Posted by: @derek-m

↑

@mjr

I cannot state for definite, but as a Control Systems Engineer I would expect both the LWT and the RWT will be used in the control of the speed of the internal water pump of the ASHP. The DeltaT between these two measurements, as far as I can see, is the only way that the controller can be aware of what is happening to the heat demand. When the heat pump is operational, it is afterall a continual balancing act, between changes in the LWT and changes to the RWT.

I think that the Ecodan controller periodically runs the water pump for several minutes, when the compressor is stopped, to equalise the temperatures around the system, to try to ensure that it is getting reasonably accurate information.

Having thought about what may be happening when a heat pump is operating at the lower end of its capabilities, I agree that using an on - off type of control under these circumstances may be more efficient than the normal controller running WC. I have that type of control for our gas boiler, but rather than a thermostat I use an industrial PID Controller, the control parameters and response of which I have adjusted to provide indoor temperature control accurate to +/- 0.2C.

I would be interested to know the operating conditions under which you tried auto adaptation. One reason for large changes in LWT could be stratification when the compressor and water pump shutdown. Changes in temperature whilst the system is operating, could be due to an inadequately mixed water flow past the temperature sensor, which alternately heats and cools the sensor as the flow rate changes. I have experienced this on several occasions in the past on industrial systems, where the temperature sensor has been located too close to the heat source.

If I run with on-off stats in relatively mild conditions using weather compensated flow then all that happens is that the stats switch the ASHP on and off but it still cycles when it's on. The ASHP achieves its target flow quite quickly and even though the stat is still calling for heat, the ASHP stops heating. The energy use is very variable with a big peak whenever the stat calls for heat. Without stats, there is cycling but it's very constant and without large peaks.  I don't think there is much in it in terms of energy consumption but I think the weather compensation approach provides a more pleasant, constant room temperature. 

If I use a constant flow and stats there is less cycling, bigger peaks, more variable room temperature and it definitely uses more energy to maintain the same room temp as weather compensation alone. 

I plan to do some analysis on my newly acquired wireless controller when I get around to installing it.

@kev-m

Hi Kev,

Thanks for the useful information, which only goes to prove that theory doesn't always work in practice.

Posted by: @derek-m

↑

but as a Control Systems Engineer I would expect both the LWT and the RWT will be used in the control of the speed of the internal water pump of the ASHP. The DeltaT between these two measurements, as far as I can see, is the only way that the controller can be aware of what is happening to the heat demand.

I think we are in vigorous agreement there. I did say RWT is used in delta T. I just haven't noticed it being used otherwise. 

Posted by: @derek-m

↑

I have that type of control for our gas boiler, but rather than a thermostat I use an industrial PID Controller, the control parameters and response of which I have adjusted to provide indoor temperature control accurate to +/- 0.2C.

I may connect a software PID controller to the heat pump this week, to see how it performs. It may not get to 0.2C when the ecodan room sensor is 0.5C resolution and my 0.1C resolution sensor is too near a radiator, but we'll see! 

Posted by: @derek-m

↑

I would be interested to know the operating conditions under which you tried auto adaptation.

Last November after the firmware upgrade, this February and accidentally last Sunday. Room target 19° the first two, starting from about that, 17° the last time, starting from 14.5°.

Sadly, I've lost the more detailed non-melcloud data recordings from the second try and never had them for the first. It was that first attempt which started me recording more data, to try to spot what was going wrong. There may have been more attempts that I forget. I'm not as rigorous about this as at work and probably more conservative at aborting experiments before anyone grumbles about being cold. 

@mjr

I am not certain how knowledgeable your are concerning PID Controllers, so please forgive me if I am 'teaching my Granny to suck eggs'.

To be able to accurately control the indoor room temperature requires overcoming quite a number of problems. Having the best controller in the World may not achieve the desired result, unless it is supplied with the correct information. It is therefore necessary to have a reasonably accurate, but more importantly repeatable, temperature sensor, which is installed in the correct location. The controller also needs to be able to manipulate the heat source with adequate accuracy so as to be able to change the room temperature in a slow and even manner.

One of the major problems experienced by home heating systems is the relatively long time delays that can occur between action and reaction. If I increase the temperature setpoint on my controller by 1C, it can be several hours before it actually achieves the desired temperature. If I lower the temperature setpoint by 1C, it can be even longer before the temperature reaches that desired, particularly if the outside air temperature is quite mild. For this particular reason a single PID control system would probably struggle to provide accurate and consistent control of the room temperature.

To provide more accurate control it is necessary to utilise two PID controllers, in a Master - Slave configuration. The Master Controller is supplied with data from the room temperature sensor, which it compares against the setpoint, if the room temperature is above the setpoint the controller will reduce its output signal, if the room temperature is below the setpoint it will increase its output signal, and if the room temperature is equal to the setpoint the output signal will remain the same. The output signal from the Master Controller now acts as the setpoint for the Slave Controller, with the second input being from a temperature sensor measuring the heat source. The output signal from the Slave Controller being used to vary the heat source.

In the above situation if the room temperature is below the setpoint, the Master Controller instructs the Slave Controller to increase the heat source by slightly increasing the setpoint of the Slave Controller. Being a dutiful slave, the Slave Controller starts to increase the heat source, but soon after starting to do so it soon finds that the heat source temperature matches the setpoint instructed by the Master Controller, and therefore does not increase the heat source any further. If the room temperature fails to increase to the specified level then the process will continue until the system is in balance.

To prevent overshoot or undershoot of the room temperature, the response timing of the Master Controller will need to be quite long and hence give slow response, whilst the response of the Slave Controller can be much faster, since the temperature feedback is much quicker.

In an ASHP system I suspect the following may be happening, but without access to the internal software it is difficult to be certain.

In WC only mode, the controller will probably act in a similar manner to the Slave Controller, in that it will receive a 'setpoint' signal from the combination of ambient air temperature sensor and WC curve, with the 'feedback' signal being the LWT, and the 'output' signal speeding up or slowing down the compressor. The actual room temperature will be dependent upon how accurately the WC curve matches the heat loss characteristics of the home.

In WC plus auto adaptation mode, the system would probably operate as described above, but with an added input of the 'Master Controller' type, that adds to or subtracts from the setpoint of the 'Slave Controller', and hence finely adjusts the LWT until the desired room temperature is achieved.

If my above assumptions are correct, then for auto adaptation to function correctly, the controller must have the ability to self tune the response times to the correct values for your particular home, which of course could take some time to achieve.

A further problem could occur during milder weather if the heat pump is not running continuously, unless the manufacturer has taken this into consideration.

@derek-m I like to keeep things simple and use a Pulse width modulation thermostat with the heat pumps WC on under floor and the heat pumps  load/WC on radiators. Unfortunatly there are very few heat pumps that have a combination of WC and LC (adaptive control) with the units. Most manufacturers are lasy.

Posted by: @heacol

↑

@derek-m I like to keeep things simple and use a Pulse width modulation thermostat with the heat pumps WC on under floor and the heat pumps  load/WC on radiators. Unfortunatly there are very few heat pumps that have a combination of WC and LC (adaptive control) with the units. Most manufacturers are lasy.

I assume that what you mention is actually a temperature sensor and transmitter combination rather than a thermostat, since Pulse Width Modulation is a technique used to vary the signal. I have not heard of this type of device, who is the manufacturer?

@derek-m It looks like a standard thermostat and connects to a standard under floor actuator head. Instead of either open or close, as with a standard thermostat, it modulates and varies the ammount of water flowing in to the under floor pipes depending on demand. We generally only ise the to reduce bedroom temperatures, not for common areas which will be controlled by the heat pump.

Salus make them.

Posted by: @heacol

↑

@derek-m It looks like a standard thermostat and connects to a standard under floor actuator head. Instead of either open or close, as with a standard thermostat, it modulates and varies the ammount of water flowing in to the under floor pipes depending on demand. We generally only ise the to reduce bedroom temperatures, not for common areas which will be controlled by the heat pump.

Salus make them.

It would appear to actually be a controller, since it is modulating the water flow to regulate the room temperature. I realise that it is a pay on words, but to me a thermostat is an on - off device. As an Engineer I always try to use the correct description so as to avoid any confusion.

@derek-m A thermostat is a device that controlls atemperature, it cancachieve this in many different forms. On-Off is the most basic form and conciquently the most inefficient and unstable form of temperature control, whist modulation is the most efficient and stable form, we use modulation.

Hi folks – I’ve been noodling on the issue of controllers and I think this thread might be on-topic. We have (brief recap) a 16kw Daikin Altherma HT ASHP, 300sqm single level open-plan house, UFH in the night zone (bedrooms), (sized) radiators in the day zone (everywhere else), 6.6kW solar PV.

I’ve followed the excellent advice on the forum (particularly @derek-m) and we’re happy with how its running but think it could be running more efficiently of course. We have the system running in WC mode (0.6 currently) programmed to provide space heating from 6am to 10pm (we both WFH); DHW cycle at 1pm, with a 30m boost at 10pm. We don’t use setback at night as the house doesn’t cool down (max measured drop of 1C). We have a 12kW log burner in the main (15m x 6m) space which we run most days in winter.

Each zone has a third-party wall thermostat installed, both of which are set +2 degrees above desired room temp (same for both zones) so essentially permanently calling for heat (which of course is not available from 10pm-6am). The Daikin ‘remote controller’ (EKRUAHTB) sits in our ‘tech room’ with the indoor unit and the thermal store. As such, the ASHP has no knowledge of the room/ zone temperature(s) – just the outside temp, LWT and RWT temps, and the hot water tank temp. I’ve attached the manual for reference that details the functionality of the ‘remote controller’ – manual states (p10) that the controller knows the indoor temp but this is actually the LWT in operation.

I’m aware that Daikin sell a wired in-room controller compatible with our system – the EKRTWA – which gives a lot more control over programming (e.g., a holiday mode) as well as providing feedback to the remote controller on in-room temperatures, and I’ve been considering replacing the current ‘on-off’ third party thermostats with one or two of these Daikin units. They would seem to help improve efficiency of the heat pump. Are they a good investment? Are they an upgrade on our current ‘on-off’ thermostats, given how we use the ASHP (on at 6am, off 10pm)?

One additional question I would have is since we have two ‘zones’ (UFH and rads) with the two on-off thermostats currently, could/ should I replace these with two EKRTWA units that would allow for separate management of the two zones? I’ve only seen examples in the Daikin literature of one unit being used.

Thanks in advance for any help.

@marzipan71 I had a Heatmiser stat (on or off only) which I set higher than desired temperature then used the LWT and the WC to control the temperature 24/7. This scheme seemed fairly reliable on my Daikin EDLA08 but has now been replaced by an Evergreen Energy Homely smart controller which is settling in at present but looks to be controlling temperature very well and more economically than the Heatmiser. In the manual, it mentions that second zone control is not yet available but on the way; I note that the Homely Hub does have facility for the second zone built in though. Regards, Toodles.

Thanks @toodles for the response. Yes, that's what I'm talking about - almost Thinking v1.0 is using the ASHP with the 3rd party stats, on-off, LWT and WC to control. That's where we are thanks to @derek-m and his advice - the kwh's we buy in from the grid are 50% of what they were in 2021 when we were trying to use the ASHP like a gas boiler (Thinking v0.0). I'm thinking v2.0 could be using the Daikin proprietary wired in-room thermostat which is designed for use with our model to provide a greater level of control/ efficiency, still using LWT and WC. I'm in Italy and Homely isn't available here. I don't think the Daikin in-room controller is as smart as a Homely but it will provide that extra piece of data back to the 'brains' (the 'remote controller') which at the moment is lacking. If anybody has used the EKRTWA I'd be interested to know if you can have two connected which can control water supply to the UFH and rads based on 'zone' temperature.