (136W per degree x 24 degree difference between -3 ℃ outside design and 21 ℃ indoor temp).

Suggesting that a heat pump with a maximum output of around 3.3kW could suit.

First thing that jumps out is that this 3.3kW result is way lower than the 5.5kW recommendation of the room by room heat loss survey.

There is also suggestions that it’s difficult to compare the energy usage of an on/off boiler and that of a run long and low heat pump.

We will round up and summarise all these various calculation results at the end of the article.

Software based HTC Opportunity

I think someone could make a killing here if they could come with a web based HTC calculator.

  • Input your postcode (to get local average daily temperatures)
  • Link to your smart meter data (to get daily gas usage)
  • Ask for your annual/daily water/cooking demand (so you can remove this from your daily total leaving only heating gas usage)
  • Ask for Combi boiler efficiency (or just use 85% default)

Then do the full HTC calculation across the whole heating season, say October to March.

Like an extension of the Carbon Coop Powershaper used earlier.

Who is up for coding that? Could we make it Open Source?

Michael De Podesta Formulas and Calculations

As you will have seen from the previous HTC section, the calculations can be long winded and laborious.

Michael, a retired physics engineer, has looked at how to condense and simplify those calculations. 

You can find Michael via @Protons4B on Twitter.

He created a video on how he came to create these simplified calculations and how to apply them.

You can find the video here:

He provides two simple calculations

  • What heating power do I need to raise the temperature by 1 ℃?
  • What size heat pump do I need?

Let’s look at each one in turn and apply my details to them.

What heating power do I need to raise the temperature by 1 ℃?

His calculator for this is:

  • Annual Gas Usage (kWh) divided by 57.3

Michael says that his calculation has accounted for, and added a margin of error if you include gas for cooking and hot water in the gas figure used.

In his video he says both cooking and hot water account for only a minor part of his annual consumption.

I agree on the cooking, this is a very minor amount for us. But our gas usage for hot water has been far from ‘minor’.

  • Total gas usage in 2021: 14243kWh
  • 365 x 14kWh summer daily average (cooking and water): 5110kWh
  • Left over (just heating usage): 9133kWh

That shows that one third of our annual gas usage was used by hot water and cooking.

Note: the 14kWh for cooking and water was derived in the earlier “HTC Calculation” section.

Obviously, the amount of people in your house and how many showers/baths they have can affect this figure either way.

So I will use both our ‘full’ and our ‘just heating’ gas usage figures in Michael’s calculation.

  • Full gas usage: 14243 / 57.3 = 248.5W
  • Heating only: 9133 / 57.3 = 159.4W

So the calculations suggest either 159W or 248W to raise the temperature by 1 ℃.

What about at -3 ℃ outside and 21 ℃ inside?  That’s a 24 degree difference.

  • Full gas usage: 24 x 248.5 = 5964W (5.9kW)
  • Heating only: 24 x 159.4 = 3825W (3.8kW)

So as you can see, using the ‘heating only’ gas consumption figure comes out close to the 3.3kW figure of the long winded HTC calculations. But not quite.

The ‘full usage’ figure at 5.9kW is closer to the 5.4kW of the room by room heat loss survey.

What size heat pump do I need?

Michael’s second calculation looks to simplify things even further by using annual gas consumption to predict what size heat pump you will need.

  • The calculation is simply “annual gas usage / 2900”

In the video he explains how he arrived at the formula including how he used ‘degree days’ to determine outside average temperatures.  It is well worth a watch.

Again, I will use both the ‘full annual’ and ‘ heating only’ gas figures in these calculations.

  • Full annual gas: 14243 / 2900 = 4.9kW
  • Heating only gas: 9133 / 2900 = 3.1kW

Because our hot water usage is such a major proportion of our annual usage I cannot disregard that, hence doing these calculations both with and without.

Your house may well be different, so please play with the figures yourself.

By not including the hot water and cooking, the result of 3.1kW seems very low compared to the room by room heat loss of 5.5kW?

When using ‘all gas’ usage, Michael’s figure comes out at 4.9kW, which is closer to the room by room heat loss of 5.5kW.

A summary of all the methods used

Here’s a quick summary of my results for all methods listed above.

  • Room by Room Heat loss: 5.5kW
  • Heatgeek watts per sqm best case: 3.9kW
  • Heatgeek watts per sqm mid case: 5.1kW
  • Heatgeek watts per sqm worst case: 6.4kW
  • Full HTC calculation: 3.3kW
  • Michael HTC calculation (all gas): 5.9kW
  • Michael HTC calculation (just heating): 3.8kW
  • Michael ‘what size heat pump’ (all gas): 4.9kW
  • Michael ‘what size heat pump’ (just heating): 3.1kW

Phew, quite a range. As low as 3.1kW and as high as 6.4kW

The lowest sized heat pump right now is the Vaillant 3.5kW model. But there are plenty of options in the 5kw and 6kW range to satisfy properties with a low heat demand.

Our plan is to go with the Vaillant 5kW model as it can easily cover the 5.5kW at -2.5 ℃ from the room by room heat loss as well as modulate down as low as 1.5kW heat output.  Which should be fine even when it is 14 ℃ outside.

It will be interesting to see when I eventually get our heat pump installed and start monitoring how close these surveys and calculations come out to be.

Word of caution though, all the methods apart from the room by room just give you a single total heat loss figure at design outside temperature.  

Whilst this is okay for ball park figures, you still need the room by room survey done so you can size radiators (and heat pump) correctly.

As shown in the diagrams previously, matching room heat loss to radiator size and target flow is imperative to getting a balanced system.

In house time based thermal performance testing

The final option for testing is probably the gold standard.

These are the likes of the Veritherm and Co-heating tests.

Both these tests use sensors placed around your house over a set time period.

The Veritherm method heats each room in the home to 25 ℃ inside (using fan heaters). They monitor exactly how much electricity was used for the fans and electric heaters that they put around the home.

Then they switch off the heat and just monitor the temp drop in relation to the outside temperature. This gives them the coefficient of heat transfer by using sensors both inside and outside.

You can then use this data in the same way as the room by room survey.  But with the confidence that the results have been measured rather than calculated.

I personally didn’t stretch to the cost of these tests, but if money allowed I’d love to see how the results of that stack up against the other methods.

Pipe sizing (both heat pump and heating)

Now we have the heat pump sized and correctly sized our radiators we move over to arguably the MOST IMPORTANT (and arguably most overlooked) part of the system design; pipe sizing.

There are two main pipe types we need to be aware of for heat pumps

  • Pipes from outdoor heat pump unit to indoors
  • Primary heating pipework around your house.

Both these pipe sizes need to be capable of carrying the heat load of the property.

Heat Geek has done a tremendous job here with an article, a YouTube video and also providing a cheat sheet that explains the whole subject. The penny dropped for me when I read all this.

If we look at the cheat sheet and concentrate on the DT5 column.  DT5 is the usual difference in temperature between flow and return when installing a heat pump.

Note: DT = temperature differential is not to be confused with DOT = Design Outside Temperature

DT10 and DT15 shown on the cheat sheet (and even DT20) are usually for other fuelled boilers (gas etc).

These higher DT listed like DT15 are useful in condensing boilers for example. You shoot out 65 ℃ water and it comes back at 50 ℃, so in the condensing zone as shown on the earlier boiler efficiency chart.

But for heat pumps, DT5 is the target.

So if you’re running at a flow temperature of 45 ℃, the return will come back the heat pump unit at 40 ℃.

Running DT5 means the heat pump isn’t having to work as hard to turn the returning 40 ℃ water back into 45 ℃ before sending it back out to the emitters again.

If you do run DT7 or DT10 on a heat pump you are going to see a big drop in performance and SCOP.

A simple pipe size requirement example from the cheat sheet, if you are putting a 6kW heat pump in, then you need 22mm copper pipework to carry 6kW of heat if running DT5.

Now this doesn’t mean that the whole house needs 22mm throughout.  As long as the main bits of the pipework can take the heat load, the ‘drops’ to the individual radiators can be smaller.

Think motorways around the floors and A-roads to the individual radiators.

Although as Heat Geek says, “the numbers in the chart are limits, not targets”.  So if you can go up a size it will give you more capacity in the system.

Heat Geek explains the chart really well and provides examples in the video, so it is well worth a watch.

If you look back to my 55 ℃ and 40 ℃ system diagrams earlier in the article you’ll see that we have upgraded the downstairs 15mm copper and 10mm plastic pipework to new 25mm MLCP pipework. 

And we plan to use 28mm copper from the heat pump to the cylinder and into the house.


The downstairs upgrade is required because 15mm pipe can only allow around 2.75kW down it and 10mm only 1.15kW.  And regardless of which heat loss calculation method we decided to use, those 10mm and 15mm pipes downstairs would not be big enough to carry the load.

This is the same challenge that a house full of microbore pipe has.

The eagle eyed amongst you will notice that our total upstairs heat loss comes out at 2.9kW. Which is around the edges of what is allowed for 15mm copper (2.75kW) shown on the cheat sheet. But Damon has done the calculations, pipe by pipe, length of pipe run by length of pipe run and we will be okay.

We also upgraded the 10mm plastic pipework that connected the first floor to the loft conversion with 15mm copper because this was a long run up there.

Remember, the cheat sheet it just a cheat sheet giving ballpark figures. Your engineer should do precise mass flow rate calculations to size your pipe work.

This is where getting a good engineer like a Heat Geek is worth it. They can design the system using maths! Nothing left to chance.

Undersized pipework seems to go hand in hand with poorly performing and horror story heat pumps.

As renowned heat pump system designer Simon Poskett reminds us, “Heat pumps are ‘flow machines’, so they need room and space to move the water around the heating system”. 

Having pipes that are calculated to be large enough, in conjunction with fully open radiators really helps with that movement and flow.

Zoning rooms off with third party controls and having inadequate pipework causes bottlenecks and restrictions which a heat pump does not like.

You also increase your chances of noise through pipes if they are undersized.

Saying all that, you don’t have to replace all pipework.  But if you do stick with undersized pipework, you will have to accept compromises and potentially lower performance/SCOP.  

And in some cases, putting more strain on the heat pump.

If you can afford it, both financially and in terms of the upheaval cost of replacing pipework, then you will get the best retrofit design (and SCOP) by doing the work.

To quote Craig Brookes Heating (Heating Installer of the Year for the South West 2022).

“The only way to do it right (retrofit heat pump install) is to re-pipe your system to allow DT5, insulate all pipework and size your radiators to a max 45 flow temp. Try to go direct so it gets rid of any buffer (or low loss header). Use weather compensation and just let the ASHP do its thing”

Anything less is a compromise.

What size hot water cylinder

In most cases there is no real need to worry about the size of the heat pump when it comes to producing hot water.  Most modern heat pumps sized at 5kW and above can quickly heat water. 

Remember: with a heat pump, you need a hot water cylinder. A heat pump cannot generate instantaneous water like a combi boiler. The water needs to be heated by the heat pump and then stored somewhere.

The target temperature of the stored hot water is usually low, say 45 ℃ to 55 ℃ to ensure a good COP from the heat pump. This will mean you will need to run a regular legionella cleansing cycle using the immersion to take the hot water over 60 ℃.

But storing cooler water means less usable as there will be less mixing with cold. So this plays into the next question.

What size water cylinder do you need?

The ‘rule of thumb’ for sizing a water cylinder is somewhere around 40 to 45 litres of water per day per person. 

MCS have a hot water cylinder selection document on their website that is worth a look at.

But there are many variables at play:

  • Temperature of stored water
  • Your daily usage and patterns of usage
  • Physical size limitations for a cylinder
  • Time of use tariffs
  • Solar PV diversion

As I spoke about in my Mixergy cylinder / plant room article, I went for the biggest cylinder I could fit in the space we had available. We only have one bathroom and are family of 4, but went for a 250L Mixergy.

This was because I wanted to make the best use of overnight time of use tariffs and also divert surplus solar PV to the cylinder using our Myenergi Eddi.  In essence treating the cylinder like a battery.

If I’d have had the space for a 300L cylinder, I would have gone for that.

If you have space limitations, you could also consider a heat battery like a Sunamp instead.

These heat batteries are a great option if you are tight on space, but you still need hot water capacity.

Your usage patterns could also come into focus when sizing your cylinder.

Do you want one charge of the cylinder per day (off peak) or are you happy with on demand top ups?  Ie, as soon as you start to empty the cylinder, it starts to heat again.

Although longer and consistent hot water runs will lead to better COP than short sharp bursts.

Also remember, that when the heat pump is heating water, it’s not heating the home.  It can only do one job at a time, so something else to think about too.

The only potential problematic combination could be a home with a low heat demand (so small heat pump) and high water usage (where a larger heat pump would help).

This is why the hot water requirements can feed into the sizing of the heat pump.

For those new to heating water in a cylinder, here is a chart showing how long it would take (in minutes) to heat the listed number of litres of water from 10C (average cold water temp) to 50C using different output power sizes.

ie, 3kW for an immersion element, right up to whopping 12/14/16kW heat pumps.

Time in minutes to heat water using various power output sizes

  3kW 5kW 7kW 10kW 12kW 14kW 16kW
50L 47 28 20 14 12 10 9
100L 94 56 40 28 24 20 18
150L 140 84 60 42 36 30 27
200L 187 112 80 56 48 40 36
250L 233 140 100 70 60 50 45
300L 280 168 120 84 72 60 54

You can create your own charts using this wonderful water heating calculator:

As you can see, with the 3kW immersion and those smaller heat pumps, you don’t get instant recharge of water. So cylinder size, charging regime and usage patterns could well dictate your choice of cylinder size.

For example, if you have a smaller heat pump and you want 4 showers straight after each other in the evening, then you’ve got a challenge to solve there if you can only have a small cylinder.

There’s much to think about and talk through with your installer when choosing the right sized hot water cylinder (or indeed heat battery) and heat pump for you.

Again, it is essential you consult with an expert and have these conversations.

Final Thoughts and Disclaimer

So there you go, lots to take in and think about.

I have personally found great value by typing out all these notes. I hope they have somehow helped you.

What did your calculations come out at? Hit us up on Twitter with how you got on.

Disclaimer:  All the methods listed here are for information only and I cannot take responsibility for your own use of them.

Always consult with an experienced heating engineer for your install, such as a Heat Geek.

Personally, we will be sticking to the outcome of the room by room Heat Engineer software survey and sizing our heat pump, radiators and initial flow temperatures based on that.

If the whole house demand does turn out to be lower, as suggested by the HTC calculations, we then have the bonus of having oversized radiators and we’ll be able to drop down the flow temperature even further than calculated.

But I’m happy that we are only a kW or so out using all the calculations, so that puts in the same size heat pump ballpark.

The key really is not to end up with a massively oversized heat pump. There is already so much headroom and margin for error in many of the calculations (air changes, insulation etc).

It can be very easy for an installer to err on the side of caution and go bigger than required.

If the heat pump is too big it could continually cycle on/off (because it produces more heat than it can get rid of, the return temp comes back high and it turns off), which is bad for efficiency.

Plus, if an engineer specs an oversized heat pump, this could leave to planning permission issues (needing a double fan unit) and electrical supply upgrades by pushing you towards the limit of a single phase supply.

There is a train of thought that suggests it’s better to slightly undersize a heat pump than oversize. And if required, just provide an alternative/secondary/backup heating source for those 10 days a year when it might hit -3 ℃. Especially if there are plans to add insulation to the property later down the line.

Hopefully using these easy to try DIY HTC and Heat Geek calculations can give you a rough idea of heat pump size before going ahead into the detailed room by room survey for sizing your radiators/under floor.

But remember, your installer will need to comply with MCS and do the full room by room survey if you want to be eligible for any grants like the Boiler Upgrade Scheme (BUS).

Final comment. Yes, on my installation I may have gone deep retrofit on our radiator upgrades. But the pipe work would have needed doing whether I’d have stuck with the older radiators anyway, as 15mm copper downstairs wasn’t up to the job of moving 5.5kW of heat with a heat pump.

I see all this work as an investment though. Now that all these major changes have been done, if we ever need a new heat pump sometime way in the future, then it would only be a heat pump swap out. The framework is all in place.

As I heard this week, you can class all this outlay as a “one-off transitional payment”. Setting your home up for a long life of low temperature heating. Arguably something that should have been the norm when condensing boilers first came out. But that’s a grumble for another day!

Any questions, comments or corrections please grab me via the contact form on the website.

Or catch us over on Twitter.

If you find this information useful and are thinking of joining Octopus Energy, please consider using my referral code or ‘buy me a coffee’ to help support running costs of the website.

Using this referral code will gift you £50 of free credit after signup:

If you’re considering owning part of a wind farm through Ripple Energy, get £25 of free credit (if investing more than £1000) using my Ripple referral link.

We can also be found on Twitter, please follow us @energystatsuk for daily Agile tariff pricing graphs and summaries.

Also, be sure to check out our Dashboard, Download Historical Data and OctoChargeCalc features to help you decide whether these tariffs are the right fit for you.

Note: The past performance of energy pricing is not necessarily a guide to the future.

Close Menu