Firstly you have come to the right place and its good that you have posted.

15kW for a 225 sq m 1990s house does initially feel like extreme overkiill but... 3000 litres = 30MWh corresponding to most likely (depending on your heating pattern) 10kW perhaps a bit more, perhaps a bit less depending on your heating pattern etc, so its not completely over the top as the house currently stands.  However your boiler may be only 75% efficient as its old, so the loss could be as low as 8kW and the uninsulated roof will be contributing a fair proportion of this.

Even as the house is you would quite likely get away with a 11/12kW and insulating the roof will make a massive difference, quite likely reducing the loss to 8kW at which point 15kW is oversized by sufficient margin to cause concern.

Unless there is something very odd about your house you do NOT want or need a buffer tank (ie a tank connected to both flow and return)!  A volumiser (connected to either flow or return) is OK, but probably unnecessary if the heat pump is correctly sized.  Check what they mean by 'buffer tank' and why they think one is necessary, the term is used a bit loosely

If you oil boiler will soldier on for another season I would be very tempted to insulate and then get some more quotes.  If you had gas I would say also collect half hourly meter readings this season, but you dont.  This being the case it might well be worth fitting a heat meter so you log heat generated to get a more accurate handle on loss, and also ensuring the property is warm most of the time to get an accurate read out.  The cost of the heat meter will easily be recovered in savings on the difference in heat pump price, and you can now get devices that log 15 months worth of data eg this one (that's not a recommendation, I have no experience of such devices!)

If your oil boiler is going to die instantly then you do have a bit of a dilemma.  If you can find someone (Heat Geek?) that will quote for a 12kW heat pump and no buffer tank that would very likely be OK.  If you are stuck with 16kW and buffer tank there is a fair probability that your performance will be disappointing unless it is set up well - but likely no worse than your old oil boiler.

Another thing you could do, if you feel able, is your own heat loss calculations.  This would tell you a lot about the contribution of the various elements enabling you to make more intelligent choices about insulation and indeed play some what-ifs.  There are various online tools available to do this, for example heatpunk, openheatloss etc

In summary (and assuming your house is warm most of the time currently so your consumption represents your loss reasonably well) settling now for a 15kW heat pump and buffer is, IMHO, a last resort.  If you can either get a 11-12kW no buffer quote now or, better still, defer a season and get the roof insulated and some actual loss measurements, that would be far preferable.

Happy to answer more questions and others may have alternative ideas.

Posted by: @seaviews

The house has double-glazed windows (installed in 1990), but there is virtually no insulation on the upper floor or in the attic (although we do plan to resolve this).

15kW for 225 sqm is about 67W per sqm of floor area. My now well insulated 1970s house has a heat loss of a little less than 6.5kW and is 150 sqm, so about 43W per sqm of floor area. As building standards improved over time, this indicates that it is highly likely that you can make some large improvements in heat loss, and commensurate reductions in your energy usage and bills by taking a fabric-first approach and insulating the house as much as possible before changing the heating system.

The biggest improvements would be from insulating the loft / roof and from ensuring that the cavity walls are well insulated. The cavities may already have insulation as this was usually done in the 1990s, however, if not it needs a highly professional approach and specific materials using due to the proximity to the coast and high exposure to wind driven rain. Your 1990s double glazing will not be particularly well insulating. The whole window U-value (a measure of heat loss) will be around 2.7 to 3.1. By comparison, new double glazing has a U-value of 1.2 to 1.4 to meet building standards. In other words you could more than halve the heat loss through your windows by fitting new double glazing. Though expensive, you may also want to consider that home improvement.

It is possible that by focusing on improvements to insulation first, you could get the heat loss down to around 2/3rds of what it is now. It is likely that 15kW is an overestimate, although perhaps not by a huge amount, so post-insulation you could be looking at a 10kW heat pump or smaller. The other factor is that the U-values for any improvements that you make will be known and evidenced. Hence the degree of overestimation in the smaller post-insulation heat loss figure will also be reduced.

The heat loss figure not only sets the size of the heat pump needed, but it is also the primary factor in radiator sizing. So if you insulate first, then the radiator sizes needed to achieve an economical flow temperature (for example max 45C) on the coldest days will be smaller. This will also reduce costs. Further, the very high water flow rate needed to transfer 15kW at the design flow temperature will also be reduced in proportion to the reduction in the heat loss figure. The reduction in the flow rate needed means that your existing pipework may be just fine, whereas it may not be for 15kW.

In short, there are many good reasons to upgrade the insulation as far as you reasonably can before getting a new heat loss survey and a heat pump.

I would echo the comments made by jamespa.  Insulate as best you can and as is appropriate for the house.  Loft insulation is normally relatively cheap and easy to fi.

I do wonder what any of the contractors have said re existing piping and radiators.  You seem to suggest upgrading radiators can wait.  Heat pumps operate best at much lower temperatures than an oil boiler and both the pipe and radiators need to be sized to emit the heat being generated.  15kW of radiators for a HP are substantially bigger than are required to distribute the heat from an oil boiler.  So if the contractors are happy to utilise the existing radiators without any changes, I would question if that the essential design has been done thoroughly.  Maybe you could explain a little more on what you are being offered?

Posted by: @editor

↑

There's nothing inherently wrong with upgrading emitters in stages, but you should understand the potential running-cost consequences before agreeing to that approach

@seaviews

Besides the ongoing compromise in efficiency by keeping radiators which are borderline.

As far as I'm aware, at present replacement radiators and the work fitting them as part of the HP installation would benefit from the saving in VAT.

Replacements after the fact would incur the standard VAT rates for the parts and labour.

Of course, the radiators may actually be large enough to work effectively with a low flow temperature and high flow rate. Which could be true at present, but more and more likely if you can easily reduce the required outputs by installing insulation.

Besides knowing they are a thing, I haven't looked into them. But a hybrid heatpump, with the oil boiler, may be a good solution here. (I'm sure others on here will have some experience)

If fabric upgrades are definitely planned but not immediately feasible, finding a HP which would suit the post upgrade output requirements (would allow the removal of the boiler at some point), but is capable of running alongside the boiler as a hybrid solution, where the boiler is called upon to suplement the HP output as needed.

Posted by: @singlespeed

↑

Besides knowing they are a thing, I haven't looked into them. But a hybrid heatpump, with the oil boiler, may be a good solution here. (I'm sure others on here will have some experience)

If fabric upgrades are definitely planned but not immediately feasible, finding a HP which would suit the post upgrade output requirements (would allow the removal of the boiler at some point), but is capable of running alongside the boiler as a hybrid solution, where the boiler is called upon to suplement the HP output as needed.

Interesting in principle but no BUS, so ~£9000 more expensive and significant extra complexity with ultimately wasted components.  There are better ways to solve the challenges presented by house IMHO - quite possibly either find someone who will fit a 12kW house now or insulate, measure/recalculate, and ultimately get a 7-8kW (numbers guessed, all subject to confirmation when we know a bit more).

If cash flow is a limitation may well even pay to take out a loan to cover early installation of insultation.  

We live in Devon and have been down a similar route. Our property is about 280 square metres, is reasonably well insulated but is old. Our annual oil usage is approximately 2,500 litres and two installers suggested a heat loss which necessitates a 15KW heat pump. My own calculations and that of the installer that we have chosen suggests a 12KW unit is OK. One issue is that many heat pumps come in a 12KW flavour and the next band jumps right up to 15. So, based upon what you have stated in your post, 15 may not be too far off. I also question the buffer tank approach - we've avoided that route and gone for a fully open loop system. 

From what I have seen in quotation, potential installers need to base what they advocate upon what they see on the day. It isn't acceptable practice to say that you plan to insulate in the future and can I have a more favourable heat loss please? So, as others have said, if you want to insulate, go ahead and do it and then get a heat loss survey.

Interestingly with ours, the 15KW advocates had a similar fabric loss to the ones suggesting 12KW. The main reason why the heat loss quotes differed is the perceived ventilation losses: the lower HL installer was more pragmatic about what our ventilation losses really are while the others made assumptions based upon houses similar to ours (even though a lot of work has been done to plug gaps in insulation and surfaces). More insulation and consideration of gaps could help you massively.

If a 12KW proves to be right for us will become evident as we go through winter. But I agree, do the cheaper work to cut your heat loss down and then get a quote based upon that rather than based upon what you have now. Be careful though, just in case the Government decides to reduce the grant.  

I’d agree with all of the above. Especially sequencing. But you may need to move fast to avail yourselves of VAT and BUS.  Our install was massively over specified (16 kWh on a 450m2 floor plan).  House is quite well insulated - but old so not perfect and we have never needed anything like 16kw.  The whole house has never consumed more than 5kw in a 30 minute period - so I think that means 10kw would have been big enough!

i guess if temperatures drop to -25 degrees I might need all 16kw!

@singlespeed I think it’s worth me adding that I’m a big fan for homeowners changing rads if they’re older than 10-15 years as part of heat pump installs because they’ll just improve efficiencies and comfort.

@carpenterstation, can I ask how you arrived at the figure of 5kW in a 30 minute period, and what exactly you were measuring