I’m trying to understand why my efficiency on Weather Compensation is worse than Auto Adapt when I try it.
Here is an example.
Last night I put our 8.5kW unit into Weather Compensation mode and then this morning switched back to AA. Can be seen here Emoncms - app view.
You can see that the WC part cycles much more but this is only around 3 times per hour so doesn’t seem excessive and looks similar to people with other units that do WC well.
At first glance WC looks better as it is maintaining a much more steady temperature and operating at a lower flow temperature.
However with AA there are longer runs but at much higher flow temperature which should be less efficient.
However when I zoom in on this morning 0h-8h I get a COP of 3.16
Do you know how big your emitter circuit is? The heatpump app reckons it’s 12 kW at dT 50. At a flow temperature of 35° the radiators can only emit 2kW, whereas the heatpump has a minimum output of 3.7 kW.
At 40° flow, 3 kW output from the emitters are more able to keep up, though is still less than it needs so some cycling is inevitable.
I reckon the target temperature for WC is simply too low for your emitter circuit, and the heat pump will not run smoothly below 43°C. Setting a higher curve may give you better performance.
Ignore the separated out Space Heating COP values when in Weather Compensation as the heating flag isn’t on the whole time it is space heating in this mode.
@F1p This flag is based on the Operations Mode read from CN105 using your module. I think when in WC mode this flag shows ‘Heating’ when the flow temperature is below the setpoint but then goes off when it exceeds this. Not a problem, just interesting although if I use WC more I will need to find a different way to set the flag for emoncms.
The orange shading on the graph in WC mode is an interesting way to see how well the emitters are able to get rid of the heat.
That makes sense but should that not just mean there is more cycling at lower flow temperatures but surely as the flow temperature is lower the heat pump should still have a higher COP.
Or is there always a hit associated with the start of each cycle? If so why is this?
I guess upping the flow temperature so emitter output is closer to min output of heat pump but in WC this would just overheat the house.
I’ll have to have a look, I think it may have been following what the operation mode being read is and show the pause / idle icon rather than the play one for heating.
I think the Auto Adapt mode has access to some efficiency-improving tactics not available in Weather Compensation mode, including running the circulating pump while not harvesting heat but other things I forget, too.
I think auto adapt is simply better becuase the target temperature is on average lower. The set point can in theory move to the lowest needed to achieve the required room temp. The curve is actually quite a blunt tool. The settings need to be hight enough to cope with warm up, and no ability to ‘throttle off’ when incidental gains are contibuting etc
I think that is probably the case where the thermostat is somewhere prone to overheating.
The only time I think it is suppressing the flow temperature in our system is when its quite cold out but there is no wind to wash away heat inside.
The rest of the time the flow temperature is a good bit higher than what I had it in in WC mode as it allows for large over or undershoots. This does prevent cycling which must be where I see the increased efficiency.
I guess when it is cycling, AA will allow the flow temperature to drop quite a bit so I guess the first part of the next cycle will be at a low temperature till it gets back up again.
Soo… maybe that’s where the efficiency gain is. A bit like the difference in COP when you heat a DHW cylinder from cold vs heating an already warm one.
The reason AA can get away with doing that is because it is checking what it ‘can get away with’ before indoor temperature is negatively effected.
So maybe I should be thinking of the average flow temperature needed over a time to maintain an steady indoor temperature and here AA might be lower than WC trying to stick closely to an on average higher flow temperature.
I guess when not cycling there shouldn’t be much difference in efficiency (assuming other factors are changing the indoor temperature) so maybe this is why AA is so much better in systems like mine that don’t have adequate emitter output.
I’ve been running Heating Flow with a radiator system (Ecodan R290 6kW, no heat meter) setting the flow temp via Home Assistant (using the MQTT dongle) to thermal comfort.
I haven’t used MELCloud but setting the flow temp via HomeAssistant is as easy as setting my Thermostat and I generally set the thermostat high so I’m just controlling indoor temp via the flow temperature and the house doesn’t reach the thermostat temperature.
I found AutoAdapt would overheat the house and weather compensation was cycling, I haven’t uploaded data to Emoncms yet but using the ‘yesterday’ Ecodan figures, AutoAdapt performs best, then Weather/Heat Compensation, then heating flow.
Averaging 3.1 COP (~100 days, since late August) from Ecodan production/consumption figures.
Using Shelly clamp consumption data and Ecodan production - 4.5 COP over 66 days (since late September, suspect the Ecodan production figure is optimistic). Mostly running with weather compensation but switched to setting flow temperature last 2 weeks.
1970s unfilled cavity semi detached, 67sqm, 9 K2 radiators. There is a radiator at the back of my office desk and I have suspended uninsulated wooden floors so notice the temperature difference from cycling at floor level, especially on windy days.
The R290 6kW seems to be able to modulate down to ~300w, pump demand seem to be 40W @14l/min, I have cycling issues when I set flow temp down to 30C.
That is interesting, and nice it goes down to 300W. I never managed to get my head around Home Assistant… too complicated for me. So,… if AA overheats the house, what happens if you turn the setting down?
Could it also be that the main controller or wireless thermostat (whatever is being used to provide the room temp for AA) is in a colder location or somewhere it takes longer to ‘see’ the heat from nearest emitters?
Also the thermal diff settings can be adjusted to reduce the amount of flow over or under shoot allowed in AA.
One thing I did notice when running WC yesterday morning was that the temperature was more even. I could perceive the larger swings in flow temperature from AA even though the room temp measured my my main controller was still steady. So I’ve since tightened up the thermal diff temperatures.
Yes, the thermostat is towards the north side of the house (hall) and one of the cooler locations. I tried AA with lower thermostat set point this morning but once it cycles off, the impact on thermal comfort is noticeable, it highlights the fabric changes I need to do to make the house more comfortable/efficient.
I played around with the ‘thermal diff’ settings initially, I think I set 47C max (initially left at 50C) and 30 minimum. My main FTC controller is located on the cylinder (packaged cylinder) is in the ‘coal store’ so certain settings like weather compensation settings for me are a pain to change as I need to go outside and into the store. The thermostat has only up/down buttons on it, so as I use HomeAssistant with the dongle, heating flow feels like it offers the best control for comfort, although I appreciate I am using more pump power proportionally.
Its bad isn’t it? It is very irregular over long time frame, and some periods of operating fine, then quite extreme ‘crash’ of output. I am in the middle of looking at EEV pulses to acertain if the valve is getting the correct sort of messages, or if it seems like one of the sensors might be intermittently telling lies… e.g. TH4,Hz, TH3, P-sensor, TH6, TH7. They probably all feed into the LEV control. There are plenty of long runs that look perfect, so it has the potential of working. Was this the unit that had software update this year?
Have never spotted S/V valve before. what is it? Compressor will have an internal safety bypass to simply relieve the discharge to the suction. These should never operate, and surely they would not open intermittently.
Power input is around 1200W and occasionally dips to 700W. The heat output drops at the same time, but then falls much lower than it should. Is the something bad happening with the refrigerant when the compressor throttles back?
Also peculiar as to why it’s throttling back at all, as it looks close to stable running.
