Earth Notes: ASHP: Domestic Air-Source Heat Pumps

Updated 2024-03-22.
Ensure that your ASHP installation is greener than mains natural gas for central heating and hot water (CH/DHW) in the UK. #ASHP #heatpump

How to get a heat pump for your house (in Great Britain) and the Boiler Upgrade Scheme.

What's the greenest space/central heat (CH) for a typical UK suburban property along with DHW (domestic hot water)? After insulating to the max... With apologies to Aretha Franklin, is it A S H P?

(Originally written in 2008.)

Assuming that you don't have masses of land for wood or even solar collectors, then mains natural gas at ~0.19kgCO2/kWh of heat is currently probably just about best. It's a lot better than plain old electrical resistance heating (eg immersion water heater and convection/fan space heaters) at a UK grid average of ~0.43kgCO2/kWh. Other fuels such as oil and coal are somewhere in between on cost and CO2.

But suppose that you don't want to be burning a fossil fuel for heat. Or you just feel uncomfortable about the energy-insecurity and price of gas. And you don't have space for the collectors of a ground-source heat-pump (GSHP) for example because you live in a flat, what then?


ASHP external fan, public domain image

In CO2 (ie 'global warming') terms alone, if the heat-pump system has a CoP (Coefficient of Performance, ie how much kWh heat you get out for each kWh of electricity in) consistently at or above ~2.3, then the heat-pump solution beats gas.

Why 2.3? Because that's the ratio of the CO2 emissions of electricity/gas per kWh, circa 2008--2016, but the grist is getting greener, implying that a lower CoP will work from a carbon-saving point of view in the future.

A belt-and-braces approach suggests purchasing the electricity to run the heat-pump on a 100%-renewable tariff. This helps continue to help 'green' the grid rather than just making everyone else's supplier a little 'browner'. Or better still, microgenerate as much possible from local PV/wind.

A GSHP solution can almost certainly manage that CoP of 2.3 or better, year round, especially if used for underfloor radiant (low-temperature) heating. It's a bit harder if DHW is being produced as well, eg over ~45°C. In the long-run the UK's grid electricity is likely to get 'greener' (ie less CO2 per kWh) and natural gas more expensive. So if you have the money and the space for the collectors and the system is installed 'right' in a well-insulated and efficient house, GSHP is almost certainly a good thing.

ASHP is likely to have a lower CoP than GSHP. Especially in winter when most used/needed. But a badly-installed GSHP can be beaten by a good ASAP.

Why? Because in winter when you most need the heating (particularly space heating) air temperature is likely to be significantly lower than ground temperature, making the system work harder and less efficiently. A typical ground temperature might be ~8°C or higher year round. But although air temperatures are rarely below zero in most of the UK, in winter and at night the air might well be close to freezing. And indeed at air temperatures of about -20°C or lower typical ASHP won't operate at all and you need a backup heat-source such as inefficient electric resistive heating.

There is a prospect of R744(CO2)-refrigerant-based ASHP for DHW with a CoP of ~4 from an external air temperature of ~0°C. That would certainly be ahead in CO2 terms of even the best condensing gas boiler for DHW. But most ASHP on the market won't get near to that, especially in winter. And you may be needing to replace an entire 'combi' system implying further compromises.

Combining with Solar DHW

Combining with solar thermal may work especially well even in the UK. No space-heating requirement at all for (say) 6 months, and solar meeting almost all DHW demand for the same period.

Combining with Solar PV

Note that at a CoP of 4 (or higher) then in terms of overall efficiency in terms of heating/DHW kWh from available insolation it may be better to use solar PV (20%+ efficiency) in conjunction with ASHP to provide heating (pulling in extra power as needed from the grid). Why? PV's ability to export all excess in times of low demand (eg summer). That cannot be easily done with solar thermal/DHW systems.

For example, in the UK using solar PV and G83/1 grid-tie inverter technology, and assuming a normal single-phase supply, with minimal red-tape up to 16A (3.7kWp) of generation can be connected. Ignoring other electricity demand and limitations for the moment, that corresponds to nearly 16kW of heat availability. That is comparable to a small (gas) boiler.

More pertinently perhaps, on a reasonably-pitched south-facing London roof that might require ~25m^2 of roof space and produce ~2.7kWh/day of electricity thus 10kWh of heat, which might well cover all DHW needs for a small family. (Even on our east-west facing roof surfaces, generating ~1.4kWh/day, that might generate the equivalent of ~6kWh/day DHW which might just suffice.)

It may of course be possible to get permission from the DNO (Distribution Network Operator, that runs the electricity cables to your house) and the electricity supplier upstream to connect more than 3.7kWp of PV. Then all you need is sufficient roof space.

On very dull/cloudy/cold days you would be importing electricity to heat your DHW (eg instead of gas). But over each month (even December), you should be approximately neutral or better. And indeed exporting in summer. and displacing fossil-fuel burn elsewhere on the grid. (When using ASHP for space heating you'd always be importing for that unless you employed a huge and well-sited PV array.)


One important point in choosing an ASHP could be its refrigerant. The 'wrong' sort in a domestic system upon leaking/disposal (a few kg but at up to the green-house equivalent of several tonnes of CO2) could easily exceed years of fossil-fuel-fired CO2 flue output. So, from that point of view, good refrigerants with low GHG/CO2(e) impacts include: R744 (CO2 itself), R717 (NH3/ammonia, not a GHG), and R290 (propane).


Again, as for GSHP, we could do with good independent reviews in UK conditions. These systems are so expensive that you can't just try two or three, ripping them out until you get it right. And the technology is complex.

Here is a good start: the Energy Saving Trust 2010 heat-pump field trial phase 1 [PDF], and phase 2 [PDF].

The Market

In private communications with one of the ASHP manufacturers 2008-08, I was told that the UK domestic boiler replacement market is ~1.5 million/year. So everyone is keen to be in it with a good range of products. However, CO2/R744 systems were proving very tough to make economically.

2020-10: Green Homes Grant

Radbot is allowed as a secondary measure alongside a heat-pump in the UK's Green Homes Grant scheme, so I'm going to kick the tyres.

It's widely accepted that getting ASHP from concept to install in the approximately six month timescale of the GHG scheme can be a challenge. This is not helped by the relatively small number of installers with the appropriate TrustMark / MCS / PAS accreditation that can do the work.

Visiting the Simple Energy Advice (SEA) website (during which I claimed to have no accessible garden because it's smaller than the site allows for):

  • Tells me that my 2009-vintage EPC shows grade B (and that the house has 5 habitable rooms, and is heated by mains gas), better than average. Most homes appear around grade D.
  • Suggests:
    • solid floor insulation, GHG primary measure at £5k (we have done the best we can above the downstairs floor)
    • upgraded heating controls, secondary at £400 (already done with Radbot!)
    • condensing gas boiler, not GHG-eligible at £2.6k
    • solar thermal, primary at £5k
    • insulated outside doors, secondary at £1k (already done eg front!)
    • ASHP, primary at £6000 to £11000 (bingo!)

I have added just ASHP to the "plan", at a suggested required investment of £8500.

I am told that you may be eligible for funding 2/3 of the total cost, up to £5,000.

Here is the PDF downloaded plan.

The link to the TrustMark site brings up 65 ASHP installers, the closest five within 10 miles, for example.

One of those is Your Energy Your Way, that I follow on Twitter! I sent them an initial enquiry, to which they have responded, but I know that they are very busy.

, I'm going to email a similar request to the top few TrustMark installers close to me that have a Web site and suitable contact details. Again, I expect them all to be stupidly busy.

2020-10-28: Enhabit Response

Enhabit came back to me today (after a little more to-ing and fro-ing) with an approximate costing of £11,500 + VAT at 5%.

However Enhabit asked again for details of insulation in each area on the floor plan, to allow us to accurately calculate the size of heat pump required even though I have provided them with heat-demand estimates based on down-to-hourly granularity of gas consumption.

Enhabit also asked for an exact location for the external unit that complies with permitted development. We need the location of the internal cylinder and other equipment, which requires a full height cupboard measuring 1,200mm wide x 900mm deep, and confirmed they will arrange for this to be constructed prior to install even though I previously said We would need advice as to where to locate the internal cylinder and buffer vessel (and indeed the rest of the equipment on our small plot to minimise planning hassle etc. Enhabit replied Unfortunately we do not do site visits at this time so it will be difficult to advise where to place the unit.

Neither perfect nor "oven-ready" but it's a start, and compares to £20k to £30k for ground-source.

2020-11-10: Enhabit Further Response

I have a further response:

It sounds like the location you have provided will not be suitable.

The heat demand information you sent over previously is very helpful, however we need to size the heat pump based on peak hea[t] load not annual demand consumption so without this information we will not be able to quote.

I have replied that I have per-hour demand figures available as previously mentioned, in particular per-hour demand on the the days with peak space-heat demand. Allowing for DHW, that peak power demand has been approximately 3kW.

2020-11-11: No Dice

I had a further follow-up from the MD:

... we don't often get sent hourly consumption data from clients. I have checked again and the data provided is sufficient. For this level of heat demand we would suggest a 5kW heat pump with 150 litre cylinder, which would cost around £9,500 +5%VAT to supply and install.

However, having looked at your property on Google maps I can see the challenge you have with locating the external unit. I think it is very unlikely that you will be able to find a suitable location and as such we can't proceed with this quote request. Sadly a lot of UK properties are not suitable for heat pumps.

He clarified that he suggested 5kW against an apparent 3kW peak demand because I would always want to include a safety factor into the design and we are restricted to heat pump sizes.

Further examination of by-hour gas consumption in months with the highest HDD suggests some mornings after cold nights show peak demand of 5kW up to 8kW. There may be some other (eg DHW) demand mixed up in there. But the demand over 3kW is usually only an hour, at most two, so at the cost of the house taking longer to warm on those extreme days, 3kW peak heat supply might still be enough. For example demand for example is 3.33kW at 8am, 3.89kW at 9am, down to 1kWh at 10am with final demand at 1.33kW at 11pm, for 30kWh over 16h, mean 1.9kW. The following day starts at 1.2kW at 7am, then 6kWh then 1.2kWh for 35kWh over 17h. 1st March averages 3kW demand (51kWh/17h), and is the highest day's demand in the Loop data, ie back to the start of 2016.

2020-11-14: Movement

I gently prodded Element 5 by email this afternoon with some more data:

I have just re-analysed the last few years of gas data for the highest heating-degree-day months, and only on four days was total gas demand over 40kWh (including non space-heat use):

2017/01/21 40.54
2018/03/01 50.87
2018/03/02 46.98
2018/03/18 41.43

On those and similar days, hourly analysis shows that peak demand first thing in the morning has apparently very occasionally been over 3kW for more than an hour in a row for space heat, but I'll stick with the original ~3kW peak demand figure for now. I'd be prepared to top-up with resistive heating beyond that.

I had a fairly fast response aiming for a chat and then viewing the property next week. Good!

2021-02: in the cold "Beast from the East II", gas consumption hit ~36kWh on 2021-01-09 (HDD12=12.5) and ~40kWh on 2021-01-13 (HDD12=13.1).

2020-11-19: Gone Quiet

It's all gone quiet so I have requested progress with an email containing: Given that the government seems to have extended the GHG deadline by a year, things are maybe a bit more sensible.

2021-02-18: Slow

2021-03-12: Dead?

The GHG scheme seems to effectively be dead.

The government claims lack of interest from the public, though that seems to have been very far from reality. Shortage of qualified and certified trades able to ramp up activity quickly, and poor central administration, seem more culpable.

2021-03-27: Officially Dead

Good Questions To Ask

Ice Energy had some good questions to ask any potential heat-pump supplier, such as:


(Count: 2)