Central heating

[Oldgreybeard]

One thing that I noticed when looking at GSHPs (we were originally going to have one with the collector down a borehole) was that there is a fair bit of power used to run the heat collector circulating pump(s). For the sort of power we needed having maybe 50W of additional pump power pretty much did away with the very small difference in efficiency from collector temperature not changing much. Probably doesn't matter much for a heat pump using several kW, but has a noticeable impact when the heat pump draws less than a kW, as ours does most of the time.

[Joeboy]

Well I'm rather chuffed with that. SH had brought home airtemp to 22.4 degs at 07.00hrs this morning. At 15.00 hrs it's only lost 0.2 degs.

All the crack hunting/ gap sealing over the years has paid off as of course has the stored energy from the charging window which closed at 02.30hrs.

[Oldgreybeard]

Warm here, too, 22.6°C in the hall at the moment, a result of the temperature overshoot we always get when the heating first comes on. It'll settle down closer to the 21.5°C set point over the next two or three days I expect, although the forecast here is for a bit of sun tomorrow and Sunday, so we may well end up with the house remaining a bit toasty until the middle of next week. Somehow I doubt the heating is going to fire up again any time soon.

This temperature overshoot is always a bit of a bugger at this time of year, and again in March when the heating goes off. It's the only downside of having a long decrement delay time, the house can very easily get a bit too warm at the changeover points from no heating to heating and vice versa. Still, can't complain, we're a great deal better off in terms of heating costs than most people.

[chris_n]

I'm guessing that the A2A vs GSHP will swap over as temperatures drop.

Currently your A2A is scavenging heat from double-digit outdoor temperatures while the GSHP is working with water that a few degrees.

At some point the A2A will be working with air that is sub-zero while the GSHP will be working with water that is still a few degrees...

The air temperature doesn't make a big difference, does it?

Being a bear of simple brain my thinking is along these lines (and I may well be wrong):

Air at absolute zero contains no heat at all.

Air has a heat capacity of around 700J/kg.K (varies a little bit with temperature).

Air at 10°C (+283K) holds about 198.1kJ/kg.K

Air at -5°C (+268K) holds about 187.6kJ/kg.K

I make that a reduction in the available heat that could be extracted from the air of just 5.3% for an air temperature change from +10°C to -5°C, probably not worth worrying about in the overall scheme of things. This seems to be supported by the fact that ASHPs are very popular in some cold countries. I saw a lot of them when in Finland a few years ago, seemingly working well in sub-zero temperatures.

Plenty of Air Source Heat Pumps here in the Austrian Alps where winter temperatures are regularly lower than -10 with occasional trips to below -20. Most of the newer houses in the village have ASHP. I seem to remember conversations where people claimed the higher humidity in UK caused icing on the condenser when it gets cold but I have no idea if this is true or not. I would like to know the answer as I am considering fitting an A2A unit at my daughter's house next year.

[Oldgreybeard]

Plenty of Air Source Heat Pumps here in the Austrian Alps where winter temperatures are regularly lower than -10 with occasional trips to below -20. Most of the newer houses in the village have ASHP. I seem to remember conversations where people claimed the higher humidity in UK caused icing on the condenser when it gets cold but I have no idea if this is true or not. I would like to know the answer as I am considering fitting an A2A unit at my daughter's house next year.

My experience has been that the humidity issue is the really big thing that hits efficiency. As supplied, our heat pump was running a defrost cycle very frequently if it was run in the early evening on cool, damp, days, when the humidity tends to be high. What would happen is that the evaporator would ice up, the heat pump would detect this and it would then run in reverse for a few minutes, pumping heat out of the house and into the evaporator to melt the ice. It was easy to see this on the display, as the flow temperature would suddenly drop as the heat pump went into reverse.

The solution was to adjust the settings to stop the heat pump working so hard. The most critical setting was the flow temperature. As supplied, this was set to 55°C when the outside temperature was above 5°C, rising to 65°C as the temperature dropped to -5°C, because of the weather compensation feature. The highest temperature we needed was only around 28°C at the UFH, and the UFH thermostatic mixing valve was holding the UFH flow temperature to about 28°C, so it was pointless feeding it with water that was a lot hotter than this. I found the thermostatic mixing valve needed the flow temperature coming in to be a bit warmer, so it would work properly, but not as much as around 25° to 30° warmer, it seemed to work fine with just a 10°C differential.

I spent a fair while trying to get into the installer menu and decode the settings, but after a bit of fiddling around I managed to turn off the weather compensation feature and set the flow temperature from the heat pump to 38°C. This significantly reduced the power the heat pump was using, and as it wasn't having to work anywhere near as hard it stopped icing altogether. I don't think I've seen it ice up at all since adjusting the settings. Made a big difference to the COP, and I think a fair bit of the improvement seen was just down to the lack of icing and the need to run defrost cycles every ten to fifteen minutes in damp weather.

[Gareth J]

My experience has been that the humidity issue is the really big thing that hits efficiency. As supplied, our heat pump was running a defrost cycle very frequently if it was run in the early evening on cool, damp, days, when the humidity tends to be high. What would happen is that the evaporator would ice up, the heat pump would detect this and it would then run in reverse for a few minutes, pumping heat out of the house and into the evaporator to melt the ice. It was easy to see this on the display, as the flow temperature would suddenly drop as the heat pump went into reverse.

The solution was to adjust the settings to stop the heat pump working so hard. The most critical setting was the flow temperature. As supplied, this was set to 55°C when the outside temperature was above 5°C, rising to 65°C as the temperature dropped to -5°C, because of the weather compensation feature. The highest temperature we needed was only around 28°C at the UFH, and the UFH thermostatic mixing valve was holding the UFH flow temperature to about 28°C, so it was pointless feeding it with water that was a lot hotter than this. I found the thermostatic mixing valve needed the flow temperature coming in to be a bit warmer, so it would work properly, but not as much as around 25° to 30° warmer, it seemed to work fine with just a 10°C differential.

I spent a fair while trying to get into the installer menu and decode the settings, but after a bit of fiddling around I managed to turn off the weather compensation feature and set the flow temperature from the heat pump to 38°C. This significantly reduced the power the heat pump was using, and as it wasn't having to work anywhere near as hard it stopped icing altogether. I don't think I've seen it ice up at all since adjusting the settings. Made a big difference to the COP, and I think a fair bit of the improvement seen was just down to the lack of icing and the need to run defrost cycles every ten to fifteen minutes in damp weather.

Very interesting, as someone curious as to how a hp would work in our house in the mainly damp but generally warm SW.

What sort of temperatures would you call cool when it's a "cool, damp day"?

Also, any further efficiency gains to be had by removing the UFH TMV altogether and running the HP at a target temperature of 28C? I know what you mean regarding those valves, the temperature reading on the knob doesn't work absolutely. Output temp varys significantly with respect to what flow temperatures it's seeing.

[Oldgreybeard]

Very interesting, as someone curious as to how a hp would work in our house in the mainly damp but generally warm SW.

What sort of temperatures would you call cool when it's a "cool, damp day"?

Also, any further efficiency gains to be had by removing the UFH TMV altogether and running the HP at a target temperature of 28C? I know what you mean regarding those valves, the temperature reading on the knob doesn't work absolutely. Output temp varys significantly with respect to what flow temperatures it's seeing.

I can't be sure, but my gut feeling is that the worst conditions were when it's been damp and the temperature drops down below about 4°C but stays above zero. That seems to be when a lot of water starts to condense out of the air as it cools.

I did think about removing the thermostatic valve, but I think it's useful to keep it there to regulate the temperature a bit better. The heat pump flow temperature varies a bit, as it cycles on and off, as it tends to overshoot by maybe 5°C every time it fires up and takes a while to settle. Most of the time when it's on the heat pump won't run continuously, as it can only modulate down to around 2kW or so output, which is a lot more than the heating needs. If we had a smaller heat pump, one that could modulate down to around 400W or so output then I think it might work OK to just remove the thermostatic valve altogether.

The temperature settings on the valve are as you say, pretty imprecise. I've added a couple of digital thermometers, with remote sensors poked under the insulation on the UFH manifold flow and return pipes, and used them to set the temperatures. The UFH normally runs with about a 4°C to 5°C temperature difference between the flow and return, which seems to keep things reasonably well under control. If the flow temperature gets too high the system tends to overshoot quite badly, it's easy to put too much heat into the floor slab and we may well then end up with the house being a bit too toasty for a day or two (as happened yesterday - the house is still over 22°C this morning, with no heating since the night before last). Heating a low energy house is definitely trickier than heating one that need more heat, it's very easy to over do it.

[Oldgreybeard]

Heating on here today for the second time this winter. The temperature dropped down to about 5°C here last night, but the house didn't get cool enough for the heating to come on overnight so I'm trying an experiment. We're generating around 4kW at the moment, and the forecast looks set to ensure the battery and hot water are fully charged, with some left over, so I've turned the heat pump on manually and added some additional monitoring to measure the power it's using and the power it's delivering (temperature differential of the flow and return and data from the flow rate sensor to measure the latter).

First observation is that after the initial spike at near full input power (about 1.8kW) that lasts around three minutes, the heat pump quickly modulates down as low as it will go, which is 30% of full output power (according to the manual). At 30% output power it's currently drawing 530W and delivering 1,930W, so it's running at a COP of about 3.65. Outside air temperature at the moment is 8°C, relative humidity is 43%.

One interesting thing is that the power it's using isn't changing at all, it's been sat at a rock solid 530W for the past hour or so. The other interesting thing is that I've been looking back through old data and there is, perhaps unsurprisingly, a strong correlation between daytime periods when it's fairly cold and the amount of PV generation. It looks like cold weather is often accompanied by clear skies, so generation is way higher than average for the time of year on those days.

I'm going to experiment with running the heating during the day on clear days, to see whether I can reduce our off-peak electricity consumption a bit this winter. The risk is that the house might get a bit too warm by doing this, as we'll also be getting some solar gain on days like today. Because the house takes a long time to respond to changes in heat input I'm not sure what's going to happen.

[Joeboy]

Heating on here today for the second time this winter. The temperature dropped down to about 5°C here last night, but the house didn't get cool enough for the heating to come on overnight so I'm trying an experiment. We're generating around 4kW at the moment, and the forecast looks set to ensure the battery and hot water are fully charged, with some left over, so I've turned the heat pump on manually and added some additional monitoring to measure the power it's using and the power it's delivering (temperature differential of the flow and return and data from the flow rate sensor to measure the latter).

First observation is that after the initial spike at near full input power (about 1.8kW) that lasts around three minutes, the heat pump quickly modulates down as low as it will go, which is 30% of full output power (according to the manual). At 30% output power it's currently drawing 530W and delivering 1,930W, so it's running at a COP of about 3.65. Outside air temperature at the moment is 8°C, relative humidity is 43%.

One interesting thing is that the power it's using isn't changing at all, it's been sat at a rock solid 530W for the past hour or so. The other interesting thing is that I've been looking back through old data and there is, perhaps unsurprisingly, a strong correlation between daytime periods when it's fairly cold and the amount of PV generation. It looks like cold weather is often accompanied by clear skies, so generation is way higher than average for the time of year on those days.

I'm going to experiment with running the heating during the day on clear days, to see whether I can reduce our off-peak electricity consumption a bit this winter. The risk is that the house might get a bit too warm by doing this, as we'll also be getting some solar gain on days like today. Because the house takes a long time to respond to changes in heat input I'm not sure what's going to happen.

That's a very good position to be in! Maybe just have to live with the response lag and keep the shorts handy?

[Oldgreybeard]

That's a very good position to be in! Maybe just have to live with the response lag and keep the shorts handy?

I think we may have to do just that, and put up with the house being a bit toasty. This is the first time since I built the house that I've tried running the heat pump during the day, so I'm not sure how it's going to respond. All the modelling I did during the design stage (about 10 years ago now) assumed only running the heat pump at night.

The model turned out to be pessimistic in terms of the amount of heating we really need, anyway, as it used the PassivHaus Planning Package (PHPP) to calculate heat loss and gain, and that seems to not be very accurate in practice (over estimated our seasonal heating requirement by around 30%, for example).