Closing remarks on the hot water heating. Even though the heat pump only reports whole kWh of energy consumed, it should have reported the energy used by the circulating pump as it was over 1 kWh but it did not. It evidently ignores the circulating pump's consumption. For the whole of May the overall hot water COP based on the heat pump's own energy consumption and transfer measurements was 1.6.  For Friday, only 1 kWh out of 2.2 kWh energy consumed was accounted for, so the actual COP was about 0.7. 

Although it is not the heating season, I recalled that it had taken about 30 hours of continuous operation for my heat pump to get my home up to temperature when it was commissioned on 21st March, and although I appreciate fully that the heating is supposed to be slow and gentle, one doesn't expect it to take over a day to get warm after returning from a holiday, or in this case, heating my home about 96 hours after the gas boiler had been removed.

To test the heating, with the sensors added, I turned off the overnight hot water cycle and set the heat pump the space heating target of 21C during the overnight off-peak electricity period instead. The room temperature started at 19.8C, so only a 1.2C rise was needed, which the sun can exceed easily in a couple of hours on a sunny day. Needless to say, it failed, raising the room temperature by just 0.5C in the 2 hours.  Here are my measurements:

Rounding the decimals, since the heat pump only reports full degrees Celsius, the water entering the un-zoned underfloor heating is 7C cooler than the water leaving the heat pump. The water leaving the underfloor heating is only 0.7C cooler than when it entered, suggesting that the heat transfer efficiency through the unwanted buffer tank is just 0.7/7.7 = 9%. This hopeless situation appears to be reflected in the power drawn by the heat pump, which resembles a fine toothed comb, oscillating between about 250W and 550W. I'm guessing this is what cycling looks like.

My underfloor heating manifolds each came with a circulating pump, but they were taken off on the understanding that the circulation would be driven by the heat pump itself and the buffer tank integrated into the bottom of the pre-plumbed hot water cylinder would be configured as a volumiser on the return to the heat pump. Instead of having 2 circulating pumps for the 14 underfloor circuits, there is just the one mounted on the cylinder, and with the flow control valves throttled right down on the shortest loops to the towel rails, shower room, bath room, utility room and cloakroom, the maximum unrestricted flow through the longer loops is only 4 litres per minute. My underfloor pipework has an internal diameter of 11.5mm, giving a volume of 1 litre for every 9.6 metres of pipework laid. It's close enough to calculate my underfloor heating volume as 1 litre per 10 metres of pipe. My longest runs are over 80 metres, so with this flow rate a water change takes over 2 minutes and the typical ones are 60 to 70 metres, which means a water change takes over 1.5 minutes. The total underfloor pipe length is about 660 metres, thus has a capacity of about 66 litres. From this, it is self-evident that the heat pump is cycling more frequently than the water is changing in the underfloor heating. Surely this is utter nonsense?

My heat loss calculations are about 3 kW, and the heat pump is rated at 4 kW. I don't know what the minimum system volume is for a 4 kW heat pump, but all pipework is 28mm, with a volume of about 1 litre per 2 metres of pipe. That gives a total volume of about 75 litres for the underfloor system, about 15 litres for the heat pump flow and return pipework.  Does the buffer tank even need to be configured as a volumiser, or is it completely unnecessary?

Lastly, if you've read this far without losing the will to live, a video for your entertainment. Exchanging the hot water cylinder introduced hammering when the first hot water tap is opened after a water heating cycle. I recorded this at 102 dB, easily loud enough to startle and wake up anyone who is still sleeping. The installer came back and added a tiny expansion vessel close to the hot water outlet from the cylinder. This video shows what it sounds like now.  No special effects have been added! https://youtube.com/shorts/29AAeSHsIVw

Sorry to hear about your problems, which your installer should be fixing however...

Posted by: @sandvika

↑

and the buffer tank integrated into the bottom of the pre-plumbed hot water cylinder would be configured as a volumiser on the return to the heat pump.

has it been configured as a volumiser or is it still a buffer tank?  There appears to be a 5C difference between heat pump flow and heating flow, there should be minimal temperature difference! 

Are there any restrictions in or bypass to the UFH.  It looks, from your temperatures, possible that very little water is actually flowing through the UFH which can only be because of a restriction, bypass, or because the pump isnt turned up.

Posted by: @sandvika

↑

My heat loss calculations are about 3 kW, and the heat pump is rated at 4 kW. I don't know what the minimum system volume is for a 4 kW heat pump, but all pipework is 28mm, with a volume of about 1 litre per 2 metres of pipe. That gives a total volume of about 75 litres for the underfloor system, about 15 litres for the heat pump flow and return pipework.  Does the buffer tank even need to be configured as a volumiser, or is it completely unnecessary?

Probably not, but a volumiser (plumbed either in the return or (better) in the flow NOT between the two) will do no harm so, assuming its been correctly plumbed this is not the problem.

Posted by: @sandvika

↑

Lastly, if you've read this far without losing the will to live, a video for your entertainment. Exchanging the hot water cylinder introduced hammering when the first hot water tap is opened after a water heating cycle. I recorded this at 102 dB, easily loud enough to startle and wake up anyone who is still sleeping. The installer came back and added a tiny expansion vessel close to the hot water outlet from the cylinder. This video shows what it sounds like now.  No special effects have been added! https://youtube.com/shorts/29AAeSHsIVw

er, the plumber needs to revisit and do the job properly.

"Never time to do it right, always time to do it twice!"

Posted by: @jamespa

↑

Sorry to hear about your problems, which your installer should be fixing however...

 ...

has it been configured as a volumiser or is it still a buffer tank?  There appears to be a 5C difference between heat pump flow and heating flow, there should be minimal temperature difference!

Posted by: @sandvika

↑

... Does the buffer tank even need to be configured as a volumiser, or is it completely unnecessary?

Probably not, but a volumiser (plumbed either in the return or (better) in the flow NOT between the two) will do no harm so, assuming its been correctly plumbed this is not the problem.

 

Thank you, James.  The installer has received the BUS grant but is not expecting my payment until the issues have been resolved. I think I'm lucky that I don't need the space heating right now.

It was agreed on the first day of the install that the buffer tank would be plumbed as a volumiser on the return, which I understand to be common practice. However, just reading up about it, there doesn't appear to be a particular reason for having it on the return, whereas having it on the flow mitigates against very cold water entering the system at the start of a defrost cycle.

Since it was actually plumbed as a buffer, and remains that way presently because the installer now maintains that it is correct, it would entail a trivial modification to the plumbing to remove the buffer and circulating pump from the return loop, whereas it would involve two separate modifications to remove the buffer from the flow and the circulating pump from the return. Based on the defrost cycle benfit, I would settle for it remaining on the flow and just making the one modification to the return.

Posted by: @sandvika

↑

It was agreed on the first day of the install that the buffer tank would be plumbed as a volumiser on the return, which I understand to be common practice. However, just reading up about it, there doesn't appear to be a particular reason for having it on the return, whereas having it on the flow mitigates against very cold water entering the system at the start of a defrost cycle.

You are right.  Thinking on this has evolved over the past 12 months following a lengthy discussion on this forum where we went through lots of different arguments.  The current view, at least on this forum, is that a volumiser is best on the flow, Im not sure whether a conclusion was reached about whether it should go before or after the diverter valve, the thread is here

Posted by: @sandvika

↑

Since it was actually plumbed as a buffer, and remains that way presently because the installer now maintains that it is correct, it would entail a trivial modification to the plumbing to remove the buffer and circulating pump from the return loop, whereas it would involve two separate modifications to remove the buffer from the flow and the circulating pump from the return. Based on the defrost cycle benfit, I would settle for it remaining on the flow and just making the one modification to the return

Firstly thanks for confirming it is still plumbed as a buffer, which certainly could explain what you observe with the system.  It needs to change!

You ideally want to use (if it becomes a volumiser) an input port at the bottom and an output port at the top to ensure that its properly at temperature, which slightly complicates the replumbing, although it looks from your picture that there is not really a top and bottom, its an incredibly squat buffer, so I doubt it will make much difference!