@jamespa the answer to your question is in our A2A podcast episode featuring @ashp-bobba, and it drops at 15:00 tomorrow. 

Definitely air to air (and we did)

We have a big solid stone 1860s victorian pile with 18" thick solid walls. It's grade II listed so a lot of insulation is simply not possible and would wreck the building.

A couple of years ago we approached one of the heatgeek people who dealt with old houses and after some initial calculations the requirements for A2W were comical (over £30K, filling half a basement room). We installed air/air heatpumps instead.

No grant, but they cost so much less it didn't matter. Even today I wouldn't bother with the grant because it means you've got to over pay some MCS "accredited" installer, remove gas, and have all the setup to meet MCS rules not your needs. We've got loads of very competent existing aircon engineers who are not going to bother with the MCS stuff and will be cheaper.

We kept the gas (it still heats some water for now, and a cooker until it needs replacing) which means if it gets really cold we have both as options although we've found we didn't actually need it. 

My experience though is that air/air installers believe every house needs exactly the maximum number of heads that the multisplit they are trying to sell you will support. The reality is usually far less especially if you leave a few doors open. You can heat a reasonably insulated small terrace with a single unit at the bottom of the stairs, although another one in the bedroom is nice.

Just avoid portable units with a single hose, that's a very expensive way to buy what is basically a glorified fan. Dual hose portable for summer is fine (COP can be about 3 which is ok).

For garden offices especially the "through the wall" type units can be self installed and are great. In theory they are "self install" for houses and flats but unless you habitually drill 15cm holes in your external walls .. then nope.

Other thing in my experience is that A2A can generally be installed okay (ie COP about 4) by someone who is fairly clueless, whereas A2W requires a magic wand and pointy hat. As the UK has a lot of fairly clueless installers that's a huge win.

For A2A it's usually cheaper to just use an immersion heating element if you have access to an EV tariff unless you use a lot of water to justify the separate heatpump water heating.

The podcast on air-to-air heat pumps with @ashp-bobba is out:

Great, I hope I did ok? 🙂

@ashp-bobba smashed it!

Thanks for a great podcast that answered pretty much all my A2A questions. Here's some of my reflections on all the new information. 

I was a little surprised @ashp-bobba was so definite that A2A isn't as good an option as A2W for most UK homes. Add to that what he says about how it's not worth much to installers to get the MCS certification for the £2,500 BUS grant and it doesn't sound like A2A is going to be the next big thing. I'm wondering if I should change my vote.

What he said about managing air flows (and expectations about them) was particularly useful - if discouraging. Similarly with cleaning the indoor units.

I did think it was a bit pessimistic about cooling bedrooms. In the UK, even in a heat wave, night-time temperatures aren't so high that a/c would need to run all night. At least not here in Yorkshire. So for me, where the problem is that my bedroom is west facing and warms up a lot through the afternoon and evening I would work to cool the room beforehand then turn the a/c off.

I appreciated the technical point about how A2A systems do not modulate to low outputs as well as A2W. I'd suspected that. For instance I've looked at the Daikin Multi+. With 4 indoor units plus hot water it would do all of my 2 bed house except the bathroom - which is pretty good. It's rated at 9kW max and 3kW minimum (from a data sheet I can't now find).

So for me a Multi+ is oversized. My heat loss is calculated at 4.8kW, for 21C indoors -3C outside. Measuring it (from gas consumption and degree days) I think the reality is more like 3.6kW. Factor in that I have an indoor temperature of 18C and the maximum load comes down to 4.2 and 3.2 kW. Put another way I'd rarely draw even the minimum amount of power.

On efficiency it was good to get comparable COP numbers. They suggest A2A would use around 75-80% of the energy of A2W. Which seems a significant difference. Especially if you use a lot of heat. I don't, only around 3,000 kWh per year.

As for installation costs. I've only had two A2W quotes. Heat Geek (not zero disrupt) came in at £17, Octopus at £11.5k that's £9.5k and £4k after a £7,500 BUS grant. The rough estimate from the Daikin website for a Multi+ system for my house was £9, or £6.5k after a £2,500 BUS grant if and when it happens. That does include adding a water cylinder as I currently have a combi boiler.

So maybe I need to get a better A2W quote. To be fair to Octopus I only decided not to go with them because of problems with where they wanted to put the water tank.

Finally it really is great to have such high quality information available.

Posted by: @springswood

↑

As for installation costs. I've only had two A2W quotes. Heat Geek (not zero disrupt) came in at £17, Octopus at £11.5k that's £9.5k and £4k after a £7,500 BUS grant. The rough estimate from the Daikin website for a Multi+ system for my house was £9, or £6.5k after a £2,500 BUS grant if and when it happens. That does include adding a water cylinder as I currently have a combi boiler.

A2A would only work in smaller properties (and I think there's a consultation running to eventually allow more people in flats to have them). 

We found that since the Heat Geek site took our EPC data to estimate heat loss, it oversized the pump. It would have been nice to be able to override the heat loss estimate. But anyway, we found a HG installer able to give a rough estimate based on another survey, which was backed by our gas usage. 

Posted by: @batpred

↑

Posted by: @springswood

↑

As for installation costs. I've only had two A2W quotes. Heat Geek (not zero disrupt) came in at £17, Octopus at £11.5k that's £9.5k and £4k after a £7,500 BUS grant. The rough estimate from the Daikin website for a Multi+ system for my house was £9, or £6.5k after a £2,500 BUS grant if and when it happens. That does include adding a water cylinder as I currently have a combi boiler.

A2A would only work in smaller properties (and I think there's a consultation running to eventually allow more people in flats to have them). 

We found that since the Heat Geek site took our EPC data to estimate heat loss, it oversized the pump. It would have been nice to be able to override the heat loss estimate. But anyway, we found a HG installer able to give a rough estimate based on another survey, which was backed by our gas usage. 

 

In relation to the grant, I think you are correct, it will probably only work on smaller flats to make it value for money, there are some consultations looking into solutions for flats, especially examining application and permissions. But to be clear, A2A limit of design for size of room and building is technically limitless or only limited by the power supply or space for the systems. The largest A2A systems we have installed are hundreds of kW's in very large office blocks, the Tate galleries, large schools, hotels just to mention some examples, most manufacturers build up to 100kW modules and then they can be control networked together on centralised controllers to form 500kW systems if needed. These systems come in all shapes and sizes but are still A2A in one form or another. The largest value job we did just for A2A on one site which was the University of Hertfordshire was £500k, thats a lot of systems and they covered a lot of space and a specialist Olympic gymnasium. Although these are commercial applications this is still the same A2A and it shows that it will heat and cool any home including a castle if you want to imagine big, its a matter of deign and application.   

With regards to the overstated heat loss from EPC's, CIBSE now released their new 2026 edition of the CIBSE Domestic Heating Design Guide changing the way whole building air permeability value works, this now reflects the air changes much closer to reality. Our new estimating software will start to reflect this from next week so if we are sending out estimates it will be a bit more accurate with the data available to us before any formal manual heat loss assessment has been taken. It will now also allow us to enter the official measurements taken if we perform a building blower test or a pulse test, this would be as accurate as it can be, in many cases the ventilation losses are more than half of all the heat loss in a property, get this bit wrong and its a large error. 

For around 3 years or so we have been editing these values manually to between 0.6 - 1 ACH once we have assessed the property as we noticed the ACH were always much lower than calculated by the guides. We became better at this job by taking risk and pushing the boundaries, on a couple of occasions we fell short but you will be pleased to know we stepped up and replaced the ASHP with the next model up at no extra cost and recorded our findings. We now share this knowledge via mentoring and support to other selected installers.

Posted by: @springswood

↑

Thanks for a great podcast that answered pretty much all my A2A questions. Here's some of my reflections on all the new information. 

I was a little surprised @ashp-bobba was so definite that A2A isn't as good an option as A2W for most UK homes. Add to that what he says about how it's not worth much to installers to get the MCS certification for the £2,500 BUS grant and it doesn't sound like A2A is going to be the next big thing. I'm wondering if I should change my vote.

What he said about managing air flows (and expectations about them) was particularly useful - if discouraging. Similarly with cleaning the indoor units.

I did think it was a bit pessimistic about cooling bedrooms. In the UK, even in a heat wave, night-time temperatures aren't so high that a/c would need to run all night. At least not here in Yorkshire. So for me, where the problem is that my bedroom is west facing and warms up a lot through the afternoon and evening I would work to cool the room beforehand then turn the a/c off.

I appreciated the technical point about how A2A systems do not modulate to low outputs as well as A2W. I'd suspected that. For instance I've looked at the Daikin Multi+. With 4 indoor units plus hot water it would do all of my 2 bed house except the bathroom - which is pretty good. It's rated at 9kW max and 3kW minimum (from a data sheet I can't now find).

So for me a Multi+ is oversized. My heat loss is calculated at 4.8kW, for 21C indoors -3C outside. Measuring it (from gas consumption and degree days) I think the reality is more like 3.6kW. Factor in that I have an indoor temperature of 18C and the maximum load comes down to 4.2 and 3.2 kW. Put another way I'd rarely draw even the minimum amount of power.

On efficiency it was good to get comparable COP numbers. They suggest A2A would use around 75-80% of the energy of A2W. Which seems a significant difference. Especially if you use a lot of heat. I don't, only around 3,000 kWh per year.

As for installation costs. I've only had two A2W quotes. Heat Geek (not zero disrupt) came in at £17, Octopus at £11.5k that's £9.5k and £4k after a £7,500 BUS grant. The rough estimate from the Daikin website for a Multi+ system for my house was £9, or £6.5k after a £2,500 BUS grant if and when it happens. That does include adding a water cylinder as I currently have a combi boiler.

So maybe I need to get a better A2W quote. To be fair to Octopus I only decided not to go with them because of problems with where they wanted to put the water tank.

Finally it really is great to have such high quality information available.

I apologise if the point came across that it was so definite that A2A is not as good as the A2W, obviously its subject to the design and application but in my view for whole heating where a customer has a heating circuit already in place and the assessment is purely based on value for money or cheapest option then its unlikely A2A will work out the cheaper option but if you want cooling as well then thats the way to go and pay the extra. 

My point about the bedrooms is customers often want these systems invisible and silent so the placement is a big design consideration, we can hide them like ducted units in the loft and all you see are grilles but they are not silent, the point I wanted to raise is if you want to operate them during sleeping hrs then you must be aware they emit around 20db of noise continuously and can often have a slight (I am not sure how to write this: swoooshing noise) of the refrigerant inside the coil when working, not to mention the 3rd party pumps that are sometimes attached to remove condensate, now they are noisy. I would say domestically bedroom installations are the most popular installation for us, most customers do not want to pay double the money for the loft version so wall units are chosen more often, I also agree with your point, it is possible to cool the room before sleep and then turn the system off.

Only Multi units do not modulate down as low as A2W, I think at the time our example was a multi system covering multiple rooms, splits will do the same as A2W modulation but you cannot have multiple outdoor units without planning permission so multi systems are much more popular for multiple rooms.

I am glad you found a large part of the interview to contain quality information. 

KR

Newbie and my brain hurts.

1970s detached extended bungalow, initially working towards air to water, calculated heat loss 5.37.

Two distinct zones  between original building older cavity wall insulation and good loft insulation, and extension/living area which has underfloor insulation new cavity wall insulation and a lot of glass and solar gain(it’s a vaulted garden room with velux and two mostly glass walls ).

Eleven rooms all with rads mostly doubles all on microbore with the garden room rads piped under the concrete floor.

Heating is oil, grant blueflame + OSO cylinder all 4 years old and pretty efficient.

We want to get rid of oil and until recently thought air to water(vaillant7/8kWh). Cost to us between £7/8k using BUS including re pipe on all but the stuff in the concrete, pump, new cylinder(which pains me) and all controls etc.

However by chance I came across several YouTubers with actual living experience of air to air and I’m very interested and due to have a survey in 2 weeks.

Accepting that every property is different the additional functionality of heat, cool, humidity control and air filtration particularly with multi room as opposed ducted looks initially like a no brainer.

Wouldn’t use the grant and keep the current heating as a back up until we knew it worked over a least a year, no revising plumbing, no ripping out the current boiler and cylinder no needing a physics degree to get/keep air to water running properly. If we eventually decommission the oil we will just use immersion for hot water.

Sorry if all this seems simplistic but I want my future needs to be simple.

We are having a 12/13kWh solar array +11/12kWh storage installed in a few weeks.

We live in Pembrokeshire so other than the occasional Beast from the East event, very temperate climate.

At the moment air to air looks the firm favourite.

Posted by: @batpred

↑

Posted by: @springswood

↑

As for installation costs. I've only had two A2W quotes. Heat Geek (not zero disrupt) came in at £17, Octopus at £11.5k that's £9.5k and £4k after a £7,500 BUS grant. The rough estimate from the Daikin website for a Multi+ system for my house was £9, or £6.5k after a £2,500 BUS grant if and when it happens. That does include adding a water cylinder as I currently have a combi boiler.

A2A would only work in smaller properties (and I think there's a consultation running to eventually allow more people in flats to have them).

That's possibly the most bogus heatpump comment I've heard this week 8)

Things we generally heat (and cool) with air 2 air heatpumps

- Office blocks

- Hotels

- Warehouses

- Loading docks

- Data centres

And in our case it's a giant 1860s solid stone victorian pile and about 500m2 of floor space. There's not a lot you can't heat with an air to air system. In fact even one of the heat geek people who had done big installs before told us A2W couldn't do the house viably because we'd need to lose half a room to some giant heatpumps, and totally replumb what is a listed building and then it would still suck.

There are buildings A2A cannot do well, but it's not about size. Where A2A falls apart on cost is if you have a lot of small rooms that are relatively well insulated from one another and where you can't just leave a load of doors open. In those cases you need an awful lot of pipework or ducting and a lot of head units. Classic examples are things like 4 storey HMOs with self closing fire doors to every room off the landings.

Flats are a tricky one. If you want to heat each flat separately then an air/air system is brilliant. However in many cases (assuming it was designed right so doesn't overheat - which is sadly rare for newer builds) the best way to heat a block of flats is usually a single big heatpump system for the building, and that's often easier to do well with radiators and water. Unfortunately it's also a nightmare because the UK doesn't yet have a lot of the needed rules and policy in place to make it work well.

@ashp-bobba I thought England now allowed 2 on a detached house ?