Room stat for heat pump

[Sultry1]

We have a Daikin Altherma HT heat pump system which drives one zone of underfloor heating and 2 zones of rads (upstairs and downstairs). This Daikin system has inverter driven pumps and compressors so can adapt somewhat to load and has weather compensation fitted and working.

The u/f works really well with the Continental room stat.

The rads are controlled via Honeywell CM901 room stats. I've read and mostly understood the other posts on this site that describe the TPI control system that the Honeywell stats use.

However, I'm wondering if these Honeywell stats are appropriate for heat pumps. The issue is that when the stat is in the PI zone, the minimum cycle frequency and minimum cycle times just don't seem to be appropriate. Our heat pump has a 10 min start up period and then takes at least 30 mins to get the flow temp up to anything sensible (say 50C) even at this time if year. In colder weather it can easily take more than an hour. We find that the stats come on and off regularly night and day without the rads heating up at all. i appreciate that the CM901 is supposed to be a learning system, but that's not how it appears to behave in practice.

And before anyone comments that heat pumps should be run continuously, could I say that I understand that but in early winter/ early spring, the heat demand is not sufficient (the heat pump can't modulate low enough) and the electricity consumption goes through the roof (not literally).

Should we be using a room stat where we can turn off the TPI, such as the Danfoss TP7000 series so that the heat pump will come on less and run longer even if that means somewhat wider fluctuations in temperature?

Any thoughts?

 

[ALEC1]

why would you want temperature fluctuation? You say the thermostats shut down, but why? is it because of another heat source?

the issue isn't really the TPI element of the thermostats, but the mixing of on-off controls with a variable output heat generator..

 

[Sultry1]

Thanks.

No we don't want temp fluctuations but if the cost of tighter control is that the heat pump is cycling frequently (with risks to long term reliability and noise from the pipe work each time it begins to very slowly heat up/cool down), then we would accept it.

The reason the stat turns off is that the zone is up to temp. Then, according to the action of the PI control, the stat switches repeated on/off to maintain the temp around the set point.

My thought was that this kind of control is more suited to gas boilers where you can get a meaningful amount of heat in just a few minutes operation. This is isn't the case with a heat pump.

 

[ALEC1]

you can vary a lot of the settings on the honeywell, I would try that first...

you will have to go to installer level though...

do you have TRVs too?

 

[dcawkwell]

We have a Daikin Altherma HT heat pump system which drives one zone of underfloor heating and 2 zones of rads (upstairs and downstairs). This Daikin system has inverter driven pumps and compressors so can adapt somewhat to load and has weather compensation fitted and working.

The u/f works really well with the Continental room stat.

The rads are controlled via Honeywell CM901 room stats. I've read and mostly understood the other posts on this site that describe the TPI control system that the Honeywell stats use.

However, I'm wondering if these Honeywell stats are appropriate for heat pumps. The issue is that when the stat is in the PI zone, the minimum cycle frequency and minimum cycle times just don't seem to be appropriate. Our heat pump has a 10 min start up period and then takes at least 30 mins to get the flow temp up to anything sensible (say 50C) even at this time if year. In colder weather it can easily take more than an hour. We find that the stats come on and off regularly night and day without the rads heating up at all. i appreciate that the CM901 is supposed to be a learning system, but that's not how it appears to behave in practice.

And before anyone comments that heat pumps should be run continuously, could I say that I understand that but in early winter/ early spring, the heat demand is not sufficient (the heat pump can't modulate low enough) and the electricity consumption goes through the roof (not literally).

Should we be using a room stat where we can turn off the TPI, such as the Danfoss TP7000 series so that the heat pump will come on less and run longer even if that means somewhat wider fluctuations in temperature?

Any thoughts?

I have a trianco heat pump. Not a fancy inverter unit 12kw. So it fires up and you get instant heating. Not sure why your pump would be taking
10 minutes to get going. I have only radiators in the house and about 100litre or so of water volume.
I can't understand why the thermostats would be coming on and off without the radiators getting hot. It sounds as if the heat isn't getting
away from the heat pump or badly balanced system.
With the heat pump on and thermostats calling all the radiators should get hot. Unless the heat pump is not sufficient to handle the heat loss in which case the heat pump would be running all the time. You have weather compensation so the radiator will only heat to a temperature to supply the necessary heating effect.

 

[Sultry1]

Thanks again.

Yes, I have TRVs on 16 of the 20 rads with other 4 as bypasses. I have tried running the heat pump continuously (by setting a high target temp) and controlling as possible with the TRVs. This produces very high electricity consumption for the time of year (>50kWh per day).

I have also adjusted the settings to 3 cycles per hour, 5mins minimum run time but left the PI zone at 1.5C. While the heat pump does not come on every 20min cycle, it does come on for most. It runs for the 5mins and then drops out again. In that time it does not get out of start up mode, let alone produce significant heat.

To the other poster, the Daikin system does have a kind of soft start followed by a ramp up in compressor frequency which lasts around 10 mins. I then have 20 rads, most of which are cast iron (but fully cleaned). Pipework is iron also mainly and the house is quite big leading to some quite long runs. I have insulated all pipes where I can get to them though. So in all, there is quite a lot of thermal inertia. Do you use a room stat with TPI control and if so, how do you get on with it?

 

[ALEC1]

if you have a system with high thermal inertia then you might just as well run it as UFH....

It sound like there is a complimentary controller for the heat pump that has been omitted in favour of on-off controls...

 

[dcawkwell]

Thanks again.

Yes, I have TRVs on 16 of the 20 rads with other 4 as bypasses. I have tried running the heat pump continuously (by setting a high target temp) and controlling as possible with the TRVs. This produces very high electricity consumption for the time of year (>50kWh per day).

I have also adjusted the settings to 3 cycles per hour, 5mins minimum run time but left the PI zone at 1.5C. While the heat pump does not come on every 20min cycle, it does come on for most. It runs for the 5mins and then drops out again. In that time it does not get out of start up mode, let alone produce significant heat.

To the other poster, the Daikin system does have a kind of soft start followed by a ramp up in compressor frequency which lasts around 10 mins. I then have 20 rads, most of which are cast iron (but fully cleaned). Pipework is iron also mainly and the house is quite big leading to some quite long runs. I have insulated all pipes where I can get to them though. So in all, there is quite a lot of thermal inertia. Do you use a room stat with TPI control and if so, how do you get on with it?

Something is not right turn all your thermostats right up to 30degrees all the radiators should get hot and should do so within about 30 minutes. They should then stay there with the heat pump running all the time. If not there is a problem. You have set 3 cycles per hour which is fine.
So this basically means when the heat pump has been running it won't call again for 20 minutes.
The problem is why is the heat pump only running for 5 minutes when it has been called it can't of provided any heat so why has the wall thermostats suddenly decided the room is now warm?

 

[Sultry1]

Exactly the problem I think. It calls for heating every 20mins because no significant heat has been produced in the meantime and hence the room temp has not shifted.

The question is, will the 'learning function' on these stats eventually cope with this? By that I mean, if the response of my system means that the heat pump has to be on for say 30mins before it shifts the room temp, will the learning function eventually pick this up and always extend the run time accordingly.

Otherwise, it appears that we may be better with a non-TPI stat and run it as we do the underfloor (per Alec).

 

[ALEC1]

I am suspecting that the TPI is having more issues with the cast iron rads, rather than the heat pump, because of the high thermal inertia..

the high thermal inertia is good for energy efficiency, but not for on off heating which is what you are expecting..even with TPI

Much better to run the whole system continuously, using temperature profiling to vary the flow temp rather than turning things on and off, but you can't do that with 240v switching.

I would try Daikins own controller...

 

[dcawkwell]

Exactly the problem I think. It calls for heating every 20mins because no significant heat has been produced in the meantime and hence the room temp has not shifted.

The question is, will the 'learning function' on these stats eventually cope with this? By that I mean, if the response of my system means that the heat pump has to be on for say 30mins before it shifts the room temp, will the learning function eventually pick this up and always extend the run time accordingly.

Otherwise, it appears that we may be better with a non-TPI stat and run it as we do the underfloor (per Alec).

Not quite sure about the learning function. Simply if the rooms are cold the controller should call for heat ALL the time until this is satisfied.
Any other settings shouldn't matter. Learning functions are unusually just so the controller can judge if you need to to be at 20 degrees by 7am it needs to start at 6.30 say to meet that.
Check that you have nothing funny set up on the timings on the controller
and read the manual on correct set up.
I know you have underfloor which will be set much lower in operation.
I assume there is a separate bending valve and pump for this circuit.
If not there should be. Underfloor could somehow be cutting out the heat pump for the rest of the house zones.

 

[dcawkwell]

I am suspecting that the TPI is having more issues with the cast iron rads, rather than the heat pump, because of the high thermal inertia..

the high thermal inertia is good for energy efficiency, but not for on off heating which is what you are expecting..even with TPI

Much better to run the whole system continuously, using temperature profiling to vary the flow temp rather than turning things on and off, but you can't do that with 240v switching.

I would try Daikins own controller...

Daikins own controllers on air con units are hopeless. We have them in the office and they can't maintain a constant temperature. Hopeless.

 

[ALEC1]

and what do Daikin say about that?

 

[dcawkwell]

and what do Daikin say about that?

They Daikin engineer who comes to it says it is right.
So why is it 22 degrees in the office and it is set a 19? It seems to have a 5 degree range as default which is far two large. I am not responsible for the installation.

There seems to be no way of controlling it other than getting up
all the time and switching it from heating to cooling.
Set it to self adjusting and we either freeze or are boiled.

 

[ALEC1]

No manufacturer releases products that don't work, but very few manufactures, or there sales agents in the UK support their engineers properly.


I am not familiar with Daikins, or many heat pumps for that matter, but modern control technology I do get!

something like a heat pump is designed to be varying its flow temp with input from inside and outside, its also designed to be run "constantly" between two internal temperatures..

they get this technology right in other countries! why not the UK?