Heat pump sizing sense check

[Dillsue]

We're planning on installing an ASHP later this year and just working through heat loss and resizing rads. Total calculated loss comes out around 4kw without any allowance for ventilation losses. Back in December we heated the house for a couple of days with 3 electric fan heaters running 24/7 while it was around zero outside and 4kw of direct electric heating kept the house at 18/19 degrees. This roughly aligns with calculated loss so am I correct in thinking this  validates the calcs???

 

The reason for asking is an MCS umbrella company have suggested an 18kw heat loss based on floor area and fabric details I've given them. I appreciate that MCS guidelines drive a very conservative approach but the 18kw seems way over the top.

 

Am I missing something??

[JohnMo]

You need to sort the losses from ventilation really also.  You would also size for 21 in the main rooms and 18 I believe in bedrooms under MCS install. Not 18 to 19 overall.

 

But sounds like a 6kW unit not an 18kW. So a big miss match of information.

 

Are you sure you have not missed something in the handover of information.  Wall, floor, roof and windows buildup? Do you have an air test certificate? MVHR?

[mk1_man]

Are you able to say what umbrella company?  The first I approached were insistent that I needed a 12kW unit when in fact I ended up installing a Vaillant 7kW and it has been great even when sub zero.  What type of property do you have and what are floor areas etc. Also is it radiators or underfloor?

[ProDave]

You have proved at 0 degrees outside, 4kW of heat is only just enough when running 24/7

 

It will sometimes be colder than 0 for several days.

 

Your heat pump can't do heating 24/7 or else it will not have time to heat your hot water.  And do you really want it on 24/7 (we prefer a quiet house at night)

 

So i would say a minimum of 10kW ASHP.

 

MCS don't work on real figures so just their own way of calculating, and the last thing they want is to have to rectify an under specified system so they probably take a cautions approach.

[joth]

What heat emitters will you use with ASHP?

If it's ufh in a decent thickness slab then it gives the option to "batch heat" the house during cheap overnight electricity. This is typically 4-8 hour window, so only 1/6 to 1/3 of the day, thus you need to 3-6x the ASHP to get your full day's worth of heat delivered in that short window. Of course on the very coldest days you'd allow it to run for a longer period, the goal would be say 90% of time do it all in cheap rate. So about 3x would be plenty.

 

Anyway our PHPP said 3kW max heat load, we installed an 8.5kW and that's been perfect. Plenty of headroom for batch heating and of course allowing some DHW time. It spend 12-16 hours idle even on the coldest days. 

I'd agree with Dave's 10kW estimate 

 

Edited February 8 by joth

[S2D2]

3 hours ago, Dillsue said:

4kw of direct electric heating kept the house at 18/19 degrees

What temp did the house start at? If it held steady then yes, if there was even a small drop then a lot of uncertainty is introduced.

 

You need the ventilation losses too which depend a lot on the house, 0.5ach is a reasonable sense check, which makes 18kW sound way out.

 

The really, really important figure is the minimum modulation of the unit being proposed. This varies a lot with manufacturer, hence why it's hard to pin down a general rule. I'd agree with others who provide arguments for aiming higher than the bare minimum heat loss, there have also been some recent counter examples to that method as the industry has tried it in earnest. There's an urban plumber video on going back to upsize a unit because of uncertainty in the heat loss / defrost losses leading to a cooler than planned house. Vaillant also recently removed the Arotherm Plus 5kW from the compatibility chart for a 250L cylinder; I haven't had clarification on why but presumably because in the extreme case it spends too long heating the hot water, meaning the power outside that window has to be increased to compensate.

 

So you're right to reject 18kW but maybe don't aim for the bare minimum.

[JamesPa]

Measurements over 2-3 days can be quite suspect depending on the preceding conditions.  Do you have any other measurements over a longer period of time which could help provide a greater level of confidence.  18kW sounds much too big based on what you say but Im not sure I would yet be confident of 4 or even 6 (which is not to say that this isnt correct, just I wouldnt be confident of it based on 3 days).

 

FWIW I used 2 years of measurement (gas consumption from half hourly readings) and plotted daily consumption vs degree days, as well as checking averages over 3, 6, 12 hours.  I ended up concluding that I needed 7kW, compared to the 16kW that two 3 hour fabric surveys told me.  I fitted the 7kW Vaillant (which claims 8.5 at the design conditions - although there recently some doubt has been case about that figure) and it seems to be 'right sized'.  So don't be surprised about a large differential, but do make sure you have enough data! 

 

Regarding ventilation I see that @S2D2 is suggesting 0.5ACH, Ive have also now heard three of four installers say that 0.5-max 1 ACH is a good value to assume, as opposed to the often much larger MCS assumptions.  That value seemed to work for me and allowed me to reconcile the various figures.  However this, of course, says nothing about your case.

[Dillsue]

Wow, thanks for all the input. 

 

We currently heat 95% with wood 5% lpg so I've no historic consumption figures. We'll be keeping the WBS and boiler to give a hybrid system although if we use the MCS umbrella outfit Im assuming we'll need to disconnect those and reconfigure after the HP is commissioned and signed off. There's wet UFH in the tiled kitchen floor set in an insulated concrete slab the rest of the house is rads which most will be getting swapped around or replaced sized to run at 35deg c. House is 270m2 built in 2001, low e DG, 75mm insulated cavity on the ground floor, 150mm insulated timber frame on first floor, 250mm loft insulation

 

I have to work out how to factor in ACH but I'm not intending to size the HP for exactly the calculated heat loss as it needs time to do the DHW. With having a hybrid system the intention is to use the WBS/lpg boiler if the HP was struggling through a cold snap but in the 24 years we've been here those are few and far between. Weve got solar thermal and not averse to putting the immersion on occaisionally if the HP was busy with the heating.

 

Based on the circa 4kw loss I'm thinking a 6Kw HP would give a bit of headroom for DHW and cope with the heating at least 95% of the time. Does that seem reasonable or is there more to consider??

 

Edited February 9 by Dillsue

[MikeSharp01]

I just thought I would ask MS Copilot to give me an overview of the sizing views here. Worked out OKish and this thread featured. Prompt was simply:

 

"Hi, what is the consensus on buildhub.org.uk about sizing a heat pump to a house please." (I always include the Hi at the start and the please somewhere because I don't want to degrade my attention to such things in other parts of my life - IE I want to avoid assuming that everybody I talk to are automata.)

 

Result was simply: 

 

" On BuildHub.org.uk, there's a general consensus that accurate heat loss calculations are crucial for sizing a heat pump - https://forum.buildhub.org.uk/topic/42699-heat-pump-sizing-sense-check/. Users emphasize the importance of considering ventilation losses, insulation quality, and the specific heating needs of different rooms - https://forum.buildhub.org.uk/topic/42699-heat-pump-sizing-sense-check/. 

Many contributors suggest that over-sizing a heat pump can lead to inefficiency and reduced lifespan, while under-sizing can result in inadequate heating. It's recommended to size the heat pump to handle 99% of the heating load, ensuring it can cope with the coldest days."

So this thread features as do others. EG this one: 

 

 

 

 

[Dillsue]

I'm just looking at factoring in ventilation losses and am struggling to understand suggested air changes! Seems a figure of 0.5 ACH is reasonable for a reasonably well sealed house. If I use this figure it puts the heat loss up by about 50% which is having a massive impact on radiator sizes needed to run at 35 degrees.

 

If half the air in the house is replaced every hour that's 4 complete changes of air in the 8 hours were in bed. We don't have the heating on overnight as the house holds its heat fairly well. Seems implausible that the house is getting 4 changes of potentially freezing outside air overnight or it would be freezing next morning, which it isnt. Only explanation I can see is theres enough thermal capacitance in the fabric that can reheat multiple changes of air??

 

Is 0.5 ACH actually what happens or is that a figure the B Regs want you to aim for??

[JohnMo]

1 hour ago, Dillsue said:

Is 0.5 ACH

If you have MVHR the air change per house is 0.3 to 0.5. But this has about 80 to 90% heat recovery.

 

An air test of building tightness of 10ACH would give you around 0.5ACH natural infiltration.

 

If you looking at a grant, the MCS spreadsheet has pre added ACH figures. You want them to move on the numbers you need an airtightness test result and then argue the point.

[S2D2]

1 hour ago, Dillsue said:

theres enough thermal capacitance in the fabric that can reheat multiple changes of air

Yep.

 

MCS will want 2-3 ACH depending on room type, it's taken as the worst case but for your average house where effort has actually been put into closing up gaping holes to outside 0.5 can be a reasonable sense check. It might actually be higher or lower, you need an air tightness test to know for sure. MVHR changes the ventilation heat loss dramatically for obvious reasons.

[JamesPa]

39 minutes ago, S2D2 said:

MCS will want 2-3 ACH depending on room type, it's taken as the worst case but for your average house where effort has actually been put into closing up gaping holes to outside 0.5 can be a reasonable sense check. 

Some mcs installers will work with a more realistic value than the 2-3 ACH that is the default.  If it matters in your case @Dillsue (which it may well not given your low loss), collect together the evidence and shop around.  There is flexibility in the 'rules', some installers choose to use it, others don't.  Don't be bullied!

Edited February 15 by JamesPa

[SteamyTea]

For the few quid it costs, why don't you make yourself a blower and see what is really happening.

Give you the opportunity to find those leaks as well.

[Dillsue]

On 16/02/2025 at 06:45, SteamyTea said:

For the few quid it costs, why don't you make yourself a blower and see what is really happening.

Give you the opportunity to find those leaks as well.

Never looked at air tightness measuring but do you not need calibrated pressure/flow meters?

[MikeSharp01]

3 minutes ago, Dillsue said:

Never looked at air tightness measuring but do you not need calibrated pressure/flow meters?

Calibration not needed, I did ours with an dust extractor unit a relatively cheap anemometer and a length of plastic tube to make a manometer ( I also used a high end differential pressure meter as well but the manometer gave a reliable result.) I got within a few % of the official test. Builhub has an anemometer in its loan system.

[marshian]

On 08/02/2025 at 22:48, JamesPa said:

FWIW I used 2 years of measurement (gas consumption from half hourly readings) and plotted daily consumption vs degree days, as well as checking averages over 3, 6, 12 hours.  I ended up concluding that I needed 7kW, compared to the 16kW that two 3 hour fabric surveys told me.  I fitted the 7kW Vaillant (which claims 8.5 at the design conditions - although there recently some doubt has been case about that figure) and it seems to be 'right sized'.  So don't be surprised about a large differential, but do make sure you have enough data!

 

I used the heat engineer software to calculate my heat loss (at -2.4 Deg C)

 

Heat Engineer software suggested that my heat loss is 4.93 kWh - but I did make a couple of mistakes in the process which when I spotted them I back calculated the difference and got to 4.7 kWh

 

However I have also been measuring my gas usage v HDD (Heating Degree Days) for nearly 3 years however until Dec 2024 I heated to a schedule rather than 24/7 which compared favorably to the heat loss calc because of the periods where heat loss is reduced due to a lower internal house temperature so my kWh/HDD was 3.5 to 3.8

 

Moving to 24/7 heating with a weather compensated flow temp has increased the kWh/HDD figure to 4.2 to 4.6

 

If a HDD value of 20 is a good fit for 24 hrs at -2.4 deg C (pretty sure it is) then my heat loss is between 3.5 and 3.75 kWh

[JamesPa]

@marshian Im assuming you mean kW not kWh in the above.  If so with calculated losses of 4.7kW and measured 4.2-4.6 you have a pretty good indication what the actual loss is, I would say.  You do, in the latter, have to make an assumption about gas boiler efficiency (75-110% depending on age/operating mode) but any fabric calculation also makes many assumptions. 

 

Many may suggest you go for 6kW so you have oomph to heat DHW in a sensible time (its worth doing the calculation on reheat time if you have a low loss house and use a lot of DHW!); also some manufacturers 4-5kW models are just hobbled 6kW machines so the lowest output is no better and there is little or no point in going for these.  Of course the latter us manufacturer dependent

Edited February 18 by JamesPa

[marshian]

49 minutes ago, JamesPa said:

@marshian Im assuming you mean kW not kWh in the above.  If so with calculated losses of 4.7kW and measured 4.2-4.6 you have a pretty good indication what the actual loss is, I would say.  You do, in the latter, have to make an assumption about gas boiler efficiency (75-110% depending on age/operating mode) but any fabric calculation also makes many assumptions. 

 

Many may suggest you go for 6kW so you have oomph to heat DHW in a sensible time (its worth doing the calculation on reheat time if you have a low loss house and use a lot of DHW!); also some manufacturers 4-5kW models are just hobbled 6kW machines so the lowest output is no better and there is little or no point in going for these.  Of course the latter us manufacturer dependent

 

I debated ASHP but decided to stick with a gas boiler so current boiler is Viessmann 100-W Heat only (16kW with Weather comp and DHWP - range rated to 4.5 kW for Space heating)

 

I was merely posting to support the statement you made that historical energy usage can be a good guide to sizing of ASHP provided you are heating in the same way - ie low and slow rather than scheduled