Calculating heatpump size (using gas readings to help?)

Hardware Heatpump
1

New to the forum and I’ve been trying to do as much reading as possible!

I’ve put my house plan into heatpunk and have played around with the air changes, but I don’t know exactly what our air tightness is, and obviously this makes quite a big difference to the heat loss calculations!

I was fairly happy with the idea that a Vaillant Aerotherm+ 7kW would cover out needs, however I’ve really been put off by all of the negative comments about noise in Vaillant Arotherm: excessive noise from compressor. And also the more I read about heat pumps, the more I realise that oversizing will mean that the unit can modulate down as well (leading to more cycling in certain conditions, which is bad).

I have started to wonder if I can get my heat loss low enough to be covered by the 5kW unit (which seems to have much better reviews?), however I have also seen the posts about baking in extra capacity for when the weather is both cold and damp, therefore I don’t want to go too low!

I was wondering if I could use my gas bill from previous cold spells to estimate my max heat loss requirements?

The most extreme gas usage I’ve seen since living in the property is:
18th Jan 2024 (between 0c and -6c outside, 19c inside)
124kWh gas used
85% efficient boiler = 105.4kWh of heat produced (including hot water)
105.4/24 = 4.4kWh average heat demand

Most of this usage is from 7am-11pm, 8kWh being the highest usage period (7-8am), fairly consistent 7kW/h between 9-12, and 6kW/h until 11pm.

So, based on this, would I be able to take the average hourly heat usage figure of 4.4kWh and go with a 5KW unit, or would I need to look at the max usage figure of 7kWh?

This year I have increased the loft insulation from ~150mm to 300mm and I have also been adding 100mm PIR between the floor joists (should be completed before this winter hopefully).

We also have a log burner which we could use in extreme periods.

The heatpunk calculations (when temperature is set to 20c and air changes set to 1) come out as a total heat loss of 5370 W, which is covered by the 5kW unit at 46C flow temperature ( 5676W), however I don’t think heatpunk take into consideration the drop outs caused by the defrost cycles Vaillant maximum output capacity testing.

Would love to hear peoples thoughts on any of this?

3

Hi @ash, and a warm welcome to the forum.

You are asking the right questions, and hopefully you will get plenty of learned answers here. I’d just make a couple of general points:

  • No matter what your heat loss actually is, remember that it’s what your installer estimates that’s important if he will be applying for the all-important BUS grant on your behalf, and that he has an incentive to oversize your heat pump. (Two incentives actually - he makes more profit from selling you a bigger HP, and you are less likely to complain that your house isn’t warm enough.) So be prepared for a debate with him if you think his estimate is much too high.
  • There is a (minor) downside to choosing a HP that just matches your maximum demand - you’ll have less scope for roomstat setback at low outside temperatures, as discussed in House Thermal Inertia and Roomstat Setback (some cautionary notes). (Personally, my economics favour a night time setback, and I’m glad I chose a HP with a couple of kW spare for house reheating in the morning.)
  • It may pay you to review several HP manufacturers/models, and then choose an installer familiar with the brand. If you choose your installer first - especially if it’s one of the nationwide ones - you’ll end up with his standard brand, come what may.
  • As you will soon see on this forum, there is no perfect heat pump (they all have good points and bad points). But in general, I suspect that you get what you pay for…

Keep us informed of your progress!

4

I think you’ve got it more or less right, other than you don’t need to factor hot water into your heat demand calculation - making your actual heat loss a little bit lower than you’ve suggested (depending on how much hot water you used that day).

Also, if you’re after the BUS grant then what you think your heat loss is will be more or less irrelevant, as the installer will do a heat loss survey to MCS guidelines and decide what they’re comfortable fitting. In this case you’re performing the calculation to satisfy yourself that they’re not putting in something crazy oversized.

5

An approach to getting a ballpark sizing from your annual gas consumption (yes, really!) is described in this article. Michael DePodesta provides supporting calculations, explanations etc. The rest of his stuff is worth checking out as well.

6

I have an oversized Vaillant- it’s the 10kW rather than the 7kW which would have been big enough. I have two observations 1. Vaillant is pretty good at copping with oversizing - my system is doing pretty well in the league table despite being oversized. 2. The cause of the 7kW having poor reviews is mainly under sizing. In frosting conditions it doesn’t maintain the documented output. This coupled with a big incentive for people on the margin to opt for the single fan 7kW unit over the double fan 10kW, that until recently needed planning permission, meant that quite a few people ended up with a unit that couldn’t keep up with demand.

On balance I would go with the 7kW as recommended by your installer I doubt you will regret it.

2 Likes
7

I was lucky enough to have 2 years worth of smart meter readings so was able to do the plots in the attached. The figures are gas consumption, its a condensing boiler which is actually set up to condense (!), so might in principle be more than 100% ‘efficient’. This gave me the confidence to get the 7kW Vaillant and conclude that my ACH was 0.5-1.0. The two installers that made a song and dance about doing a 3 hour fabric survey told me me I needed ~16kW; my own fabric calculations using MCS ACH assumptions got to 10.5kW. If I substitute ACH = 0.75 I get close to the measured loss.

Its turned out to be just right, it can cope down to -5 against a design temperature of -2. I run it in low noise mode for 8 hours at night to suppress the compressor buzz, this does cause the IAT to drop a degree when its <-2 outside and its defrosting a lot (we had freezing fog/-5 for a few days last winter), but it has just enough spare to recover the following day.

Although your numbers do definitely suggest that the 5kW will be sufficient I would personally would be nervous about drawing conclusions from very limited data; houses have significant ‘memory’ and so you really do, to be certain, need a reasonable dataset. 2 years is much more than you need but I wouldn’t personally want to do it with less than several weeks. Obviously it depends on what you have/can get and in your case as the figure from measurement is similar to the heatpunk figure you do have an extra degree of confidence.

The data in the attached was sufficient to convince two more open minded installers, who were anyway pre-disposed to assuming a lower ACH than the MCS assumption.

The Mitsubishi R290 8kW has a 6kW (variable) and a 2kW (fixed) compressor. That may be worthy of consideration as irt covers the uncertainty, although the Vaillant controls are better IMHO.

Measured consumption.pdf (785.8 KB)

8

Hi Ash,

I am in a 1935 solid walled 3 bed semi and my 7Kw Vaillant has been good at -9c with 20c inside. Still only consuming 2Kw out of the 2.4 Kw it draws heating hot water. In ECO mode its almost silent, and even running flat out its only about the same as a freezer.

If you are putting PIR between the joists and sealing the pir into the joists with builders foam (that is what I did) you will do a huge amount of draft reduction and air changes will drop drastically. So much so that you must ensure that all air bricks are clear so that you get good air circulation under the floor. I am a bit paranoid about this, having seen dry rot eat my brother-in-law’s house.

I insulated external walls internally with 50mm PIR and to head off any possible condensation issues I installed a MHRV system, which works a treat (Brookvent) That was 6 years ago and still working fine. If you take a look at their website, they now do a new version that has the now obligatory app to control it, but last time I looked, they still did the old hardwired version at nearly half the price. The app is pointless as after initial calibration it never needs adjusting and runs 24/7.

I think your calcs will come in very handy as you can then compare them with an installer’s high estimate and ask them to prove you wrong if they can.

Regards

Nigel

9

I think that looking at historic gas usage is a very good way of estimating heat pump size and I’m really surprised it isn’t done more. There are some complexities around variability of demand / peak demand / defrost impacts but another way of looking at it is to take the installers calculations and to compare their estimated total heat demand with your historic total gas heat supply. To the extent there is a difference you can broadly scale the heat pump size they recommend.

Another factor is your willingness to use top up space heating of any sort on the coldest days of the year.

10

I agree that personal historical heating demand is a good yardstick, but it doesn’t account for the BUS grant eligibility being predicated on standardised assumed occupancy and heating settings.
In the case where the present occupier is a frugal, single occupier of a large house who only heats part of the property, historical consumption will reflect that case.
BUS requires that the HP installation will satisfy 100% of the heating and hot water demands of the follow-on family occupiers of that property.

If the occupier wants their HP install to qualify for BUS and satisfy future owners’ reasonable demands then some adjustment of historical usage MAY be required.

As far as I’m aware, supplementary non HP heating can’t be factored into the MCS heat loss calculations.

2 Likes
11

It can in limited instances - I have a plinth heater (that isn’t plugged in) in the kitchen, which allows me to have a slightly undersized radiator. The larger radiator size needed would have been too disruptive to the kitchen layout.

12

I used the approach you suggest here very succsessfully for my heat pump project. It really gave me confidence the heat pump would work.
I have the 7kw Vaillant my brother, who has a smaller house, has the 5Kw. I have noticed that his goes through many more defrost cycles than mine in the same conditions. This obviously impacts its efficiency significantly.

13

Hi @ash

I’m very Interested in assessing whether gas and electricity usage can be used to predict whether an ASHP will operate at a reasonable SCOP when it is correctly sized.

I’ve been trying to develop a way of helping homeowners to make the decision to get the calculations done by first evaluating how they currently use the house. This revolves around how they heat the house, cook and use other appliances. This is best taken from their actual annual gas, electricity and other forms of heating. I’ve been working on the basis that the gas heat input is related to the efficiency of the boiler and similarly for wood burners and that all electricity contributes to heating the building envelope. This can then be related to the gross internal area of the house to derive the energy usage intensity – EUI – in kWh/m2/yr.

In order to test the spreadsheet model (currently called House Master), I need annual before and after an installation data and whether the ASHP performance meets the customer’s expectations for performance and cost. Let me know if you (or any other readers) would like to participate in the data collection.

The output will look like this:

Screenshot 2025-08-08 at 16.54.17
Screenshot 2025-08-08 at 16.54.171451×617 191 KB

[image]

We are hoping to build up a network of case studies of different house types which we will cover in a new website and ASHP’s are key to big carbon reductions.

I covered some of this in a previous post Average electricity cost p/kWh

[Edited to “@mention” ash - prefix their username with a ‘@’ symbol.
also - drag and drop your image into the post – Moderator (RW)]