@dur Welcome!

You have the foundation of a good install I believe! You must be careful to keep it that way. 
Your heat pump will run best when you have the highest possible water volume in the system. That means all zones open as much as possible, and running your home as 1 zone if you can. Avoid thermostats in every room, and avoid too many valves and pumps.

A flow temp of 28C is on the lower end for a heat pump. Cycles of 25min are definitely acceptable at that flow temp. 90% of systems would cycle much more at that flow temp.
The reason a heat pump cycles: you ask a flow temp of 28C. The heat pump turns on. The temperature delta between the emitters and the room is low so heat emitted is relatively low. Flow temp increases because the heat pump is producing more heat than the emitters are emitting. When flow temp overshoots by a certain amount, the heat pump turns off. Flow temp decreases until it hits 28C (or lower) and the heat pump turns on again. Repeat.

You will have less cycling but lower efficiency at higher flow temps. That's the balance.

Is your heat pump internet connected? Do you have Melcloud?

Just to add I would review you UFH design you do not want any zones in your UFH system, you will just waste money on thermostats and a wiring centre that you don't need.

I have all of these and bypassed them all, took all the zoning out by removing the actuators, decoupled the wireless thermostats from the wiring centre and now control from 1 thermostat in the hallway.

Now the UFH can run for hours at 32C without any cycling, what it can't do as HCAS explained is run any cooler than that without cycling as you just can't get rid of the heat fast enough.

@hcas thank you very much for replying.

I guessed it was essentially that the Ecodan was thinking that the flow and return temperatures were too close. I watched the thermistor readings during a cycle (refreshing every few minutes) and it seemed to have been stable at c. 29 C out and 26 C return before it then shut off.

I set the 28 C flow temp on the basis that lower is better for efficiency. It is enough to get the floor to around 20 or so in 2 or 3 hours which is fine when the outside temp is only going down to 8 or 10 C overnight at the moment.

If the Ecodan controller is to be believed it is giving a COP around 4

I guess eventually we'll put it on to the WC curve and see what it then does.

I got the message about not trying to over control. I thought I might fit the zone valves only to shut them as one when the pump is off as there seems to be some reverse direction thermalling going otherwise.

I wonder if I can open all the zone flow valves a fraction to increase flow but maybe I'll do that when we have all the floor sensors wired into Home Assistant and I can see what is happening.

We do have radiators upstairs but don't really want heat in the bedroom so those are shut off.

Obviously at 28 the tiny towel rail and the other radiator are only just warm to the touch.

There is a WiFi box plugged into the Ecodan controller but nothing set up as we have not had the proper final commissioning yet as we have taken so long to get to this point in the project. Techie son has figured that he can plug direct into the controller so we can get the data out to Home Assistant so that will happen at some point...

So I increased the flow temperature to 32 and it ran for over 90 minutes before going off. It  then cycled but I'll have to watch it to see how frequently.

Anyway. I'll run at this for a while and see how it goes.

One question please  - does heat transfer from pipe to floor quicker as flow rate increases or vice versa  (or something else)?

I am wondering if I ought to try and increase or decrease flow through the UFH manifold

@dur If you have sufficient volume in your UFH loops, which it sounds like you have, and if you don't zone it, it's OK to close the radiators in the bedroom. I would start by leaving them open and only close them when you're actually too warm. 

On your question around heat transfer, heat transfer increases with higher flow rate and higher delta t between emitter and room temperature. 
If you have a buffer tank, be careful with distortion. 

My company has built software to control the heat pump optimally and automatically, 24/7. We connect to the heat pump via the internet connection. Let me know if you want to give it a try.

@hcas

Thanks.

So if I can increase the flow rate by the correct proportion across all the circuits that should help the transfer and therefore reduce cycling?

I am not sure what flow rate the Grundfos UBM can manage. Our total across all circuits is 19 LPM at the moment. I can call Grundfos and ask as I can't find the detail in their manuals.

We don't have a buffer tank btw.

The software sounds potentially interesting but I think at the moment I'd like to get the basics right and we are not even properly we wired up yet.

@gary 

If you are able to run continuously at 32 C (or whatever) how do you control the overall house temperature?

Presumably at 32 the floor eventually warms up to a certain point where input heat and heat loss are in balance.

At the moment, having followed your suggestion and tried 32 in preference to the 28 previously, we can get the floor up to around 21 in about 3 hours. In the winter it might be fine to run constantly but not at the moment so see some level of control will be needed.

Thanks!

@dur My thermostat turns it off at 21.5C comes back on at 21C like you said means it runs for several hours then if off for several hours and repeats any lower flow temps means I get cycling every 20 mins.

Hello. Update and questions...

So we have been running as previously described ie manually turning on the ashp and the UFH pump. The Ecodan is running on fixed flow temperature set to 32 C. No great finesse - on if we are in and remember and then leaving on until we see the slab temp (at one particular reference point) increase to 21 C in the mild weather and now it's cold to about 22.5 C.

At the moment the slab, at the ref point, is dropping to around 19 C in 24 hours and is taking maybe 4.5 hours to get back to temperature.

The delivered energy/consumed electrical energy according to the Ecodan controller is over 4.0.

So, given that the cycling issue seems to demand a minimum flow temp of 32 C and given that we can cope with the colder weather by leaving it on longer at 32 C, is this going to be the most efficient way to run the system?

If I put it onto weather compensation it seems that the minimum flow temp will still need to be 32 so I don't know what I would set the high and low flows to. Also are we realistically going to better the 4.2 to 4.4 we are seeing on the controller?

When we get around to controlling with Home Assistant we could maybe alter the run time according to outside temp rather than flow temp - weather compensation but according to time.

Very interested to hear any thoughts and apologies for being long winded!

Posted by: @dur

↑

Hello. Update and questions...

So we have been running as previously described ie manually turning on the ashp and the UFH pump. The Ecodan is running on fixed flow temperature set to 32 C. No great finesse - on if we are in and remember and then leaving on until we see the slab temp (at one particular reference point) increase to 21 C in the mild weather and now it's cold to about 22.5 C.

At the moment the slab, at the ref point, is dropping to around 19 C in 24 hours and is taking maybe 4.5 hours to get back to temperature.

The delivered energy/consumed electrical energy according to the Ecodan controller is over 4.0.

So, given that the cycling issue seems to demand a minimum flow temp of 32 C and given that we can cope with the colder weather by leaving it on longer at 32 C, is this going to be the most efficient way to run the system?

If I put it onto weather compensation it seems that the minimum flow temp will still need to be 32 so I don't know what I would set the high and low flows to. Also are we realistically going to better the 4.2 to 4.4 we are seeing on the controller?

When we get around to controlling with Home Assistant we could maybe alter the run time according to outside temp rather than flow temp - weather compensation but according to time.

Very interested to hear any thoughts and apologies for being long winded!

 

The efficiency gain through weather compensation gets less at low flow temperatures, and is probably around 2% per C or less at 32 and you are already seeing a good COP.  That said the majority of your heating consumption will be when its warmer outside than it is at present, so if you can run at an even lower flow temperature (ie employ weather compensation as its intended) then it should give an improvement.  There again you may be in the zone where fixed loads (pumps, electronics) are a significant proportion of your consumption and, as we have seen from others, this can sometimes argue in favour of part time working, particularly if the heat pump is oversized and/or you are out for long periods during the day.

If you can say how much you are currently consuming/outputting to the property, and what your pattern of occupancy is, that will give more of a clue to which direction is more likely to be most cost effective.  Ultimately you may have to perform the experiment - which is very difficult in practice because there are far too many uncontrolled variables.  

Thanks James

The truth is that I am probably being a bit premature in my thinking as we still have a couple of small but critical areas of the house to seal up / insulate so I can't really give usage figures that would be representative longer term.

Also a couple more days of cold weather make me think we might need to increase the flow temperature just a little in such conditions. I guess a pretty flat compensation curve could end up as the way to go.

Hopefully  a bit more data and experience will give some clues and maybe some trial and error too!