How big a heat pump do you need? A simple Rule of Thumb.

As ever more people become interested in installing Air Source Heat Pumps, one of the continuing anxieties is around the size of heat pump required.

In this video I explain that a simple rule of thumb can give an answer within about 10% based just on the previous year's gas usage (in kWh).

The rule?

Heat Pump Size (in kW) = [Annual Gas Usage (in kWh)]/ 2900

This applies to the UK south of Manchester. It's very simple and in the video I explain how it works.

You can download the Powerpoint slides and get more information about degree days from my blog at:

[danielholloway]

This is very interesting and a great rule of thumb but should only be used as a guide. Heat pump installers are required to meet specific temperatures within the house. This rule of thumb is based upon the present occupant consumption, this method will work very well for that occupant. But what happens if they sell their house and it turns out, that they liked their house cool, or were very frugal, as a lot of people are. This would result in a very low consumption and hence a low heat pump calculation. The new occupants however, perhaps elderly or people in need of a warm house will have an undersized heat pump. I’m all up for simplicity so thanks for this guide.

[michaelduvwiama]

This is very interesting! My final year dissertation is on heat pump retrofitting for energy efficiency and I definitely need some of these experiences you have 😅👏

[symonchester]

Awesome video, glad to be reassured that the calculations that Octopus quoted me are correct, can't wait to get mine installed

[peteglass]

Gosh, it was back in the late 70s when I was given a review copy of a self-published 1976 book from a chap who dug up his garden and put in his own ground source heat pump. I always been interested but never quite had the confidence to do anything. With my difficult 3-storey 1860 terrace which I did partial internal insulation 30+yrs ago [50mm PU foam] it'll be an outlier for most installers to understand. Your rule of thumb will give a first order idea of where I stand before getting out my R-value and U-value tables. I think I'll have to do my own spreadsheet before letting any installer through the door ... meanwhile the ancient boiler staggers on ... 😁
Thanks for the help, Michael, I'll try and find time to read more detail in your blogs.

[kingsleyimo]

That was simply wonderful. Your presentation was simple and straight to the point. It highlighted all the very important points I needed to make an informed decision about selecting a heat pump. Thank you so very much as this came at a crucial time of making a decision on what size of heat pump I might need.

[eddyd]

I've just had a heat pump survey carried out and your figure came out pretty much bang on.

[mortenchristensrn]

Thank you for making this video. It kept me from buying an greatly oversized heat pump.

[tonyhunter]

Brilliant! Thank you for such an informative vid.

[mikekelly]

I opened the video expecting some kind of guesstimate like 100W/sq.m but I got a pleasant surprise: this is actually a nice condensed sensible way to calculate equivalence that makes a lot of sense for end users who can't or don't want to do the more complex load calculations. well done!
I'd suggest leaving a bit of margin to account for differences in Winter temperatures, unless the reference year was unusually cold. Choosing the closest capacity above the calculated load might be enough.

[DrMatthewhannah]

Thanks Michael, great presentation, sent here from BetaTalk Podcast which was also excellent 👍

[jeremyallen]

Really useful explanation, I think focusing on real usage is the way forward ( if you have the history ) You have helped me make a decision on my heat pump selection.

[neilafram]

Excellent to-the-point presentation. It should be the bases of any selection and types of dwelling. The calculation is simple but most importantly easy to understand and apply especially to design and build companies. MCS requirement is extremely complicated, especially for installers. The method needs to be adapted for domestic application. Thank you

[kasparbradshaw]

Hi Michael

It's a great rule of thumb with the exception of well insulated properties, those which don't heat 24/7 and those with a bigger hot water demand. If hot water is being circulated around the house this could skew the results alot too as that can add a big chunk of energy consumption. A few comments.

1. I don't think average gas boiler efficiency is 90% but closer to 75 or 80% as few people run their Heating system with a return temp below 55deg. Also running a boiler with a return temp at 55deg is not 90% efficient. Condensing efficiency only starts at a return temp of 55 and increases the colder you run the system. If you have a gas boiler on weather compensation then hopefully it will also run down to 35 and 45deg and you will get the 90% efficiency. The rule of thumb using 80% would decrease the output of the rule of thumb so reduce contingency a little which is not critical.

2. More importantly there is an easy way to know the heat loss by experimentation. If you know your daily space heating demand in kWh you are almost there. A heat pump is generally designed to take 1-2 hours to heat the hot water tank. Assume 2 hours worst case. A hot water tank should have your daily useable volume but assume it is heated fully twice a day that is 4 hours of DHW operation per day. That leaves you with 20 hours to work in space heating mode, assuming you continuously heat the house. Divide your daily demand during the coldest day by 20 and that is the peak power heating required. If you have gas consumption for both space heating and hot water you can remove 2kWh per person for hot water.

In your case If it's 50kWh purely for space heating per day so divided by 20 hours gives 2.5kW.

Bear in mind that this only works on the actual coldest day that the heat pump is expected to bring the house to temperature. If it's only 5 deg outside and you are sizing to -2 (standard for London) then you would need to compensate for this.

If you don't heat 24/7 then the operating time is less. If you have the heating on for only 8 hours per day then the heat pump would need to provide the winter daily space heating demand in 8 hours (or a few more If you start heating in advance). 50kWh/8h = 6.25kW.

I don't agree that continuous heating uses the same energy consumption as intermittent heating. I guess it all depends on the thermal mass of the property and how long it takes to get to temperature. Masonry buildings with EWI have alot of thermal mass but timber frame which is lighter will have less.

The heat emitter will also affect intermittent vs continuous. If you are using UFH then it will take a long time but radiators are quicker to respond. If you benefit alot from solar gain then a fast responding heat emitter will be more efficient.

I would say generally, the same way you can lower your heating consumption by reducing from 24 to 21deg target room temp you can also lower energy consumption by reducing night time temp to 15deg and reducing daytime temperature when you aren't home.

In terms of calculating your own heat loss based on U values it doesn't have to be so hard. For a quick rough estimate I calculate the whole house heat loss rather than individual rooms. I calculate the total floor, wall and roof areas and multiply each by relevant U values. For window and door area I use an approximation of 30% of total wall area. I also do the room by room calculation but only after having a rough idea first.

It's a shame that alot of installers oversize but to be fair there are lots of mistakes you can make as an installer if you aren't interested in how the heat pump actually performs after its installed. Radiator or UFH sizing, pipework size, pump speed setting (where required), pipe insulation, distance from heat pump to tank to name just a few.

If you are interested in installing a heat pump in London and want an installer who cares, you can contact the company where I work, www.Nextstepheating.co.uk

[dhunt]

I think i would suggest to do all three and try to quantify the level of error with each and see how they compare!

[gavthegas]

Michael what a fantastic video. I have been doing a lot of monitoring on my own home and the MCS heat calculations are always far too high.
Your rule of thumb is spot on. Thanks

[chemicalandy]

Thanks for this video. I've just had an indicative quote for a 16kW heat pump with the size based on a basic description of my house. They did say that this may come down to 12kW after a survey. Your formula gets me to about 4kW!

[peterdoyle]

Outstanding presentation - thank you very much.

[karencarpenter]

Excellent. Thank you. Simple rule (but the curious will want to know why!). Very good. This all stems from 57.3! Who found that?!!!

[aprobin]

Thank you very much for the video, really useful. Can I offer a couple of suggestions for the heat pump sizing:

1. I believe your calculation assumes a heat pump running 24hours/day. I appreciate that’s probably the most efficient, but practically It might be better to assume a lower figure than this.

2. Instead of using the annual degree days, how about using the degree days for the coldest month (or week) of the year. The degree days web site will generate monthly or weekly figures and that way you’ll be assured your heat pump will cope with the very coldest parts of the year.

[Gary-tsdh]

5:24 - You have obviously put in a great deal of thought an effort into this problem. This looks quite a simple way of determining need; but, I'm curious as to how you arrived at the constant used in the denomenator of your 'formula'. Thanks for sharing your thoughts.
10:43 - Thank you for answering my question later in the video. A little calculus never hurt anybody! I could have just watched until the end and dropped the original post.