the cheat sheet on heat geek is a reasonable starter but its not too hard to actually do the maths. the equations are all public.

For me this is all about the primary pipework from heat pump to distribution point where it all fans out to radiators, whether that is a buffer, LLH, manifold or just an interesting jungle of pipes in your airing cupboard.  None of this is anything to do with radiators or their pipe work.

14kw with a DT of 20  gives (14 div by 4.2)/ 5  which  0.16 l/sec of water = 10 l/min = 0.6m3 /h  0.6m3/h is is the typical reading that I see on my (single) CH pump with the current boiler. (4.2 is the specific heat capacity of water)

14kw with a DT of 5  gives (14 div by 4.2)/ 5  which  0.6 l/sec of water = 40 l/min = 2.4m3 /h Thus 4x the flow volume needed in return for 4x less DT.

then this (google engineering toolbox hazen williams) to get flow rate and head loss.

Flow velocity in even 22mm plastic at DT20 is 0.7m/sec, 20m of primary pipe head loss 0.8m , which is fine for a boiler.

But to get same flow and head loss at DT5 needs 1.25m/sec in 28mm which is too fast, noisy etc, so you needs to go up to 35mm copper which is one size beyond what is viewed as "normal" within domestic plumbing. Its stupidly expensive, hard to work with, requires pro crimping tools that not all have. you're all of a sudden needing quite a specialist.

Thats why I was asking the question because if DT can just go to 7, flow rate comes down to 0.9m/sec in 28mm copper and we're good.

In my house, Getting 35mm copper from where the  heat pump needs to go to my airing cupboard is bonkers expensive and difficult. My existing boiler primaries are round umpteen elbows, mixture of 28mm and 22m copper and plastic, a 5m section even has to bend under a tiny roof valley in a loft. one heat pump installer wanted to do 30-40m of 35mm copper literally "round the houses". would have cost thousands.

At the moment to replace this for heatpump I'm thinking parallel identical length 28mm plastic (2 runs each for F and R) for 90% of the distance as that has same total bore as 35mm and only a tiny bit worse head loss. T'd into 28mm copper for the last and first few metres. and will be so much cheaper and easier with the places where bends are required.

Secondary side I have taken actually care of and don't see any issues there. (maths done, all rads upgraded to K2, all pipework to rads 15, 22 from distribution point to rad groups).

@editor your bigger pumps will be increasing available head and thus flow rate through that particular run, in order to move enough water to still meet the heat demand with a lower DT.  See if your pump has a flow volume over time readout on it. If you do the maths you may find that the flow rate (in m/sec) through that pipe is quite high, and as you say that pump is working hard.

What happens if the DT is 7.

Does the heat pump stop working? Is it less efficient? Will you get cold room? 

@jeff no it doesn't stop working, it will fight you, its whole mission is to make a 5 degree delta t so it will slow the compressor therefore output kWs fall. you will be marginally colder and it wont run flat out. its all about flow rate

@editor dont worry to much about it, its just maths, if the delta t is 5 degrees and the flow rate is 30l/ min you are going to be able to deliver 10kW. you just wont get anymore than that, you either have to have a bigger delta t or a higher flow rate. it doesnt change anything it just means you cant deliver the units full potential. it will never go flat out

@grahamh surely not "just maths" though? he's got an 18kw specced heat pump, but he can only get 10kw out of it if the pipework will only do 30l/min? so wasted the ££ on 44% of the unit's capacity?

if the heat demand from the control side is there, and current flow rate is insufficient for demand, do these heat pumps tell their circulating pump to keep on pushing harder until either target flow rate or pump max capacity is hit?

@iancalderbank

There are some heat pump controllers which allow the required DT to be set to a particular value, within a specified range. Have a look through the installer manuals from the most common manufacturers.

Posted by: @grahamh

@jeff no it doesn't stop working, it will fight you, its whole mission is to make a 5 degree delta t so it will slow the compressor therefore output kWs fall. you will be marginally colder and it wont run flat out. its all about flow rate

I would have thought that the opposite would actually occur. Consider the following.

A heat pump is running happily with a Water Flow Temperature (WFT) of 35C and a Water Return Temperature (WRT) of 30C, hence a Delta-T (DT) of 5C. If the fixed required WFT or Weather Compensated (WC) WFT now increases to 37C, this would speed up the compressor to increase the WFT, initially producing a DT of 7C. If the heat loss in the home has increased, producing the larger DT, then the water pump within the heat pump will also increase its flow rate to transfer more heat energy. Eventually the heat supplied will balance the heat lost and the WRT will start to increase until it achieves the specified DT of 5C.

Therefore the compressor speeds up to produce more heat energy, and the water pump speeds up to transfer this heat energy to the heat emitters, until the system once more achieves equilibrium.

The reverse will of course happen if the WFT is reduced, until the system reaches the point where it cannot pull back any further and cycling will start.

Good points but if flow is restricted the pump is flat out and compressor slows, if there is spare pump capacity (very rare) what you say is true. the secret is to squander days of your life watching all this happen with a service computer connected to the unit. its dull and its what i used to do but im over this now. let the people who sell the kit do it, at least they get paid for it.

@derek-m any specific pointers? I also want to avoid units that require proprietary room controllers - I need one with an open standard (230v or volt free) input for "heat demand".

Posted by: @iancalderbank

@derek-m any specific pointers? I also want to avoid units that require proprietary room controllers - I need one with an open standard (230v or volt free) input for "heat demand".

Heat pumps do not work particularly well with thermostat control, which is on-off. Most if not all of the modern heat pumps incorporate weather compensation control, which varies the amount of heat energy produced by the heat pump, dependent upon the outside air temperature. In basic terms as the outside air temperature gets colder, the heat pump works harder and produces more heat energy. When correctly optimised, this will try to match the heat energy produced to the present heat loss, to try to keep the indoor temperature fairly constant at the desired temperature.

Some heat pump controllers also have what is often called 'adaptive control', which has both indoor and outdoor temperature sensors to keep the indoor temperature close the desired setting. For adaptive control to be used, the controller needs to located in the area whose temperature is being controlled.

As far as I am aware, most controllers come with the system as a package, since they are often required as a Human Machine Interface (HMI), to allow the heat pump to be configured and commissioned. It also allows for settings to be adjusted and schedules to be set. Therefore adding a thermostat, in many ways, is counterproductive and additional expense.

If you watch the 'Heat Geek' videos, as he states, and I have been advising for quite some time, the first think to do is improve your level of insulation as much as possible. Then perform, or have carried out, detailed heat loss calculations. Only then will you be able to ascertain what size of heat pump is required to meet your particular requirements.

Posted by: @grahamh

@kev-m thats not true , you dont measure it as a total box, you need to take into consideration the lumps and bumps and more importantly the dips. its 0.58m^3

The regulations say "The volume of the air source heat pump’s outdoor compressor unit (including housing) must not exceed 0.6 cubic metres", the Samsung spec sheet says it measures 1270x1018x530 mm and in its picture the housing looks like quite a smooth box-like shape.  It's not obvious how it comes in under the PD limit.  Where is the missing 0.1 m3?

Posted by: @iancalderbank

if the heat demand from the control side is there, and current flow rate is insufficient for demand, do these heat pumps tell their circulating pump to keep on pushing harder until either target flow rate or pump max capacity is hit? 

In our case the distribution pumps are set manually and the heat pump has no control over them: low, medium, high. On high, we’ll never exceed the 30l/minute. Our metres show that our average flow rate is 28l/minute. Enough for the heat pump to stop alarming and turning off.