In short they will be making different assumptions about the fabric leading to different loss estimates.  The only ways to reconcile are either to inspect their calculations, do your own of do a sense check vs current fuel consumption. You do not want seriously oversized so it's worth doing what you can.

In terms of the latter:

What's your current annual gas consumption?  Do you have half hourly readings, does your house stay reasonably warm all the time or does it chill a lot eg at night?

Can you tell us a bit more about the extension to help estimate whether it will reduce, increase or leave loss about the same.

Re DHW if you want a bigger cylinder they will almost certainly oblige, the problem tends to come if you want a smaller one.

@annab 

Regarding the brands, the ones you mention are very common. Mitsubishi make R32 and also the newer R290 models, Vaillant Arotherm are all R290. I stated a preference for R290 as it is the one that eventually will become mandatory. 

Installers often works with various brands. I mostly looked at the controls when choosing... 

In terms of installers, you would want to make sure they are registered with MCS. The external temperature should be very similar for all the heatloss calculations, as they are expected to follow the same standard. It is quite common for installers to charge anything from 250 to 500 or more for a heat loss calculation, but once you have one, they can usually quote based on that (even if, if you decide to progress with them, they will do their own survey). 

@jamespa Thanks for your reply. I have had a really detailed look at the heat loss calculations, and there are minor differences throughout that may contribute to the difference between to middle estimates and the higher ones, plus the fact that they have used different temperatures (-3 to -1.5). I can't see how one company has found the heat loss to be as low as 28w/m2 though, however unfortunately they are the ones who don't want to release their data!

Our annual gas consumption is 8847kW and electric 5116kW. The extension will almost double the size of the house, up to approx 200m2 (ground floor and first floor combined), but will be up to current building regs for the insulation etc. We will also be replacing all the windows and adding redoing the roof/loft insulation. At present it does get cold in winter, but probably more because we don't put the heating on enough!

@batpred Thanks, that is interesting about the refrigerants. Sounds like better to go with one if the r290 options then...

Posted by: @annab

↑

@jamespa Thanks for your reply. I have had a really detailed look at the heat loss calculations, and there are minor differences throughout that may contribute to the difference between to middle estimates and the higher ones, plus the fact that they have used different temperatures (-3 to -1.5). I can't see how one company has found the heat loss to be as low as 28w/m2 though, however unfortunately they are the ones who don't want to release their data!

Our annual gas consumption is 8847kW and electric 5116kW. The extension will almost double the size of the house, up to approx 200m2 (ground floor and first floor combined), but will be up to current building regs for the insulation etc. We will also be replacing all the windows and adding redoing the roof/loft insulation. At present it does get cold in winter, but probably more because we don't put the heating on enough!

Thanks.  If they wont release the data I would be wary of using them.  That said I find 28W/sq m very believable.  Our house is 200sqm originally 1930s solid wall with modest 1980s extensions affecting no more than 30% of the area, incomplete IWI (ie there are some solid walls left) excellent loft insulation and good double glazing, absolutely no floor insulation.  35W/sq m (7kW) measured - 7kW Vaillant a perfect match.

Assuming that the electricity is not used for heating (5116kW seems high) then your gas consumption would, if it were warm 24x7, equate to probably 3kW and almost certainly between 2.9kW and 4.4kW.  If we assume your house cools to say 10C overnight (unlikely) that reduces the 24hr average indoor temperature by ~ 25%, suggesting your loss may be 25% higher ie between 3.6 and 5.5kW.  Its difficult, without evaluating the detail, to estimate what effect doubling the size and raising it to current standards would have, but its going to be much less than double the loss and possibly not much more than at present.  Given you will end up with a 200sqm house that is better insulated than ours, I would say 6kW is a very reasonable number and the 7kW Vaillant likely a very reasonable choice, although the 8.5kW ecodan would also likely be OK.

Some surveyors struggle with fabric upgrades and refuse to consider those that they cant physically verify.  Several who looked at our house would calculate based on solid walls even though most of the house has IWI - we had two full surveys each coming out at 16kW, when the actual figure is 7kW.  Where the work is 'to be done' they struggle even more.  Its essentially backside covering - they reckon that there is no harm in oversizing and its true that the householder will be warm if the heat pump is oversized, albeit they will likely have higher running costs. 

I don't understand why they are using different design temperatures, these are specified in the MCS design document, although there is some ambiguity because they are specified only for major cities so the designer needs to decide what area you are in.

Both the 7kW Vaillant and the 8.5kW Ecodan are reputedly good machines, I personally think the controls on the Vaillant are better and its probably closer to the right capacity, but I may be biased as its the one I have.  There is a new 7kW Vaillant machine out/coming out (the Arotherm pro) its smaller, quieter and apparently cheaper!

More important than the choice of machine is the system design and in particular:

• No buffer tank (a 2 port volumiser in flow or return but not both is OK).  Also no LLH (low loss header) which I think excludes anyone proposing to use the ecodan preplumbed cylinder as that has an LLH fitted.
• No external controls, not even a thermostat unless they are specifically heat pump controls ie Homely, Adia, Passiv, Havenwise only
• Not grossly oversized - anyone suggesting more than the 8.5kW ecodan should certainly be discarded

Then, once its fitted, operate on weather compensation 24x7 with all or most of the trvs (if fitted) fully open (or better still with heads removed) and rads balanced - we can give you further help with this when you get to that stage.

Does that help, are there other differentiators between the quotes to consider (eg some sort of performance guarantee which some installers now give)?

@jamespa Thanks again, really helpful. I should have said - our electricity use is probably high because of the electric car. Also, once the extension is done, we will have about 2/3 (70m2) of UFH in the ground floor, and the rest of the house heated by radiators. 

Can I ask why you would advise against a low loss header? And would the information about UFH/rads change this?

Finally, looking at the CIBSE guide, I think I now understand why installers have used different design temps, as we are pretty much half way between Plymouth and Southampton, which have different temps!

I think I am making progress, as do have it down to 2 installers now!

Thanks, Anna

Posted by: @annab

↑

information

Posted by: @annab

↑

Can I ask why you would advise against a low loss header? And would the information about UFH/rads change this?

Low loss headers and buffers can, and frequently do, cause 'system distortion',  code for mixing of flow and return water thus reducing the flow temperature to the emitters relative to the ft from the heat pump.  In addition they complicate fault diagnosis and are rarely if ever actually necessary in a domestic installation, as the better installers will tell you.  Since ft is everything to efficiency, reducing the ft to the emitters relative to the ft from the heat pump increases running costs costs so, if it isn't necessary, why do it?  It also adds an additional water pump and the question if how to control the secondary pump (frequently botched).  There should be only 1 water pump in the system which is the one in the odu in the vaillant or external I think in the ecodan.

With a mix of rads and ufh you want to design both to work at the same low flow temperature.  Do not be seduced by rads operating at higher temperature and a ufh mixer valve, this is another mechanism to avoid proper system design at the expense of increased running costs and more complex system control.  If the ufh comes with a mixing manifold, which it probably will, you will need to disable it by setting the mix temperature to max, which you can only do if the rads are designed also to operate at a low ft.

And no it does not change the advice.

Posted by: @annab

↑

Can I ask why you would advise against a low loss header? And would the information about UFH/rads change this?

As water flow temperature increases, heat pump COP decreases, therefore you pay more in electricity for the same amount of heat energy if you want it delivered as 40°C water than delivered as 30°C water.  Basically the hotter the water, the harder it is for the machine to push heat into it, which means more work, so it needs more electric energy in order to do it.

Therefore Rule #1 of heat pumps is: you shall not pay extra for hot water, then mix it down with cold water to get a cooler water temperature. This kind of mixing happens in buffer tanks, low loss headers, UFH mixers, etc.

In addition, in very cold weather, the heat pump's max power limit depends on water flow temperature. The hotter the flow, the lower the max power limit. When mixing occurs in an uncontrolled way, for example in a buffer or low loss header with mismatched flow on both sides, which is pretty much unavoidable, you may need the heat pump to output 50°C for the rads to actually get 40°C. Requiring hotter water will reduce the heat pump max power limit when you need it the most. So you pay more in electricity, but if you want the heat pump to actually heat the house on the coldest days, you also pay more for a bigger heat pump to compensate for this reduced power in cold weather. Then this bigger heat pump becomes oversized for clement weather, so it cycles and works less efficiently, doesn't last as long, etc.

In other words, it's a can of worms better left unopened unless there is no other choice.

Thanks again everyone. All very informative. I guess the main dilemma I have now is that the people recommending the system with LLH were recommended directly from this site! Very difficult to know what to do!

Posted by: @annab

↑

...

I guess the main dilemma I have now is that the people recommending the system with LLH were recommended directly from this site!

...

A good recommendation is not something to be dismissed out of hand, of course, so bearing in mind @jamespa's advice you are still in a good position to go back to that company and ask for their reasons for specifying a LLH. Their answer will likely tell you even more about them than what they've included on their quote.

@annab 

@majordennisbloodnok has hit the nail on the head with best next step.  Feel free to post their response here for comment.

Feel free also to post their quote, redacted as you see fit.