Friends, around a year ago I wrote an article and made a YouTube video about using a ‘Rule of Thumb’ for estimating the size of heat pump required to replace a gas boiler in a dwelling.
The ‘Rule of Thumb’ is splendidly simple: one just divides the previous year’s gas consumption by 2,900 to give the heat pump size in kilowatts. So if a dwelling used 10,000 kWh of gas the previous year, then one would estimate that it needed a 3.4 kW heat pump. The YouTube video explaining why the rule works has been watched an astonishing 37,000 times, and many people have left comments telling me they found the rule helpful and accurate.
The basic reason the rule works is because (a) most gas consumption is spent heating homes (rather than heating hot water or food) and (b) the climate of the southern half of the UK does not vary that much. The rule of thumb uses gas consumption as an indicator of the amount heat which enters a dwelling and uses climate data – in the form of heating degree days – to estimate how cold it gets in a particular locale. You can find a detailed description here, here, here and here!
But one or two people have told me that it gave them answers they thought were quite wrong. It turned out that these people often only put their gas boilers on for an hour or two per day, and so most of the time their dwellings were unheated. Alternatively, some people – particularly with families – used a lot of hot water every day – and so this formed an unusually large fraction of their gas consumption.
So I thought it would be nice to develop something just a little more sophisticated than the ‘Rule of Thumb’ that would take account of some of these factors. I did this last summer and sent it to an academic expert for feedback. The feedback was devastating: they basically told me that everything was wrong. And despite trying to modify the spreadsheet to meet their criticism, they seemed unmollified. So, shaken, I abandoned the idea for a while.
But recently I have been thinking about the idea again and decided that in fact I thought the spreadsheet was useful after all, and that it could also help with one other problem: sizing of radiators.
The reason I think this endeavour is important is that people who are thinking about installing heat pumps have faced a campaign by the fossil fuel industry and their knowing (and unknowing) shills, a campaign designed to instil fear, uncertainty and doubt (FUD). Every year of delay in installing heat pumps keeps the profits of fossil fuel companies healthy, and impoverishes the world in which our children will have to live.
This is not to say that there are not legitimate questions and uncertainties about installing a heat pump. So this spreadsheet is a transparent tool that can help people make rational choices and – I hope – help them to overcome the FUD.
- You can download the spreadsheet here:Link
- Spreadsheet updated to version 6.01 on 3/3/23
I have tried to make the spreadsheet Good For Nothing™ 🙂 . But mistakes will have slipped through: if you find one, please accept my apology in advance and let me know in the comments.
Spreadsheets Galore
The ‘spreadsheet’ is actually six spreadsheets linked together in an Excel™ Workbook. Each Spreadsheet has its own ‘tab’. Six spreadsheets may sound daunting, but really this could all be on one spreadsheet. Using several sheets actually makes things simpler.
Click on image for a larger version. The introductory ‘tab’ of the Excel™ Workbook showing the other 6 tabs. Users are recommended to save the downloaded copy and experiment with a ‘working copy’.
- The first spreadsheet helps people estimate the average temperature in their dwelling, and also the maximum temperature they like.
- The second spreadsheet helps people estimate the amount of hot water they use.
- The third spreadsheet uses the ideas behind the Rule of Thumb, but modified to take account of the estimates on the first two spreadsheets. It suggests a likely required size of heat pump and a few other building parameters that specialists might find interesting.
- The fourth spreadsheet allows people to see how the area of radiators and the type of radiators affects how hot the water flowing through the radiators needs to be in order to keep their home at the maximum temperature they desire.
- The fifth spreadsheet allows people to make more detailed calculations based on the number, size and type of radiators in their own dwelling.
- Finally, the sixth spreadsheet summarises the results from the previous spreadsheets and estimates the likely savings in cost and carbon dioxide emissions.
Let me show you each spreadsheet works in a little more detail.
Sheet 1: Household Temperature
Click on image for a larger version. Spreadsheet designed to allow a user to indicate the temperature changes in their home throughout a typical winter day.
Click on image for a larger version. As above, but showing a different temperature profile.
On this tab of the workbook, one can specify how the temperature varies inside a dwelling on a typical winter day. There are four times periods and each one can be set to one of three user-chosen temperatures.
The spreadsheet then calculates:
- The average temperature in the dwelling which is useful for calculating the average heat loss and hence energy consumption.
- The maximum temperature required which determines the required power of a heat pump able to heat the dwelling.
Sheet 2: Domestic Hot Water
Click on image for a larger version. Do you know how much hot water your dwelling uses each day.
I have been told that – in the absence of any other information – a good guess for the amount of gas used to heat hot water in a household is 3 kWh per person per day. This tab uses this figure to estimate how much of the annual gas usage is for domestic hot water.
If a user somehow has a better estimate, they can use their own estimate instead.
Sheet 3: Main Calculation
Click on image for a larger version. This ‘tab’ carries out the main heat pump size calculation.
This tab carries out the same calculation as the Rule of Thumb but now with a little more information about a particular user’s dwelling. It incorporates the data from the first two tabs on average and maximum temperatures and domestic hot water usage. It asks the user for the annual gas consumption and their approximate location (within around 100 miles). The location is used to estimate how cold the weather is likely to have been based on analysis of the heating degree-day records from 21 locations in the UK and Ireland.
Click on image for a larger version. This tab carries out the main heat pump size calculation.
The spreadsheet then estimates several parameters that characterise the level of thermal insulation of the dwelling and – most importantly from the perspective of this article – the heat pump size required for the dwelling.
Sheet 4: Radiators
Click on image for a larger version. This tab allows users to see how the area of radiators, and the type of radiators affect the performance of the heating system.
This tab allows users to see how – in general – the area of radiators, and the type of radiators affects the performance of the heating system. First one sets a maximum flow temperature for the system – this is the temperature of the hot water as it enters the radiators.
Heat pumps typically use weather compensation, which means that when the weather is cold, the heat pump increases the temperature of the water flowing in the radiators. For a heat pump the maximum flow temperature required in the coldest weather should ideally be below 50 °C.
Click on image for a larger version. This tab allows users to see how the area of radiators, and the type of radiators affect the performance of the heating system.
The table above shows – for the heat pump size calculated on the previous tab – what combinations of total radiator area and types of radiator will be able to heat the dwelling adequately.
For heat pumps to work at their very best, the temperature of the water flowing in the radiators should be as low as possible while still allowing the dwelling to be adequately heated.
In the example above the heat pump needs to transfer 5,296 watts of heating power to the dwelling.
- The table shows that this would require 9 square metres of single-panel/single-fin (Type 11) radiators, but the same heating could be done with just 5 square metres of double-panel/double-fin (Type 22) radiators.
- Alternatively one might use 9 square metres of double-panel/double-fin (Type 22) radiators because this would require a flow temperature in the radiators 39. 8°C rather than 49.2 °C – and this reduced flow temperature would result in increased heat pump efficiency, and lower running costs.
Sheet 5: More Radiators
Click on image for a larger version. This tab allows users to see how the number, size and type of radiators in their dwelling affect the performance of the heating system.
The previous tab allowed users to see in general terms how the area of radiators, and the type of radiators affect the performance of the heating system. On this tab a user can input the size (width and height) and type of their existing radiators and see whether – for the flow temperature set on the previous tab – they can release enough heat into their dwelling.
Click on image for a larger version. This TAB allows users to see how the area of radiators, and the type of radiators affect the performance of the heating system.
By putting in data on their existing radiators – the radiator type is input via a drop-down menu – the heating power of each radiator is calculated at the maximum allowed flow temperature. The heating power of each radiator is then summed up to see if the assemblage of radiators in the dwelling is capable of providing enough heating power to keep the dwelling warm on a cold day. This is shown as a percentage on a bar chart.
If a figure of 100% cannot be reached with existing radiators, then users can see whether 100% can be achieved by either adding radiators, or replacing radiators with larger ones, or radiators with more panels and fins.
Sheet 6: Summary
Click on image for a larger version. This tab summarises the results from the previous tabs and compares the cost and carbon dioxide emissions of systems using a gas boiler or alternatively, a heat pump.
Nearly finished! This summary tab collects together the conclusions from the previous spreadsheets. If a user enters the cost of their electricity and gas, the spreadsheet will then estimate the likely running costs of a gas boiler and a comparable heat pump.
The annual costs of the gas installation are estimated based on the users estimate of their own gas consumption. The running costs of the heat pump installation are based on an estimated seasonal coefficient of performance (SCOP).
The coefficient of performance (COP) of a heat pump is a measure of the efficiency of a heat pump measured over a period of typically an hour, a day or a week. In mild weather, the COP will be high (perhaps 4) and in cold weather the COP will be low (perhaps 2.5). SCOP measures the efficiency of a heat pump averaged over a whole year.
If a user experiments with different flow temperatures they will find that the lower the maximum flow temperature they plan for, the higher the achievable SCOP and the lower will be their running costs. Typically users will find that with the relative costs of electricity and gas as they are now (April 2023) at a ratio of roughly 3 to 1, a heat pump installation will commonly be a little bit cheaper to run than a gas boiler, but the difference is not very large compared with the capital cost of the installation.
Click on image for a larger version. This tab summarises the results from the previous tabs and compares the cost and carbon dioxide emissions of systems using a gas boiler or alternatively, a heat pump.
And finally – and this is the point of the entire endeavour – the spreadsheet makes a comparison of the carbon dioxide emissions from a dwelling heated either with a heat pump or a gas boiler. It is here that the entire point of running a heat pump becomes clear: carbon dioxide emissions from a heat pump installation are generally around 75% lower than an equivalent gas boiler. And that’s why this matters.
Click on image for a larger version. Graph showing the annual emissions of carbon dioxide from a gas boiler and an equivalent heat pump installations.
Protons for Breakfast 
April 2, 2023 at 8:02 pm |
Hi, this looks excellent but one query. I have a number of double panel radiators with no fins. Are these Type 21 or should there be a Type 20?
April 2, 2023 at 10:38 pm |
They are type 20.
I am guessing they are quite old because nowadays if someone was installing a 2 panel radiator they would almost always install a 22.
Good luck with the spreadsheet!
M
April 3, 2023 at 6:48 am |
This is great but I have oil fired heating / water and wet underfloor heating, no radiators.
April 3, 2023 at 9:20 am |
Mike
Good Morning. I nearly extended the spreadsheet to include UFH but I wasn’t confident enough of the figures.
The heat pump size is easy. The calorific value of heating oil per litre is almost exactly 10 times the calorific value of gas per kWh. So if you know your annual usage of heating oil in litres.multiply this number by 10 to get the number of kWh to enter on Line 1 of the heat pump calculation (Cell D10). I did include heating oil in the original Rule of Thumb article.
Regarding UFH, it is like a large very inefficient radiator I didn’t include an option for it because I wasn’t quite sure how to treat it. But if you enter the area of the UFH and change the radiator constant on Line 6 of Tab 4 or on Column F of Tab 5 to around 10 W/m^2/°C this should be ‘about’ right.
Best wishes
Michael
April 3, 2023 at 8:48 am |
Hey – great spreadsheet.
I’ve noticed one very minor error in the calculation of the electricity cost in the summary worksheet (cells D29 and D30). This uses the unit cost and the standing charge the wrong way round.
As an aside, as (pretty much) everyone will already have electricity they won’t pay any additional standing charge for fitting a heat pump, but might no longer have to pay a standing charge for gas if they don’t have a gas cooker. That makes the savings potentially greater than they first look.
Thanks
Gary
April 3, 2023 at 9:26 am |
Gary: Thank you both for the correction which I have implement in 6.01 and for the suggestion which I will think how to implement.
Best Wishes
April 8, 2023 at 8:22 am |
Gas standing charge removal requires that the gas supply is disconnected by the supplier.
April 8, 2023 at 8:29 am
Indeed: Happily now completed!
Best wishes
Michael
April 3, 2023 at 10:15 am |
What a fantastic spreadsheet, I do love a good spreadsheet! Unfortunately I have a slight issue. We have just moved into a house which has various different column type radiators – some 2, 3, 4 and even 5 column rads.
How can I find/calculate the W/m^2/°C? I have had a look on the retail websites but can’t seem to find this figure, only btu(T50). I can convert that to watts and then find the W/m^2. However, how do I then get to W/m^2/°C?
Is there a calculation I can do based on the btu (or watts) to calculate this constant?
Again, a fantastic and comprehensive spreadsheet, I love it!
April 3, 2023 at 10:59 am |
Jimbob, thank you of your kind words.
Well done for converting BTUs (actually BTUs/hour) to watts! The final step from W/m^2 to W/m^2/°C is to divide by the difference between the average radiator temperature, which for BTU(T50) is 50 °C above room temperature.
So 1000 BTU/hr at Delta T 50 is 290 watts with the average radiator temperature 50 °C above room temperature.
So that corresponds to 290/50 = 5.8 W/°C. Then divide by the area to get W/m^2/°C.
One feature you may see is that designer radiators may not be all that good in terms of W/m^2/°C. I think making radiators taller is not as effective as making them wider. So those per unit area, those lovely 50 cm wide 2 m high radiators are not as effective as an equivalent 50 cm high 2 m wide radiator. I think it’s to do with the fact that as air rises over the surface of the radiator, the heat transfer to the air is practically completed in the first 50 cm or so.
Does that make sense?
Once you have a figure, you can just type it in the spreadsheet wherever the ‘radiator constant’ is shown.
Best wishes
Michael
April 3, 2023 at 12:19 pm
Thank you for that, it makes so much sense when you think about it!
Anyway, the comment about the tall radiators makes sense as well, however all of ours are 600mm high so the calculations should be OK.
Just realised that you calculate need to input historical temperature data not desired temperature data as this will effect the ‘heat transfer coefficient’. Sorry if I am pointing out something that you have mentioned.
Well, according to the calculations we are at 100% with 50 degrees. I guess the insulation journey must continue so that we can reduce the energy requirements of the house and the insulation figure of merit, which is currently 74! Old houses are problematic!
Again, Thank you for your work on the spreadsheet and thank you for sharing it.
April 3, 2023 at 12:31 pm
Jimbob,
If you are at 100% with 50 °C flow temperature, and Figure of Merit of 74 kWh/m^2/year then I think you are about as ready for a heat pump as you are likely to get. Remember that 50 °C flow is for the coldest days of winter.
All these figures are estimates and have uncertainty associated with them and in the end what you get will be somewhat different.
A heat pump installer will do a heat loss survey, estimating heat losses room by room. This is very likely to result in an overestimate of the required heat pump size, but if it’s done well it won’t be by too much. This will tell you whether you have 100% of the required heating in each room.
Anyway: can I wish good luck with your endeavours!
Michael
April 3, 2023 at 11:28 am |
Perfect timing for us! Thank you Michael. We are currently about halfway towards a heat pump, but taking it steady to see what effect the changes we’ve already made will have (EWI & triple glazing on the leakiest part of the house – the rear – & double glazing in the front sashes). We’re in a conservation area, with tight passageways on either side of the house, so realistically that’s as much EWI as we can do. The next stage is to replace about half the radiators. And on we go…
Just a note to say that your work has been hugely influential in us getting to this stage. Thank you!
Jon
April 3, 2023 at 11:38 am |
Jon,
Thank you for your kind words which I find very moving. I think you are absolutely right to take things step by step and I wish you all the best in your endeavours.
Michael
April 3, 2023 at 12:30 pm |
Hi
Thanks for the spreadsheet, and the first thing that I learnt about my
1950s detached house with cavity wall insulation and 300mm loft
insulation is that it uses more kwh heating than it should so I am going
to look urgently at insulating the suspended (ground) floor as the
obvious place for heat to be escaping
Regards
Stuart
April 3, 2023 at 12:38 pm |
Stuart,
Good Afternoon. First of all, I would urge you not to anything urgently. Assuming you are in the northern hemisphere, it’s spring and taking a few weeks to think things over will still give you plenty of time to prepare for next winter.
Insulating the floor is a great idea, but can be very disruptive. Be sure to do all the underfloor work you want at the same time: think about underfloor heating/ radiator replacements.
If you are considering such expensive work, it’s probably worth spending a couple of hundred pounds to get a heat loss survey done (not an EPC! – a proper survey). This will guide you as to where your investment can have the most benefit most quickly.
Good luck with your endeavours.
Michael
April 3, 2023 at 3:37 pm |
In North America, most of the heating system is forced air. It would be nice to have this spreadsheet to include a forced air system.
April 3, 2023 at 4:22 pm |
Kim W, yes, this is very much a UK focussed spreadsheet.
If a US dwelling is heated by gas, one can still use the annual gas consumption to estimate the amount of heating required (i.e. kWh/year). If you know the degree-day value for your location you can then estimate the required heat pump size.
This is the heat output size and should be the same for air-to-air as for air-to-water pumps.
You won’t need to worry about radiators, but I don’t know an easy way to estimate the seasonal coefficient of performance.
April 3, 2023 at 6:50 pm |
Re your comparison in the final paragraph:
I am an OVO customer and apparently I have a greater than average carbon impact at 3.0 tonnes versus an average OVO home at 1.6 tonnes. Both these impacts are less than your 3.45 figure. Have you used a worst case versus best case scenario?
April 3, 2023 at 7:08 pm |
John, Good Evening. I am not entirely sure what you are comparing with what.
3.45 tonnes of CO2 emissions in the final graph is just a figure for a household currently using 15,000 kWh of gas a year. This is a very typical UK figure and corresponds to what this house used to consume before insulation.
So your figure of 3.0 tonnes comes because presumably you are using about 13,000 kWh of gas per year?
I’m afraid I don’t know where the average figure of 1.6 tonnes comes from: it seems very low. UK average household emissions are around 3.4 tonnes per year – typically 0.7 tonnes from electricity and around 2.7 tonnes from gas.
Does that make sense?
Best wishes
Michael
April 4, 2023 at 5:45 am
Yes, makes sense to me. However, the 1.6 tonnes average comes from OVO comparison of my house with their claim for average UK house. (from the billing system)
April 4, 2023 at 8:24 am |
Hi Michael,
You are doing some very interesting work and this latest spreadsheet is very familiar to me as it’s very similar to a spreadsheet I have used from Freedom Heatpumps. Their’s is a full heat loss calculator split into similar tabs as you have done with very similar treatment of radiators with inside/outside temperature differentials and running cost, CO2 etc.
Where it goes into more detail is a room by room heat loss calculation wth entered temperature, wall perimeter, floor area, ventilation, ceiling height, window dimensions, with U values chosen from drop downs giving heat loss percentage of each loss added to room and a house total.
This, apart from desired room temperature variation, I’ve found the most potential for error with one of the biggest unknowns is the ventilation heat loss – which is basically a measure of air exchange per hour with figures of 1.5 being suggested for hallways and kitchens but plausibly this could go down to 0.2 if the draughts have been excluded well. I found at first the kWh/year figure for the house was way over my known gas boiler consumption and only became realistic when I reduced the ventilation figures. This air exchange is difficult to measure and I can see some homes having huge variations. I know I’ve sealed some rooms well because the doors don’t slam because of the cushion of air that whistles between door and frame as it closes.
Anyway, that’s why your “rule of thumb” is a good idea because it anchors the heat loss to reality whereas any calculation of heat loss, although superbly useful, is prone to significant error from guessed inputs.
I notice you have an input of “insulation figure of merit” for the whole house but also an entry for the old gas boiler energy consumption and wonder how they are reconciled?
April 4, 2023 at 9:43 am |
John H: Good Morning.
Your comment is exactly the reason why I created the spreadsheet and the ‘Rule of Thumb’. The standard calculation of heat loss is fine *IF* one knows all the parameters such as heat loss and underfloor temperatures. This spreadsheet is really a complement to the standard calculation working backwards from the known gas consumption – a figure which is often known very well.
Regarding the rate of air changes per hour, I wrote about how to estimate this here:
The technique involves measuring the level of CO2 in the house with a CO2 meter – available from amazon for £50 and upwards.
Regarding the insulation figure of merit, larger dwellings typically have larger heating requirements. So to estimate the leakiness of the dwelling one divides the estimated annual space heating required by the floor area: it is a very rough measure.
I am not sure about your last question, but unless you undertake additional insulation, the heating requirement does not change when changing heating method from a gas boiler to a heat pump.
In any case: best wishes: this whole endeavour never seems to be quite as straightforward as it ought to be.
M
April 5, 2023 at 6:45 am
Thanks for the reply. To clarify, what I meant about the last comment was that the parameters, heat energy input (gas boiler kWh), delta T (internal-external temperature differential), thermal resistance (insulation figure of merit) are all effectively linked by an energy equation analogous to ohms law I=V/R or kW = delta T/Thermal resistance. Although this may not be obvious with a complicated spreadsheet.
For the Freedom heat loss calculation spreadsheet I used, having entered desired room temperature with the known external temperature (setting delta T) and having entered insulation U values and areas (so defining thermal resistance) the spreadsheet told me what the gas boiler kWh/year figure was.
But this didn’t match reality – the gas boiler figure was 25% too big and since delta T was also a known measured quantity from a set thermostat and stated outside temperature, then obviously the effective thermal resistance was wrong, so as the areas and dimensions were a fixed definition I had to make sensible alterations to the entrees U values in order to make the thermal equation balance and match the resulting gas boiler kWh/yr result to reality. That’s when I discovered the air exchange per hour figure made a big difference and the presumed values I had set were wrong.
I haven’t tried your spreadsheet but all the factors; boiler kWh, thermal resistance (insulation figure of merit and floor area) and delta T from set room temperature during winter, seem to be inputs all freely adjustable. So I wondered how the thermal equation within the spreadsheet was being balanced or if some factor I’ve not spotted was effectively the equation output? Thanks.All very interesting stuff.
April 4, 2023 at 1:18 pm |
Hi,
Thank you for all the fascinating and useful ideas and work you are sharing.
“I did this last summer and sent it to an academic expert for feedback. The feedback was devastating: they basically told me that everything was wrong.”
I would be very interested to understand and explore the issues in more detail. Given your career and evident attention to detail I am surprised that “everything should be wrong”!
Would you be willing to share these ideas and why they were so wrong 🙂 either publicly or privately?
Thanks,
Patrick
April 4, 2023 at 8:00 pm |
Patrick, Good Evening, and thank you for your kind words.
You asked “Would you be willing to share these ideas and why they were so wrong 🙂 either publicly or privately?”
Not really. I don’t think that would help anybody. I mentioned it just because I had been quite enthusiastic back then, and the experience had been really depressing for me.
Writing this blog involves a balance between the personal and the technical. I mention some aspects of my life because all the blog readers whom I care about are human and I feel this keeps things ‘real’ rather getting lost in technical details. But the blog is not really about me or my travails.
In any case, thank you for your kind thoughts.
Michael
April 5, 2023 at 2:02 pm
Michael,
Thank you very much for your email.
I absolutely did not mean to pry or enquire about the personal side of this story. Reading this blog again and following what you in your email, I realize that the communication with this academic expert was very upsetting and I am sorry to have touched upon this. Given the wonderful tone of your blogs and your passion to contribute something positive this makes me very sad.
I have been thinking about how to approach this problem for some time now. I was and remain dubious about the survey approach which I see as fraught with uncertainties and likely a good sprinkling of guesswork in real/widespread application.
I was contemplating going down the experimental route but could see many pitfalls and problems in general use. Thankfully I came across your blog and have been fascinated and marvel at the approach you have explored and documented so clearly 😊. Considering the bigger picture I hope that some of the big players in this have read your blogs and are thinking hard about your approach.
My interest here is to see how to refine your approach and investigate where/when/why it is not giving a reliable answer. I have a background in engineering and physics and two close friends who are professional physicists and I am itching to present them with this challenge! Should you reconsider at any point, I can assure you my friends are extremely kind and positive and would never say or write anything negative.
Please keep blogging 😊
Patrick
April 6, 2023 at 1:58 pm
Patrick
Thank you for your kind words.
And yes, I would love to collaborate to find out when the approach gives poor results. I get very anxious about this!
Drop me a line at house@depodesta.net and perhaps we can arrange a time to chat.
Best wishes
Michael
April 8, 2023 at 8:29 am |
As a designer and installer of heat pumps, you should note that heat pumps are sized by the designer as it the heating and hot water requirement that determines the size of heat pump required.
This should be done based on the building fabric, with a counter check on existing consumption where available.
These spreadsheets can be useful and can help guard against the cowboys. Any decent designer will and should provide you with the calculation they have made and explain any differential.
April 8, 2023 at 8:36 am |
Good Morning.
If installers routinely used a counter check on consumption, then this spreadsheet would be redundant. But there is a problem with heat pump sizing as implemented on the ground in the UK that has led to many instances of significant oversizing by a factor of 2 or 2.5. This is due to poor heat loss assessment or over-pessimistic assumptions about ACPH or underfloor temperatures.
But I agree there is no substitute for a heat loss survey by an experienced engineer.
Best wishes
Michael
April 11, 2023 at 8:24 pm |
What difference does microbore radiator pipes make to these calculations? Octopus declined to consider my mid-90s detached 4 bed home for that one reason.
April 12, 2023 at 9:41 am |
Dominic, Good Morning.
First of all, I don’t know details of your home, but I can outline the generic problem.
Suppose that your home requires 5 kW of heating power on average to keep it warm during a cold night.
The heating power delivered is proportional to (a) the temperature drop between water leaving the boiler/heat pump and water returning to the boiler/heat pump and (b) the rate (litres per hour) at which water flows through the pipes and radiators.
For a heat pump to operate efficiently, it needs to lower the temperature of the water leaving the heat pump to 50 °C rather than 70 °C as might be the case for a boiler. To compensate for this, it therefore needs to increase the rate at which water flows through the pipes and radiators. This means the water pumps need to work harder.
The ease with which water flows through a pipe varies very strongly with its diameter. If the pipe internal diameter is 8 mm then this 12 times harder to push water through than the same length of 15 mm diameter pipework. My guess is that this is beyond the capacity of the heat pump to pump water through your radiators.
If that’s not clear: do get back to me.
Michael
April 26, 2023 at 9:53 pm |
Hi – thank you for a great explain and spreadsheet. This may be a dumb question. In the Summary tab why do the Vaillant heat pumps have a higher output than the model name. For example the 5kw model has an output of 6.1 at 45 degree flow and -3 outside design temperature. If the heat pump required calculation returned 5.3 why would I not pick the 5kw model vs the 7kw model?
April 26, 2023 at 11:28 pm |
Adam, Good Evening.
Your comment highlights a feature of the problem we all face.
Heat pumps have a so-called ‘Name Plate’ capacity. This is just a name. Their capability to deliver heat to a home varies as (a) the outside temperature varies and (b) the required hot water temperature varies. No one has any control over what a manufacturer names their heat pump.
There are some standards such as 7/35: this is the capacity of a pump to deliver heat when the outside temperature is 7 °C and the hot water temperature required is 35 °C. Typically the ability to deliver heat to a home decreases as outside temperature falls, and the temperature of the hot water required increases. So a heat pump which delivers 5 kW at 7/35 might only deliver (say) 4.5 kW when the outside temperature is 0 °C and the hot water temperature required is 40 °C.
Different heat pump manufacturers have different attitudes towards the nameplate capacity of their heat pumps. Some manufacturers say their heat pump delivers 5 kW at 7/35 and give it a name plate capacity of “5 kW”. Other manufacturers – such as Vaillant – have heat pumps with a name plate capacity of 5 kW which deliver more than 5 kW under the 7/35 condition.
One needs to look carefully at the heat pump specification sheet to understand the ability to deliver heat at the lowest likely exterior temperature.
Does that make sense?
Best wishes
Michael
May 1, 2023 at 8:15 pm
Thanks Michael – got it. Thank you for the really clear explanation. Adam
May 1, 2023 at 11:12 pm
Adam: You’re welcome. Thank you for taking the time to let me know: I appreciate it. M
June 5, 2023 at 8:35 pm |
is possible to update your excel file with this info from Heat Geeks
This Insulation Secret Will BLOW YOUR MIND
June 5, 2023 at 9:18 pm |
Is possible to add water volume in radiators and to calculate minimal water in the system to eliminate puffer/buffer ?
https://global.purmo.com/en/tools-and-services/item-selector?heatoutput=1832
The Temperature drop across each radiator: A typical value is 10 °C …………..for HP dt is 5C.
September 29, 2023 at 9:05 pm |
Great tool. Can you add something which computes the flow rate required to achieve the heat delivery for a given pipe diameter. The spreadsheet tells me my home is currently suitable for a heat pump, but I have plastic microbore and I am worried the required flow rate will be too high
September 29, 2023 at 11:08 pm |
Gary, good evening.
AS it happens the original version of the spreadsheet did include that calculation but I took it out because I worried that people would get confused.
The formula is that mass flow in litres/sec (numerically the same as kg/s) is equal to
Heat Pump Power (in watts) divided by {Specific heat Capacity of Water x Delta T}
Delta T is the difference between the flow temperature and the return temperature and is normally targeted to 5 °C for heat pumps.
Here’s an example calculation: Let me know if I can explain more. M
Power 5 kW
Power 5000 W
Specific Heat Capacity 4,200 J/°C/kg
Delta T 5 °C
Mass flow 0.2381 kg/s
Volume flow 0.2381 l/s
Volume flow 14.3 l/min
Volume flow 857.1 l/hour
Volume flow 0.857 m^3/hour
September 30, 2023 at 5:26 am
Thanks. Do you also have information on what the recommended max flow rates are for different diameter radiator pipes. In my house I believe it is 22 mm copper pipes from the boiler around the house but it ends up with 10 mm plastic pipes into the radiators except the towel rails which are 15 mm
September 30, 2023 at 11:43 am
Gary,
Sorry: that one is beyond me. A HeatGeek installer (or similar) should know: it all depends on the pressure characteristics of the hydraulic pump in the heat pump. If the main pipes are 22 mm, you may be in luck.
Alternatives are to change the plumbing – and take the opportunity to upgrade any radiators – or to add a Low Loss Header, which uses a separate pump to drive the water around the radiators.
In any case, good luck with your endeavours.
Michael
October 10, 2023 at 8:40 am |
Not sure if I am doing something incorrect but the higher the temperatures or more people that use hot water I enter, the heat pump size required gets lower. and based on the calculations a heat pump would cost more to run?
October 10, 2023 at 9:07 am |
Michael, Good Morning.
The spreadsheet works by estimating the heat transfer coefficient (HTC) for a dwelling.
It does this by assuming that the heat you put into the dwelling from a boiler leaks out, and it makes assumptions about how cold it was locally (based on heating degree data)
1. If all the gas used goes on heating (no hot water, no one in the swelling) then the HTC will take on a particular value.
2. If some of the gas was used for hot water, then less was used to heat the dwelling – so the dwelling required less heating to maintain its internal temperature so its HTC must be LOWER i.e. better insulated.
Does that make sense?
Currently with electricity about 4 x more expensive than gas, a heat pump might well cost more to run than a gas boiler.
All the best
Michael
November 1, 2023 at 10:37 am |
Michael,
I’m a new fan of your blogs. I arrived here by following an interest in EV -> Solar PV -> Tesla Powerwall -> and now Heat pumps. I’ve just paid a deposit to Octopus and am doing all I can in advance to understand and check what they will quote me for soon.
To the point; I have just started using your radiator sizing spreadsheet but have a problem.
On sheet 5, Radiators, I have eleven but when I overwrite 9, columns C,D,F and I and others return #NAME? in may cells. Is there an error in the sheet? or am I doing something wrong?
I have looked at some of the cell formulas and can do some work-arounds but can’t see the problem. Could you take a look please in case a gremlin has crept in.
Love your Youtube videos by the way they are inspiring me to do my bit for the planet and our kids.
Peter J.
November 1, 2023 at 12:12 pm |
Peter,
A fan?! I am blushing.
Regarding teh spreadsheet, it is only set up for 10 radiators. To increase the number go to CELL D17 which contains the number “1”. Look at the code for that cell and you will see that it is a SEQUENCE statement. To add (say) two more radiators change it from:
=SEQUENCE(10,1)
to
=SEQUENCE(12,1)
For this to work, the cells in column D must all be initially empty. You should see a list running from 1 to 12. Alternatively, you can just type in all the numbers you want starting at 1 and going down.
To get the cells in columns E to I working, select the cells in the bottom row (E25 to I25) and copy-drag them downwards to duplicate the formulas.
Does that make sense? If it doesn’t work, please drop me a line again.
All the best
Michael
November 1, 2023 at 2:19 pm
Alternatively: use this spreadsheet
https://protonsforbreakfast.files.wordpress.com/2023/08/radiator-worksheet-1.xlsx
that I wrote for this article
M
January 3, 2024 at 8:04 pm |
Great work! Love the spreadsheet although it would be really nice if it included underfloor heating rather than just radiators. Thanks for sharing! 🙂
January 3, 2024 at 8:20 pm |
Dear Myatix, Good Evening.
Regarding Underfloor Heating, the problem is that different construction techniques give quite different ‘Raditaor Constants’. I did look at this later in this article about Hydronics, and published a spreadsheet in which you can use ‘custom’ radiator constants. This is illustrated below:
And you you can find the link to the spreadsheet someway down this article
Underfloor heating is a ‘bad’ radiator – probably similar to or worse than the heat transfer from a single panel radiator without fins.
Good luck with your experiments
Michael
January 4, 2024 at 2:54 pm |
Hi Michael, this is a great spreadsheet. I noticed that as the number of heating days increases, the size of the heat pump decreases. For example Aberdeen is lower than Cardiff. Can you explain why this would be?
January 9, 2024 at 2:23 pm |
Maire, Good afternoon. Please forgive my delay in replying. For some reason I seem to be busy – but in honesty I don’t know how. Anyway, allow me to explain.
The key idea is the Heat Transfer Coefficient (HTC) for a dwelling. This figure characterises the heating power required to raise the dwelling by 1 °C. Suppose a house in a cold location (say Aberdeen) uses 15,000 kWh/year of gas to maintain a temperature of 20 °C. Now imagine a different house in a warmer location (say Cardiff) that also uses 15,000 kWh/year of gas to maintain a temperature of 20 °C.
Over the year the heating demand on the house in Cardiff will be less than the heating demand for the house in Aberdeen – but they both use the same amount of gas. So this implies that a house using 15,000 kWh of gas/year in Cardiff has a higher HTC than a house using 15,000 kWh of gas/year in Aberdeen.
So if one wants to warm these dwellings to 20 °C above the outside temperature, the house in Cardiff will require more heating power.
So the resolution of the apparent anomaly is that a house in Cardiff using 15,000. kWh/year is less thermally efficient than house in Aberdeen that uses 15,000 kWh/year.
Does that make sense? I do hope so.
Best wishes
M
February 8, 2024 at 1:23 pm |
Hello Michael!
Could you explain the table with cities listed on sheet 3? What does column 20 mean, also the column Slope? Could you elaborate please?
I would like to adapt that table for location Zagreb, Croatia. How would data in the table change for that location?
Also what is the meaning of insulation of merit graph and the little table with ” for plotting”, ”passivhaus” etc.?
The table itself is amazing!
February 8, 2024 at 3:52 pm |
Pablo, Good Afternoon.
To understand the table on Sheet 3, you need to look at the calculation carried out on Sheet 7 – labelled “Reference UK HDD data”
Sheet 7 contains the same calculation carried out for different UK locations. I downloaded heating degree data for these locations for three different internal temperatures.
HDD15.5 corresponds to roughly 19 °C internal temperature,
HDD16.5 corresponds to roughly 20 °C internal temperature,
HDD17.5 corresponds to roughly 21 °C internal temperature,
I downloaded 3 years of data and averaged them to find the average HDD’s /year for each internal temperature.
As the internal temperature rises – the number of HDDs rises too – i.e. it requires more energy to heat. To estimate the number of HDDs for any chosen internal temperature (e.g. 21.5 °C) I calculates the SLOPE i.e. how many extra HDDs are there for each extra one °C rise in internal temperature. Using this I can estimate heating demand for any location and internal temperature by using the HDD’s for 20 °C and the slope for that location.
So for example, consider the UK location of Newquay which is very mild.
For HDD 15.5 (internal temperature 19 °C) the average number of HDDs is 1594 °C-days over the previous 3 years)
For HDD 16.6 (internal temperature 20 °C) the average number of HDDs is 1902 °C-days over the previous 3 years)
(this is 1902 – 1594) = 308 °C days extra compared with 19 °C)
For HDD 17.5 (internal temperature 21 °C) the average number of HDDs is 2231 °C-days over the previous 3 years)
(this is 2231 – 1902) = 328 °C days extra.)
So if you look at cell E13 on Sheet 87, you can see the slope function is estimated midway between these two figure at 318 °C. This is unlikely to lead to large errors as long the extrapolation to higher or lower room temperatures is not too extreme.
Back on Sheet 3, the right-hand table (columns H to K) is just copied from Sheet 7, The column labelled 20 is the number of HDDs for 20 °C and this is used with the SLOPE to estimate the number of HDDs estimate for any arbitrarily chosen temperature.
February 8, 2024 at 4:01 pm |
To adapt the spreadsheet for application in Croatia, I would
(a) On Sheet 7 replace data for a location in the UK – e.g Newquay with data for (say) Zagreb.
Download the data from https://www.degreedays.net for Zagreb
Select 3 years of weekly data and average them to find the average number of HDDs for HDD 15.5, 16.5 and 17.5: It’s all free!
If you cut and paste the data in to columns H I and J on Sheet 7 then it should calculate the annual averages for you,
You will then need to work out the slope. You can then put the numbers into the Tables on Sheet 7 and Sheet 3.
Then I think you are done.
February 8, 2024 at 4:05 pm |
There is a figure of merit for houses in the UK about how much heating (kWh) should be required per year per square metre of occupied space in a dwelling.
In the UK, a superbly insulated Passivhaus much have less than 15 kWh/year/square metre. It’s barely possible to do much better. A very thorough retrofit (Enerphit) should have less than 25 kWh/year/square metre. My own home is now somewhere around 50 kWh/year/square metre. And most UK homes are ~ 100 – 150 kWh/year/square metre.
In Zagreb – with its colder climate – these figures will be higher.
February 23, 2024 at 2:42 pm |
[…] was that you say? You want even more data?! Ok, go back over to the Protons for Breakfast blog and there’s a spreadsheet called HTC Estimator. This spreadsheet has loads of tabs so you’re […]
March 30, 2024 at 10:51 am |
Hi Michael,
Congrats for this great work! I live in Belgium near Brussels and have filled the worksheet with my data.
One thing seems counterintuitive to me though: the Temperature Drop Across Each Radiator set @ 5°C provides more KWh than when @10°C. Is that right? Do I miss something?
Thanks already for your response.
Yvan
March 30, 2024 at 8:43 pm |
Yvan, Good Evening from the UK.
The heat output from a radiator depends only on the difference between
the average temperature of the radiator and
the room temperature.
This temperature determines the radiative power from the front surface of the radiator and the convective heat transfer from the vertical surfaces of the radiator.
If the input flow is at (say) 50 °C and DeltaT is 5 °C, then the return flow will be at 45 °C and the average radiator temperature will be close to 47.5 °C.
If the input flow is at (say) 50 °C and DeltaT is 10 °C, then the return flow will be at 40 °C and the average radiator temperature will be close to 45 °C.
If teh room temperature is 20 °C then the heating power in the room will be proportional to either (47.5 °C – 20 °C = 27.5 °C) or (45.0 °C – 20 °C = 25.0 °C) i.e. about 10 % different.
Does that make sense? You can read about the physics of radiators here:
Best wishes
Michael