@ian33a — a tricky one. Any or all of your hunches may be correct. Can you get the raw data Installers B and C used for their fabric heat losses, and look for discrepancies there? The air change differences are already explained, and I favour Installer C assessment.

I would also ask both installers B and C if they can explain the difference in results given they used the same method, and then scrutinise their explanations for anomalies. 

How confident are you in your 11.3 kW loss based on your past oil boiler and wood burner use? Which way are any errors likely to have pushed the result?

Of course getting the heat loss estimate as accurate as possible is very important, but installer character and style also matters. You need someone you can work with. On the basis of the information given, on that basis perhaps Installer C is a better bet? But the final choice has to be yours, we can only suggest things to consider.

You probably know this, but when going with a Heat Geek trained installer that is not operating through them, you could end up taking the risk of a not correctly sized pump. 

With Heat Geek they give you a SCOP guarantee for heating and hot water (hot water being a very small share of energy use). So if your figure is below their guarantee (exact figure depends on what they calculate your typical flow temperature will be), they will swap the pump at their own cost. 

If you have rads undersized in a room you care about, you will probably end up turning the temp up. Ultimately, the pump will end up operating at a higher temperature to keep it comfortable and so run with a lower efficiency.. 

We made sure we increased rad output on the rooms that were a bit low so that the whole thing will run more efficiently. So I would look at the specific advice they give in terms of upgrades (based on loss calculations, ideally very close to room level detail). They should calculate a flow design temperature at -2C external temp (for SE, less depending where you are in the country). This is what MCS requires them to do before they complete a proposal. I know many just get data from your epc (like spruce does) and give you an estimate. But after this, they ask for £250-500 for a proper heat loss report. For a detailed heat loss report, I found it money well spent. 

@batpred I couldn't get Heat Geek interested on account of the perceived high heat loss of our property. I had to seek out installers, instead, who would come out and look. Interestingly, Installer A is a Heat Geek Master and Installer C says that he has been through the whole training and was affiliated to them. Both have left the affiliation because of cost and being forced to advocate certain hardware and business practices.

Almost all of our house, at present, is under floor heated. We have two lounges which only have wood burners and we have asked all installers to add radiators in those room (UFH impractical) and in a third bedroom, which runs a little cold, we have also asked them to install an additional radiator. All of the UFH rooms (aside from the slightly cold one) run nicely warm with a flow temperature of around 45-48. In the two lounges we have asked for radiators which will drive the room stated heat losses so that the wood burner become an option. Installer B has a design temperature of 0 (which I think is too high). Installer C has a design temperature of -3 (which seems correct). My HeatPunk analysis is also -3. 

@cathodeRay - I pulled all of their data out of their reports and, as is often the case, the numbers are all over the place. The trend though, is as I listed in the opening post, Installer B tends to be more pessimistic generally and only a few rooms have a heat loss less than installer C.

I got a load of help from JamesPA to get an accurate assessment of our perceived energy usage. While I don't have per day data for oil, I have four years worth of figures and the average is taken from that. The wood burners are more of an estimate based upon typical usage but they contribute less than 15% toward the overall estimate so become somewhat second order, but I am reasonably confident. Average oil usage per year is 2,500 litres and I'm very happy for people to double check (and correct) my mathematics !!

@ian33a  Given the nature of your house I think you did well to get estimates that close.  I have to agree with this sentiment:

Posted by: @cathoderay

↑

Of course getting the heat loss estimate as accurate as possible is very important, but installer character and style also matters. You need someone you can work with. On the basis of the information given, on that basis perhaps Installer C is a better bet? But the final choice has to be yours, we can only suggest things to consider.

and also

Posted by: @cathoderay

↑

How confident are you in your 11.3 kW loss based on your past oil boiler and wood burner use? Which way are any errors likely to have pushed the result?

The other things I would add are these: 

1. The key design issue is where loss differences result in a material design change.  So the difference between 10.7 and 11.3kW matters not a jot, you are getting a 12kW pump.  Possibly even 12.3  with a bit of massage depending on the actual (as opposed to sticker) pump output.   However the difference between 11.3 and 14.7 is the difference between a 12kW and a 16kW pump, so really matters.  At this leve,l emitter differences, so long as they are in the right ballpark (or as big as reasonably possible) are a bit academic, adjustment of FT and balancing will always sort these out and if upgrading a couple makes a material difference to FT, its an easy retrofit.  If you can focus on this it may help.

2. If you end up a kW or two under, its not, depending on your attitude to risk, a complete disaster.  It will matter on only a few days of the year.  You have a woodburner and can play games like using the immersion for DHW on coldest days, thereby supplementing the heating.  In the limit you can fit a 3kW auxiliary electric ("Backup") heater.  If you are prepared to take this approach then going with the installer you feel comfortable with, even if he may have undersized, becomes much less of a risk, and anyway he is on the hook to heat your house so should be prepared to do soemthing about any shortfall if it occurs. 

Ultimately its down to you but perhaps the above is of assistance.

Posted by: @ian33a

↑

@batpred I couldn't get Heat Geek interested on account of the perceived high heat loss of our property. I had to seek out installers, instead, who would come out and look. Interestingly, Installer A is a Heat Geek Master and Installer C says that he has been through the whole training and was affiliated to them. Both have left the affiliation because of cost and being forced to advocate certain hardware and business practices.

That makes sense. Heat Geek seems to adjust their business based on their operational data and are also strategic in their options (like when their introduced the Arotherm Pro). The Head office customer interface is not great. not much explanation is given, I had the feeling they listen to customers, spew out their AI aided replies and then do what they want (perhaps the suggestions will only help other customers). But when we engaged them, we already knew what we needed and the installer was very professional and detail oriented. HG handled the paperwork and all the back office. I can confirm it only took the installer a few hours with their software to do the commissioning tests, etc. 

Your home heating is not the typical case they seem to focus on (if any are..), so it would seem someone with the HG training but no longer tied to their hardware choices may be the best..

Posted by: @ian33a

↑

Average oil usage per year is 2,500 litres and I'm very happy for people to double check (and correct) my mathematics !!

Can you post your mathematics and all the data you used? Even things like how you ran the oil heating eg always on or timed? Was the indoor air temperature (IAT) constant? I ask because getting a heat loss from oil usage is a far from simple reliable exercise (I also had oil heating before I had a heat pump, and I struggled!) and looking into how to improve the process for you and indeed others make a lot of sense.  

Thank you, everyone one, again for your very helpful replies.

I had a lot of confidence in installer A. They came out, were professional, started off answering questions in lay terms until they realised that I had done my research and understood far more than they imagined I would. After that we were able to talk in great technical detail. I was more and more annoyed as the days went by with no proposal. They asked for a couple of weeks before they could turn to my job sizing, which I accepted. In the end it took more than double that time and, only then, by persuasive pushing did I get a proposal. I got the impression that they came up with a number and, literally, doubled it. I kill the conversation flat at that point.

Installer C came out next. A reasonable blend between confidence and professionalism. Again, happy to talk in technical detail and willing to listen to and accept my own proposals. There was an obvious preference to Vaillant, but I can accept that too, provided it can do the job. He suggested fancoil K2 radiators, which would have been quite big and I told him that we didn't want the fan noise and would rather pay a bit more for aluminium heat exchanger units and cut the size down - again accepted. When the proposal came through I asked for more in depth detail about the heat loss, which was promptly sent. Pricewise, it was encouraging. Installation could not be for a few months which, again, shows that they are busy. There was no charge for the survey - which was done with Lidar and Spruce.

Installer B came last. We knew that they were a consultancy but didn't know the dynamics of how a successful acceptance of proposal would work. Again, Lidar and Spruce, but a lot more time taken to do the survey. Professional though. We had to pay for this survey, which is fine. We discovered that, if we accepted the quote, they would put is in touch with one of their preferred installers and that the installer would take full control of everything from there using the consultancy design. If we did go with them they would visit again, with the installer, before hand over of the project. Up until the CIBSE respecification they had been using their own Excel worksheets for heat loss. They'd move to Spruce two weeks earlier - there was confidence in using the software but the lack of experience with using it rang distant alarm bells. It was clear that they were following MCS to the letter, or claiming to. The consultant was adamant that several of the rooms that we considered to be plenty warm enough in the four years we have lived here needed upgrading. One of them, which is a bedroom but which only has a turbo trainer had to be treated as a study and needed a higher room temperature. What? If we moved, it would likely revert to a bedroom! Equally, another room, which is currently being used as craft room (even though it is a bedroom) was seen as bedroom and could be measured using a lower temperature. This room was considered adequate in heat provision even though we had told them that the room was often cold and we wanted an emitter added. Our main bedroom, always nice and warm, was considered inadequate and needed an additional radiator. Bear in mind that all of these rooms have UFH with a flow temperature currently in the upper 40's. We also said that we wanted low profile aluminium heat exchanger radiators added in our lounges, they specified large steel K3's. Their heat loss was about 13.7 but their proposal had several wriggle out of responsibility clauses. The consultant was very capable as an assessor but I doubted if he had ever used a pipe cutter. That he was handing over all responsibility for the installation to another company was also something of a louder alarm bell. The final proposal was a Grant 15KW unit, but with a 250 litre tank (300 in the case of installer C) and with those K3's. Price wise, they were pretty similar.

And I do agree wholeheartedly with what JamesPA is saying. A few hundred watts of difference is nothing but several KW is more concerning. That installer C has what I think is a more tailored for us proposal and with decent kit is good, as is the fact that his name is written over the door and he has to deliver is also good. That installer B is suggesting larger sized heat pumps and his numbers are closer to installer A is a worry against installer B. But my own heat loss survey comes out pretty much in the middle of both and my estimation based upon oil and wood used is pretty close to installer B would suggest that B may be more on the nail - if my numbers are right. 

To answer @CathodeRay, I used a calculation that I think I got from somebody on here:

Heat loss assumed to equal heat created to balance that loss.

Approx heat loss = Contribution from oil                                                                         + Contribution from wood burner

Approx heat loss = ( (Annual (averaged) oil usage * boiler efficiency (0.9) * 10.35)/2300) +  Estimate of heat produced over heating period * Number of days used

Approx heat loss = ( (2402 * 0.9 * 10.35)/2300)                                                            +  4 (hours) *40% average power generated * 4 KW (maximum heat) * 90 (days)

Boiler efficiency of 90 is probably very optimistic. 2300 is a factor that one forum member considers a reasonable multiplier as I don't have a per day usage measurement.

Obviously, with a less efficient boiler the energy created drops - 80% gives something like 10.22 KW whereas 90% gives 11.31 KW.

Posted by: @ian33a

↑

. He suggested fancoil K2 radiators, which would have been quite big and I told him that we didn't want the fan noise and would rather pay a bit more for aluminium heat exchanger units and cut the size down -

I have one Mitsubishi Ilife2-slim fancoil to solve a problem that otherwise would have required either a large K3 or a superlarge K2.  Its really quite dimunitive, only 130mm thick and only just over 1m wide, and the fan modulates according to load.  Its in the 'living room' that we use constantly and not at all disturbing.  I wish I had specified a couple more for the adjacent room which also has large radiators. 

I have only one moan about it, namely that the controller switches the fan off when the FT goes below 30, and this threshold is not adjustable.  I run down to 27.  A resistor wired in parallel with the (NTC thermistor) temperature sensor fixes the problem, but really the threshold should be adjustable not fixed.  I would say its well worth considering if you have a problem area.  There are other similar products of course.

Posted by: @ian33a

↑

I used a calculation that I think I got from somebody on here

It is always a good idea to quote sources, some might say it is essential. This time Wattson did get the answer, and it turns out that it is our very own @jamespa who came up with the 2000-3000 factor. However, it was not intended to be used as bottom line heat loss calculator, rather as a sanity check against "silly heat loss surveys which still appear to be common". @jamespa freely admits that he invented the numbers, and they had "only a tiny amount of science behind [them]". The discussion can be found in and around this post. 

But if he invented them, and the methodology, how did he invent them? he will of course tell us, but I think it is for oil heating very loosely:

heat loss = (litres burnt pa x calorific value x boiler efficiency)/guesstimate of the number of hours the heating was on (I think it is in hours because often the 2000-3000 number has an h after it eg 2900h)

It is as well in passing to be clear about how we describe heat losses. Almost invariably, we use kW, which is an instantaneous measurement of power, not energy, and so is not really a loss at all. The loss has to happen over time, and the time we normally use is an hour, so when we say the heat loss is 10 kW, what we actually mean is the house will loose 10 kWh of heat energy over an hour, and that is then balanced by a heat source that delivers 10 kWh of energy over the hour, and to do that it needs to run at a power of 10 kW for that hour, hence the 10 kW number.

It is worth noting that even 'the professionals' get this wrong. Only this morning I came across this on a "Heating Oil Usage Calculator" website:

"Before we dive into our heating oil usage calculator, we need a quick science lesson – starting with energy.

Watts measures energy. The most common measurement is kilowatts (kW), which is equal to 1000 Watts. It takes many Watts to efficiently heat a home."

Oh dear...but that's what happens when you 'leave it to the professionals' (albeit CH oil suppliers in this case, so they are not claiming to be 'engineers', but if you claim to be giving a science lesson, please get the science right).   

Going back to the formula, the only number with reasonable certainty attached to it is the calorific value, which is around 10.35 kWh/litre, but even that is subject to a caveat, is that the gross (ideal world) or net (real world) value? it is probably the latter...

The other numbers are subject to more or less error. If, as indeed I often did, you fill the tank in the summer, and it is filled to the top, then the amount put in will be close to the amount of fuel you burnt during the previous heating season (assuming you you an immersion heater for summer DHW, as I did). But even then it will vary year on year, depending on whether it was a mild or cold winter. Nonetheless, I used, on average, around 1100 litres per heating season. 

Then we come to boiler efficiency. My boiler was fitted in 1994, what was its efficiency in say 2014, when it was twenty years old? Who knows, but a google search finds plenty of guesstimates that the efficiency is around 60-70%. Let's say 65%.

That means that when I had oil CH, I delivered around 1100 x 10.35 x 0.65 = 7400 kKw to the house each heating season.

But to achieve what? Pre-retirement, I ran the CH morning and evening during the week, and all day at the weekends. This means the indoor air temperature (IAT) was far from constant, indeed condensation on the windows overnight confirmed that. What was the mean IAT? I don't know... But what I do know is that I run the heat pump differently. It can't do the fast reheats that a fossil fuel boiler can do, and so it is on most of or all of the time (the low and steady principle). What this means in practice is that a heat loss estimate derived from oil use is almost never going to estimate heat loss when using a heat pump, because the heating regimes are different (having said that, if the fossil fuel boiler was always on, it might get closer, but most folk run them on a timer).

That said, let us nonetheless push the formula which we now have little confidence in to its logical conclusion. We need the magic correction factor, the divisor in the equation. My hunch is this number is derived by a sort of trial and error. Indeed, we can even do that with my numbers, now that I know my actual heat loss, as determined by monitoring my heat pump output. It is about 10 kW, so in my case the magic number is 740h (7400/740 = 10).

But let's take another house, say yours. Given an annual oil usage of say 2400 litres, and a similar older boiler efficiency of 70%, and for the purposes of this discussion a 'known' heat loss of 12 kW, your heat delivered to the house using oil was 2400 x 10.35 x 0.7 = 17388. To get this to 12 kW, the magic number needs to be 1449h (17388/1449 = 12). The magic numbers vary by a factor of two, which pretty much tells us there is not such thing as a universal magic number (except, perhaps, 42).

Which leads me to the conclusion I always come back to, there is, most unfortunately, no practical way of estimating heat loss from oil usage. 

PS @jamespa, you know full well this is not a pop at you, instead it is a pop at the idea oil usage can be used to determine heat loss    

Posted by: @cathoderay

↑

@jamespa freely admits that he invented the numbers, and they had "only a tiny amount of science behind [them]". The discussion can be found in and around this post. 

To clarify - the generally accepted divisor for SE England is ~2900 which is scientifically derived from a degree day calculation based on the SE UK typical winter climate.  I can't recall how much the divisor varies by location in the UK but it's not much except at the climatic extremes.  I did once check the calculation and I am satisfied that it's sufficiently sound to use as a check if you don't have  more granular consumption data.

What I invented were the bounds around this for a sanity check.  This was based largely on reasonableness for oversizing. There is a sound argument for making the lower bound something closer to 2500 if it's the heat loss calculation that is being sanity checked, but the granularity of heat pump capacity is an argument for the lower figure.

Posted by: @jamespa

↑

To clarify - the generally accepted divisor for SE England is ~2900 which is scientifically derived from a degree day calculation based on the SE UK typical winter climate .  I did once check the calculation and I am satisfied that it's sufficiently sound to use as a check if you don't have  more granular consumption data.

When you say 'scientifically derived', does this mean from first principles (as you imply, using degree days, and if so what is the actual derivation, and how much whatiffery was involved), or empirically, ie take a bunch of houses, see what you need to divide the annual heat delivered to the house by to get something close to the heat pump rating? But my real concern, as shown by the examples in my post, is that for oil heating this number is far from a constant number. Might as well read the tea leaves to get the number...