Oversized ASHP (Samsung), flow rates, LWT and dT


We have just had a Samsung 12kW ASHP installed with a Joule pre-plumbed cylinder. We have 11 new radiators, 28mm copper from ASHP to cylinder, 22mm copper dropping to 15mm for each radiator tee. There are two Wilo pumps fitter on the pre-plumbed cylinder, one on the flow side, and one on the return and flow rates are fixed at ~26L/min. There is a 50L volumiser fitter to the return.
Heat loss calcs for the property are 7.4kW at design temp of -2C, and our radiators can achieve this output at a LWT of 40C for a set room temp of 20C (dT 20C).

LWT, flow rates and dT
I have no specific monitoring (yet), so am just taking readings from the Samsung control panel and physical flow meter.
My intention was to run constantly at as low a LWT as possible, using WC (Water Law). I have read some excellent “oversized heat pump” threads here and learned a lot from the very knowledgeable posters who have contributed.

I tried the heating out for a couple days last week when it was a bit colder (12-15C). I initially set my WC curve for a LWT of 40C at -2C ambient, and 30C for 15C ambient.

We ran the system for a couple hours - I set the heat pump NOT to cycle on LWT, and to be controlled only by the room thermostat, set high to 21C to force operation (and prevent cycling). The system happily warmed the house from 17C to the target temp of 21C. If it cannot run 24/7, and is going to cycle, I want to control the cycling by turning on/off at the room stat.

However, the unit was never able to achieve its set LWT of 30C. The lowest the LWT dropped to was 33-34C with the system fully open, and dT on flow and return was 2-3C. Flow rates are fixed (non-PWM pumps) at ~26L/min.

From my reading here, I understand:

Heat output = specific heat x flow rate x dT
Heat output = (4200 x 0.95 (glycol)) x (26/60) x 2.5 = 4.3kW

which matches the output of our radiators at 33-34C flow temps. Our radiators can only emit 2.9kW at 30C flow temps.

The minimum observed power input to the heat pump was around 900W, and from the observed LWTs above I infer the system is generating somewhere in the range of 4.2-4.6kW.

So I conclude that I’m never going to be able to run a LWT below 33-34C. Is there any benefit of setting the mild end of the WC curve for LWT’s below 33-34C (e.g, 30C @15C ambient), and just let the heat pump find it’s own minimum (is this bad for the heat pump?), or should it be set slightly above at 35C? My goal is to minimise running costs.

I appreciate a target dT of ~5C between flow and return is recommended. Is my dT of 2-3C an issue, or should I not worry? (as an aside, presumably since dT is so low, I can infer my pipe run between ASHP and cylinder is very well insulated with low losses?). I should be able to achieve a dT of ~4C in winter as our heat output demand rises.

From the heat output formula above, halving the flow rate would double the dT. Is reducing my flow rates something I should explore in order to increase dT? There seem to be two valves?? either side of the two pumps - maybe I could close off slightly to reduce flow rate?

So for now, during milder weather (likely Autumn now) I plan to run the system at as low LWT as possible, and run for a few hours and then turn off once the house is at temp. At ambient temps of 10C and above, I think the heat pump is oversized and I will not be able to run continuously, but below 10C I should be able to match LWT to the heat loss of the property to maintain a steady temp.

One other observation - the unit is sighted south facing in direct sunlight. The ambient temp (on top of home screen) rises significantly in direct sunshine (not accurately reflecting air temp). If this is the temp the unit uses for it’s WC/LWT calculation, it will be significantly skewed. Today it was reading 24C when it’s more like 19C ambient.

I am largely new to this, and learning as I go, so please feel free to highlight any mistakes in my logic above and point me in the right direction.


I also have an oversized heat pump and its a Samsung
I have a 16kw where as my heat loss is around 12kw
I have been consulting with Samsung tech on occassion and i asked the question of running the heat pump continuously on room stat only and they said that it was better than cycling.
So with that in mind i do what you have said by using the stat as the control but i use a third party stat as when you do this you get the option on the screen to alter the flow temp by + - 5 degrees so you can slow the warm up down or increase it as you please. So in the winter i tend to lower it by 5 at night and then increase it in the morning as i have my display at the bottom of the stairs so its not a problem
It does take time to get the right water law settings but once you have them about right its easy to just tweak with the display
As the heat pump has a minimum compressor speed you will find it cant go to the low set point as it cant get rid of the heat, it maybe your emmitters could be increased in size to allow more heat to be released and then you can lower your flow temp. So if you require 7.4kw you require a flow rate around 21lpm
I am not a fan of the Joule set up a friend of mine has one and it looks very restrictive and over complicated but you have what you have.
The Wilo pumps have different settings marked by the 3 boxes and then i believe they have 3 set points for each type of control. One is constant curve one is constant outlet pressure and the other proportional pressure.
You could have a tweek on them to see what you can improve.
I do have a PWM pump on mine and only one and run completely open circuit.
You might find balancing your rads can improve flow rate and this can take a while
There is an app that gives you flow rates for rads but if you know the kw output at 40 degrees then you can work it out
Ill have a look to see what the app is called and send it
You had a lot in your post so may have missed a lot but happy to help where i can and there will be others who may say different to me but i have had mine nearly 5 years and normally in the winter it runs around the 800 watt which is close to minmum speed to keep a 200 sqmetre 220 year old solid stone house at 20 degrees in all rooms. We just have a smart stat on our bedroom that closes at around 7pm and opens again at 7.30am
Hope that gives you some help

The app is IMI hytools i find it very useful as i do install heat pumps

@johnfresh Thank you for your reply.

Yes, as I use an add-on room stat, I too have discovered I can easily adjust the flow temps up and down by 5C on the Samsung controller to tweak the WC (weather law) curve. Useful tip, thank you, and it certainly makes fine tuning the WC curve easier!

Interesting that you use this in winter as a set back overnight - I had envisaged just turning the room stat back to 18C overnight and up again in the morning, which would turn off the heat pump (until the temp dropped back to 18C - effectively long cycling), whereas your method allows the unit to run continuously at a lower flow temp but putting out insufficient heat to maintain daytime set temps so the house slowly cools. I wonder which method is more efficient / costs less to run?

I also managed to find a manual for the Wilo pumps (one is on the flow and one on the return). The Samsung/Joule manual suggests the primary pump (on the flow) should be set at max and flow controlled by the other on the return. Mine were set at max (III) on the return and II on the flow by the installer giving a flow rate of 26L/min. I calculated a flow rate of around 21.5 would give me a dT of 3 (better than the ~2.5C I was getting) at minimum output, and dT of 5C in winter. I dropped the primary pump on the flow to II which reduced the flow rate down to 24L/min, and then dropping the secondary pump on the return to it’s slowest speed (I) keeping the primary flow at II give a flow rate of 21L/min, at least on the DHW (I’ve not had occasion to try the heating again yet, and probably won’t now until autumn). So I have plenty of easy adjustment to manually set the flow at 21, 24 or 26L/min to achieve a decent dT (Samsung manual recommends 20L/min plus, and the unit will error at 12L/min, so my lower end of 21L/min seems safe). Reducing the flow does not seem to have had a demonstrable effect on the DHW cycle.


I am going to a customer next week who has a Joule tank set up so will have a closer look at how it works
I’m going to alter pipe work on the primary which was done by a so called professional installer ŵho is supposedly MCS accredited and its an awful mess but that’s a longtime rant of mine as I have seen many MCS installs that are not good and it goes to prove that word of mouth is a better guarantee of a good job than a certificate
I will also find the joule manual on line to have a look through

I firstly tried the control by using the stat but as the house took a long time to drop in temp the whole system had dropped so low that when it did call again in would run flat out to get back up to the flow temp which seems to make it need a defrost quicker.
The one benefit of having a two fan unit is mine doesn’t seem to freeze up as easy as a unit down the road that I walk by which is single on the cold days. I often see that one frosted up when mine hasn’t done all that day

I’m not sure why Joule doesn’t have a one pump PWM pump set up but maybe that will become clear when I have a look next week

Hi @johnfresh,

I’d really appreciate your thoughts on the Joule set up. I have a pre-plumbed slimline Joule Kodiak, all installed in the same footprint previously occupied by our oil combi-boiler (photo taken before pipework was lagged):

Overall I was very happy with the job the installers did (to my untrained eye). The set up seems quite straight forward. As mentioned, there are two pre-fitted 50W pumps, on flow and return, on the Joule cylinder. I have two expansion tanks, and a 50L volumiser on the return (I knew I did not want a buffer tank or LLH). I probably don’t need the volumiser (they insist upon it to protect those who may shut off TRVs), but it should at least help on defrost cycles. It is plumbed on the common return so I am unfortunately heating an extra 50L of water to 55C during the DHW cycle and then leaving that heat to dissipate into the house in summer - not ideal - but the extra slug of 55C water will end up in the radiators in winter.

Interesting on the two fan unit and defrost cycles. Presumably not running the unit so hard (flat out) and having double the surface area helps reduce frosting of the unit?

I just need to build a closet around it now to box everything in.

Hi Only one quick observation

Is there a direction arrow on the strainer as they usually go the other way round if its coming away from the bottom of the volumizer, they usually do have one

I only have the primary pipework and the pipes to the tank lagged as if the heating pipes are in the envelope of the house it all helps to release heat, cupboard may get warm unless you have a louvre door on it

I will get my head round it next week and let you know my thoughts but from memory I did have a few concerns with the set up like the flow meter looked right next to the pump and isolation valves so that may not be reading correctly as they do recommend a number of pipe diameters of straight pipe in and out of them so there isn’t turbulence in the flow meter but not sure if the Samsung is differential pressure either side of an orifice or electronic waves like the ones im more used to on the water industry. Im guessing pressure.

If there is a manual flow gauge and the two match up then I guess its ok

Will mail next week


Thanks @johnfresh

I think the Y strainer is fitted the correct way - it’s in the return pipe just before it enters the volumiser. The other pipe leaving the volumiser on the LHS is the 28mm return to the heat pump.

I take your point about not lagging any pipes on the central heating side as they can effectively act as additional emitters to help dissipate the heat, which in turn helps with dT. I can remove the lagging on the 22mm CH pipework.

There are two flow sensors, both on the primary flow side and are part of the pre-plumbed cylinder. There is a mechanical flow meter fitted just to the bottom right of the front cover that you can see, just below the primary (flow) pump, and the electronic flow sensor is fitted just above the pump (you can see the black cable showing to the RHS of the Samsung controller). Both are connected directly to the pump, so no pipe runs on the outlet of the mechanical meter or the inlet side of the electronic meter. Both meters read the same, so at least they are consistent.

Not sure why they didn’t fit the strainer after the volumiser as the volumiser acts as a dirt separator due to the velocity being low it allows dirt to settle to the bottom. Maybe it was an access concern to be able to clean the filter

That’s good that at least both are reading the same


1 Like

Hi. I have been to re hash the pipe work on the system with the Joule tank and i have also had a conversation with Samsung as i vouldnt get enough flow for 12 kw through the joule system
They replied that in their opinion the Joule tank is only ok for up to 8kw as the hydraulic set up restricts flow to below 30lpm
I can confirm this as i re made the pipe work so it was only 3 metres to and fom the heat pump and couldnstill only get 27lpm through 28mm pipe work
My customer was told they were fitting an 8kw but they turned up with a 12
I dont suppose it would be hard to remove the Joule set up and fit either 2 solenoid valves ot 1 diverter valve and hopefully they havent made too many changes to the control board
I prefer to use the actuatedbT port ball valve as they are full bore

Not sure that helps you much

@johnfresh Interesting, I was looking at my Joule tank too the other day, and the integrated pipework on the tank and manifold is 22mm, which obviously limits the amount of heat that can be transmitted to around 6kW max at the recommended 0.9mps and dT5 (further reduced by 10% for use of glycol)
My heat loss was calculated at 7.4kW, and the integrated flow and return pipework in the heating system loop from the Joule tank is, you guessed it, 22mm before it branches off into two 22mm runs for each end of the house (so the two runs can carry the required 3-4kW each, but are throttled at the connections to the Joule pre-plumbed tank)
Hence I am somewhat annoyed that they insisted on fitting a 12kW unit, which is oversized for our property based on their inflated heat loss figures using a system that can theoretically never deliver more than 5.4kW due to the restricted pipe sizes used on the pre-plumbed cylinder!
In reality, I suspect I will be able to get sufficient heat output from my system in winter (we will see). I have flow rates of 26L/min from the two Wilo pumps, and can turn the pump speed up further if required. But I would have much preferred to have had an 8kW unit sat outside, than the 12kW I have.

I guess it would be best to take 28mm to where it branches off to the two loops.
Obviously your hot water flow is determined by the coil inside the tank which im not sure what size they use. My tank has two coils but is also restricted by 22mm pipe but i do get 35lpm through it on hot water and over 40 on heating and thats with just one pump
But unfortunately as mine is 16kw and the flow for that is 46lpm the pwm pump doesnt modulate as it never gets to balanced flow to delta t 5 but it does tick along bicely in the coldest days at about 800 watts and keeps the whole house at 20 degrees
Mine really should have been a 12kw but they didnt have any in stock at the time so i got what they had.
Although talking to samsung the 16 and 12 are just different in software but if i reprogrammed it the warranty would be invalid. Maybe next time ill have the right one its been in 4 years and other than me altering things to improve flow its been great.

@johnfresh Yes, the Samsung 16kW and 12kW units are mechanically identical, just the firmware in the 12kW unit limits it’s max output.
I can try turning both pumps up to max to see what flow rates I get. The Joule tank has two Wilo pumps fitted in series, one on the flow pipe and one on the return. They are non-PWM fixed speed pumps, and the installers left them set with one on max setting, and the other turned down to number 2 (setting 1-3 available). On those settings I get flow rates of 26L/min on both heating and DHW loops.
I had turned both pumps down to 2 (seemed logical to me that two pumps in series should be set to the same speed so one isn’t trying to push harder than the other), which reduced the flow rate to 24L/min, thinking that may help improve my dT (which was only ~2.5C). I’m guessing I may need higher flow rates in the colder winter weather? (I’m still learning trying to understand how different parameters affect overall performance).
The system has performed great on the test runs I’ve been able to do in the mild weather since installation, but obviously the heating has not had a full test yet at winter temps.

The pumps may have different friction loss dpending on pipe runs/ exchange losses so very hard to know what they are doing when in series
I dont get the thinking behind the Joule set up