@jamespa trvs are electronic and I have the upstairs limiting to 18c at night but I’m going to try removing this and have them all open, even the downstairs ones that are set to about 22c for solar gain limiting.

heat pump is a Vaillant 7kw arotherm plus.

I had the senso on expanded, which is how the installer set it, I’ve tried inactive and now on active and doesn’t seem to make much difference. 20c setpoint.

House heat loss via cibse rules was 7.35kw.

That was a snapshot of just the night and some of this morning. Will post another pic below which will show what it’s done from 00:00 to 18:45. There is a dip at 14:00 for hot water.

house is about 125m2. 1967 Alison build with cavity filled and about 300mm loft. It’s a 3 bed that’s been extended over the garage so I have an insulated outside wall inside the house between about 1/3 of the house. Some of the ceilings in the 1/3 are also insulated, as are some of the ground floors due to garage conversion. House sits north south almost with a slight tilt to the west, mean we get crazy solar gain.

no buffer or llh just open loop.

242hrs and 350 on/off cycles.

@jamespa

The first chart in your post above shows average 0.25kWh consumption.  The second confirms that the unit is delivering less than 1kW output.  Yet the OAT is 10-14 C.  The output (and thus presumably the demand) is rather lower than one would expect for a 7kW house.

My heat pump, also the 7kW Vaillant, has been averaging about 2kW output over the same period albeit at slightly colder night time temperatures but slightly warmer daytime temperatures.  I know from extensive data that my heat pump is well matched to the house loss.  

The most likely explanation for not so stunning COP is that its oversized, and thus cycling a lot.  All the data, including your description of the house, is consistent with this diagnosis except possibly 242hrs and 350 on/off cycles.  That still says that the average cycle time is only 40mins, which suggests that many 'on; times may be 20 mins or less.  That may not be enough to get up to full performance.  From the description of your house and the consumption/production I would guess its likely more like 4kW than 7kW, maybe even less.  There is a possibility that load is being reduced due to solar gain, you mention its quite high, but that wouldn't explain 0.3kW at night when its 10C outside and there is on solar gain, unless the house were overheated from the night before.

Do you have any monitoring other than through the app?  Looking at flow temp or consumption vs time on a granularity sub 10mins in the time domain would confirm what its doing, but you need home assistant, openerergy monitor or similar to do this.

Was this a retrofit and if so what was your previous oil/gas consumption.  This will give another data point.

The good news is that with such a low loss house you aren't going to be spending that much on heating.  Also it may be more efficient than a smaller one would be when its really cold, so SCOP may not suffer much or at all.

It may be that part-time heating in the shoulder season (but only in the shoulder season) works better for you, because it will mean that it has to work harder to deliver the required energy, when otherwise it would be cycling madly.  It might be worth trying this.  You could even try a forced 2 hrs on 2 hrs off regime.  Without monitoring its going to be difficult to be absolutely certain.  You can do enough with Home Assistant (which is free) to see whats going on if you fancied setting that up.  Alternatively the Havenwise algorithm may deal with oversized heat pumps (there are plenty around), @hcas perhaps you could comment.

Certainly open up the TRVs, its unlikely to help if the TRVs are interrupting the natural cycle by switching the demand on and off.

Other possible things you could try are to increase the compressor hysteresis in heating mode.  This sets how far above the target FT the pump can go before switching the compressor off.  It defaults to 7C, if you increased it to 12C or even the maximum 15C it will increase the cycle on time at the expense of additional flow temp deviation.  You could also consider increasing the 'compressor start heat from' parameter.  This controls when the compressor switches back on after it has cycled off.  Its measured in degree minutes and defaults to -60 which means that the compressor will switch back on when the sum of (actual flow temp - target flow temp) * time in mins is less than -60.  If you were to make this -120 it will be off for longer, the house will cool more, so it will need more energy to get it back up to temp.  These are both in the installer menu in the HPAI - so care needed.  Difficult to see the effects of these changes unambiguously without some monitoring however.

Hope that helps, feel free to post some more.  I definitely think the most likely explanation is that the heat pump is cycling more than ideal because it has a significantly greater capacity than is strictly required, so the question is how best to manage that situation, but could yet be wrong if eg solar gain is significant.  It would help to post a weeks worth of data rather than just a day.

@jamespa one thing ive discovered is the two plinth heaters in the kitchen we bought that have a FT stat set at 35c rarely come on unless early morning when it like 6c outside. this is causing the kitchen cupboard above to warm up, not good for food storage. i might ask thermix if its possible to adjust it to maybe 25c or 30c. this would help lower the cupboard temp and maybe soak up some of the heat pump. they have a remote room stat that also controls if they turn on there fans. ive set that to 25c ish today to make sure they come on. suppose when you asked about thermostats id forgotten that one.

one of the surveys i had was 5.5kw but i think they missed half the lounge or something. the 7.35kw was done by the Heat Geek installer so assumed should be accurate, though one room where they said fit a type 22 1200x600mm rad when i worked out the output against the heat loss it was near double, so might have to move it to the hall.

one bedroom upstairs ive noticed only manages about 19.5c, but thats running a very large old single panel rad that doesnt have fins even.

we also tend to have most of the doors in the house open so some heat moves that way between rooms.

last thing ive just thought is the upstairs bathroom doesnt have a rad currently as were going to renovate it. its a tiny space though, L shaped and about 2mx2m.

i will post a pic of the power draw last night on the powerwall 3 which does show a series of spikes about once per hour.

ill have a look at home assistant as i have that setup for my heat pump tumble dryer and induction hob.

@jamespa

Posted by: @adamk

↑

one thing ive discovered is the two plinth heaters in the kitchen we bought that have a FT stat set at 35c rarely come on unless early morning when it like 6c outside. this is causing the kitchen cupboard above to warm up, not good for food storage. i might ask thermix if its possible to adjust it to maybe 25c or 30c. this would help lower the cupboard temp and maybe soak up some of the heat pump.

My single Mitsubishi ilife2, which is otherwise perfection, suffers from exactly this problem and the threshold (30C) isn't adjustable.  I've just (yesterday) bought a couple of connectors so I can reversibly splice a resistor in series with the flow temp thermistor to fool the controller into thinking that the water is warmer than it actually is.  In my case the 'fooling' needs to be reversibl,e because the fancoil also has a cooling threshold, and I don't want to change that.  Im not sure fancoil makers have fully caught up with weather compensated systems and how low the flow temperature may go!

The spikes in your power plot confirm what I thought, although its not as bad as I feared it might be.  By way of comparison here is my FT plot, somewhat longer on periods than yours in relation to the 'off' periods.  The long off daily is because its still in extended mode and so switches off altogether after a day of solar gain plus the evening cook.  This will go soon as I operate on pure WC during most of the season and only use extended mode either end to deal with solar gain.  The big spikes are of course DHW including the legionella cycle.

Posted by: @adamk

↑

the 7.35kw was done by the Heat Geek installer so assumed should be accurate,

TBH I don't trust any survey that has not somehow been sense checked, particularly with a house like yours (and mine) that has been subjected to various retrofits at various times.  I think its fair to say that the reports here confirm that. 

My house is an extreme example.  Although 1930s solid wall originally, is probably of a similar insulation standard to yours (partial IWI and modern extensions have improved most but not all of the walls, floors are totally uninsulated and loft like yours has 300mm).  Its 200sq m and 7kW verified several ways.  I had 2 full 3 hr surveys done, both of which came out at 16kW.  Correcting for fabric errors and double counting of room to room losses in the surveys (ie doing my own calculation) gets to 10.5kW, the remaining gap must be due to ACH which defaults to 2-3 and must in fact be 0.5-1, a figure which several installers have subsequently confirmed is what is commonly measured when a blower test is done. 

If you do have your fossil fuel consumption and its in the region of 10MWh/yr this would confirm that your house loss is more like 4kW than 7, but even without that data I think the evidence is pretty clear.  I suspect that if you can fix your fancoils and maybe tweak the FT hysteresis you will get some improvement and as I say it might not affect SCOP too much because at peak season you should be operating at more or less the most favourable part of the compressor curve.  Part time heating in the shoulder may also be worth playing with, reducing (ie making more negative) the 'compressor start heat from' parameter will effectively force this perhaps with an hourly or two hourly cycle, or you can experiment with programming it.  The other approach of course is to add a volumiser.  If you have space for a 100-200l tank somewhere this will extend on times quite considerably, which should help with COP.  

For HA the myVaillant integration will pick up most of the information reported or adjustable in the app.  Annoyingly there a couple of data points, in particular instantaneous power, that it doesnt pick up.  It works by polling the same website poled by the app, so it must be possible, one day if I'm very bored I might look at the code, but its more likely someone else will fix this first.

Hope thats useful

I feel like ive been moved. Whilst it’s easy enough for me to find where my post went I’m not sure why my separate questions keep getting moved to my original post about finding an installer.

@jamespa I think your the chap I saw on one of Mars yt episodes as your issue with heat loss matches. One question, does your 7kw Vaillant have the same vibration issue mine has? I’m guessing not otherwise you would have mentioned it.

one thing I don’t get is why it cycles, ie if I set it to inactive so it’s pure wc surely it would just run continuously and the house would get to a balance point?

on the plinth heaters, I’m going to call them Monday to see if I can reduce the ft switch on. It did say somewhere they can turn on as low as 30c, but then later says 35c. Hopefully it’s adjustable otherwise I might have to hack something to override the ft stat.

I did have some nice data on gas usage but Octopus deleted it when I had the gas meter removed. I think it was 8.5mwh but that was with an un heated integral garage which is now a room and also was heavily zoned and timed so it wasn’t running 24/7. Also think that included hot water heating once per day that used around 10kwh of gas.