Yes, sorry Tim, a complete mis type as I was either thinking of my return temperature being lower as I typed it or meant to type higher flow rate gives lower dT, I meant bigger or wider dT.
Either way, it came out wrong, sorry!
2 Likes
How much more? Can we have numbers?
Would be interesting to see how much higher the flow temp needs to be to eliminate cycling. My own heat pump doesn’t like to go below 37°C.
Hi All @SarahH @matt-drummer @Timbones
First, Sarah,
A= Problably have a great impact and on top you run with an external Roomstats - you seem to get a completely different type of control - this is actually confirmed by my supplier.
B= I have been fuzzing around with 2093 (No use for me to change 2091/2092 since I have and use the MIM controller for room stat) and I can not measure any significant change in Cycling - I went from 8 minutes cycle to 9 minutes - so a little improvement, but still over 6 cycles per hour.
C= My supplier have stated that a 3rd. party room stat could also be in the game to force the ASHP to act different than it does now - I will wait for their reply to an email that will be sent tomorrow.
Tim, I have 8 emitters + Floorheating (FH) - FH is around 10 m2 - rest of the house is 116 m2 so total of 126 m2.
I have 100 mm. insulation in the floor, FH is 300 mm. - 100+100 mm. insulation in walls and 300-350 mm. at the roof - All windows are latest 0 energy windows.
According to my calculation - I have around 3500 watts for my use in the house at 38/35/20 degC.
dt at 3 degC - which is quite stable when the ASHP runs - goes a bit over - up to 4 and a bit under 2,95 or so - so during that 8 minutes cycle it is steady in around 3,5 minutes after it have started the compressor and until I can see the compressor shuts down - rest of the time is pwering up, and down - and around 30 sek. where it is completely off - No water circulation no nothing…
With this Samsung I have steady production around 42-45 degC, when I push it up there on manual. Then the house climbs to around 26-27 degC and that is killing me…
I would prefer a dt at around 5 but I am not able to make that - if I was able to slow down flow, I might would be able to do it, but that will affect the COP in a negative manner - so I surely wont do (Accept) that.
I was told that flow was king when it comes to ASHP - and I replaced all the valves on the emitters from Danfoss RA-N to RA-G - this way I get 2,5 more flow through the emitter - this way I have reduced flow temp from around 40-43 degC to 32-34 at 7 degC outside temp. Significant savings in kwh.
Allthough I do agree, my emitters could be better and no dought, if I replaced all to 3 layers - I problably could get a better dt - I still dought it would be like magic and solve all the cycling issues I experince.
My calculation leads to the ASHP is to large - 5 Kw would have fitted better. On top we deal with an 8 kw which uses the same Scroll compressor that is to found in the 2 larger ASHP in the series - 12 and 14 kw - So minimum power is around 0,8 kwh input which produces around 3,2 kwh at cop at 4 - which is the minimum COP I readout during production - many times it is 4,5 up 5,2 - this produces more heat power than my emitters are able to consume.
I am not so damn strong in the theory of the different factors we are dealing with - but my guess is that if the amount of produce kw matches or are below the emitters capacity - we would look at a heatpump that can get rid of the heat produced and that way create slightly better dt and COP because it will at no doughts, get longer runs…
If I am on a wrong track - let me know
/Arne
The emitter size directly affects the minimum flow temperature of the system. If the emitters aren’t big enough, then flow temp will rise until the emitters can output that heat. Or the heat pump will cycle because that flow temp goes above the set threshold.
Here’s a chart from my system showing generated heat vs. emitter output. I only get stable running when emitters exceed the heat pump’s minimum. This is way above the heat loss for my property.
Side note: heat pumps tend to not get best performance at their minimum output, so better to run at slightly higher kW output.
Raising the weather comprehension curve will eliminate cycling, but you’ll need a room thermostat to turn it off when the house is warm enough. Basically by cycling the room temperature instead of flow temperature.
Increasing emitter size allows heat pump to run at a lower flow temperature without cycling.
1 Like
this would mean increased used kwh, correct.
thinking I might should try another approach - raise Compensation Curve by 5 degC, like you propose…
Just saw an issue - My kamstrup shows 1050 l/hour - and the ASHP shows 18,6 l/min = 581 l/hour - Could this be the PRV that I have making fuzz in that?
Not necessarily. Short cycling will absolutely ruin COP, so it may actually be better to run hotter for longer periods, then turn the heat pump off for a bit while system cools a little before starting the next cycle. Overall, this could mean the same amount of heat produced for less electricity.
See Improve efficiency: Turn the heat pump off
Try raising the curve by a few degrees and see what happens.
Also, have you looked into extracting data out of the heat pump so you can have a better view of how it’s performing? Glyn wrote up how here: EmonHub support for Samsung Heatpump Monitoring via Modbus (though not as good as a proper heat monitoring package)
2 Likes
…and that’s the weird thing @Timbones.
Arne heats his whole house (including his garage!) whereas I only heat my living room (1.84m2 rads total ) plus one very small toilet room (0.3m2) - all the other TRVs stay on frost setting 95% of the time. Conventional wisdom would predict massive cycling from 4kW input (at the stated max compressor turndown) and only 2kW typical heat loss, but for some unexplained reason my cycling is ~45 mins on/45 mins off (which I am comfortable with) rather than 10mins on/off, like Arne sees 
.
I run my WC very high (40LWT @ 20 ambient, 50LWT @ -2 ambient), so the LWT spends most of its time at ~45degC, versus Arne’s 34degC. I do this to warm the living room up from ~16degC (my setback) to 21degC each day in a reasonable time (note - no roomstat influence during this time!). This gives me a lousy CoP (typ 2.75) for 14h per day, but averaged over a whole day (no ASHP usage for 10h per day) that’s an effective 4.7 which I’m happy with. (Put another way, I’d have to achieve an average 4.7 CoP or better if I ran the ASHP 24/7 to get a lower electricity bill - again totally against received wisdom.)
Go figure… 
Sarah
1 Like
Hi Arne,
Just a thought - where is your remote controller located? If it’s in a small heated utility room, or a cupboard containing your Indoor Unit which might heat rapidly from the integral hot water tank, could the (Samsung) roomstat be switching on/off rapidly (and causing your rapid cycling)?
My 3rd party roomstat is in my (large) living room which has a big thermal inertia (~11000kJ/degC) so it takes a long time to change temperature (and activate/deactivate the roomstat).
Sarah
Hi Sarah
Guess you are right - the indoor unit is in the Utility Room - around 10 m2 with Floorheating I made just before the installation of ASHP. So in there are around 22-24 degC - but as I understand it, the Thermostat is not in use as it is set up - not 100% sure though, as I write this I get uncertain of this…
If I remember correct, the only thing that controls the cycling is sensors on the flow and weather temp.
I could be wrong though…
If the stat in the MIM controller has anything to say, I have to ask… The temp shown is compensated, so I guess it was showing around 26 degC and a Thermometer showed around 22,4 degC that day, so it is compensated by 3,6 degC, so it shoulkd show the correct room temp. now.
Just made a video this morning of the cycle - 8 degC outside - cycle lenght is 8 minutes and 35 seconds.
If supplier decides to steps out of this case to make a solution - I have several things I can test
- Create better dt by change emitters from 1 and 2 layer to 3 layer. This way I get better dt, so the ASHP have more to work with - allthough I then will have to lower the compensation curve even more, and then I could be afraid I get further out of range for the ASHP to work with, but it is worth a shoot I guess…
- Get the system back with a buffer of about 300 liters and have a Circulation pump seperately to drive the emitters - internal pump then would only cycle to the buffer and not beyond that - It will demand a bit higher flow temp. than needed without buffer - but it would at least get better and longer runs, so I guess it will have a break even that is acceptable. I want the long runs
- Activate internal roomstat and move the MIM controller into our living room - to test this, like you are using a 3rd. party stat…
lets see…
1 Like
You can get a bigger dT by just lowering the flow rate.
This will not harm your COP, it will probably improve it.
10lpm at a dt of 8c is the same amount of heat as 20lpm at a dt of 4c
I still don’t understand why the high flow rate is needed.
I don’t believe that changing radiators will result in “better dt”, unless you mean something else. More radiators will mean more heat output and less cycling.
Adding complexity to a heating system rarely makes it perform better. I would be wary about going down this path.
2 Likes
Good luck with your supplier, Arne. (I’d probably suggest to mine - tactfully of course - that he read this whole thread. It contains lots of useful experience and ideas. 
)
I don’t think that you can disable the roomstat. On my MIM, FSV 2093 allows you to disable Water Law, but every option uses the roomstat. Sounds like your utility room temp might be hovering very close to the roomstat sensor’s setpoint?
Don’t forget - if you have 2093 set at “roomstat or Water Law” then your cycling might be due to LWT repeatedly hitting WL, or your utility room repeatedly hitting roomstat setting. Can you try to eliminate one of these to test which? (E.g. set the roomstat way up high temporarily?)
Most folk on this forum hate buffer tanks (even though I’m happy with mine), and if your rads in your main rooms are already 22s, I’d guess that 3 layer rads may not give you much payback.
I await your installer’s views with interest…
Sarah
1 Like
Hi again Matt.
Am I not right in thinking that a higher circulation rate (through the rads) will result in higher average rad temp, thus heat duty (from deltaT = Q / m / Cp)?
If so this would allow you to lower your LWT and thus get better CoP.
Main downside to higher circulation flow is increased pumping power, but (within limits) this is tiny compared with the reduced compressor power.
Just a thought…
1 Like
As far as I am aware, and you or anybody else can correct me if I am wrong, the lwt from the heat pump is irrelevant to heat production and therefore COP.
A lower lwt may use less electricity but that is the only way it will influence the COP.
Heat produced is just the dT between flow and return and the volume of water.
Radiators can deliver more heat at a higher dT between room and mean flow temperature.
I will get the same amount of heat flowing 42c water at a dT of 4c and a flow of 20lpm as I will flowing water at 34c with a dT of 8c at 10lpm - the heat is the same in both cases. You just need emitters capable of doing this of course.
The only reason we flow at higher water temperatures is to allow the emitters to deliver enough heat.
We would all run at the minimum lwt possible to deliver the heat we need.
Arnes problem is that the heat pump produces too much heat when running at the minimum heat output and the house gets too hot.
But Arnes radiators cannot deliver the full heat produced at the flow temperature being used, they aren’t big enough.
If Arne reduced the flow rate the Dt would increase between flow and return and the heat pump will probably run for longer before needing to shut down.
But there is no getting away from the fact that the heat pump produces more heat than Arne needs when it is warmer outside.
You are right about a higher average radiator water temperature at a lower dT and higher flow rate but the minimal gain from this is completely wiped out by a heat pump that cycles 8 times an hour.
It must be terribly inefficient and damaging running like this.
2 Likes
Sorry Matt, but I think you’re wrong at least in one respect.
The higher the LWT (for any given ambient temp) the harder the compressor has to work, and the higher the energy it needs (almost exponentially so).
However the heat generated only slightly varies with LWT (at least with the Samsung compressor).
So at a higher LWT you get slightly better heat generated, but at the cost of a big increase in energy required, so the CoP goes down (quite dramatically).
Hi Sarah,
I did caveat with regard to energy consumption., that is all that affects COP.
If the dT between flow and return remains the same then the actual temperatures are irrelevant.
When you say you run at 45c to warm your room quickly as opposed to Arnes 34c that isn’t the whole story.
You have to run at 45c because your radiators aren’t big enough to deliver enough heat at a lower flow temperature. You could heat your lounge just as well at 35c lwt as 45c lwt if your radiators were big enough.
Arne runs at 34c and gets less heat from the heat pump than it wants to deliver and it short cycles.
If Arnes radiators were big enough they could run at 34c and the heat pump would continue to run. Arne would probably melt though!
Nearly but not quite, Matt.
I want my radiators (admittedly smaller than optimum) to 1) warm up the fabric of my living room (walls, furniture, air etc) from 16degC to 21degC each day, plus 2) overcome heat losses to the environment and the surrounding rooms (which of course increase as the living room warms up - an interesting numerical integration).
I know how much heat I need to warm up the room (I’ve back-calculated the thermal inertia as ~11000kJ/degC, so to warm the room by 5degC in 4h is 3.8kW - that’s quite a lot compared with room heat loss).
I could run my ASHP for longer each day, or have a smaller setback, but my calcs show that I minimise my electricity bill by running as I do.
But thanks for your concerns…
Sarah
@Madsen68 @SarahH it would really help to have some real time LWT / RWT / power graphs to compare rather than just descriptions
I’m confused now. please post a photo of your controller.
Hi Matt
Sorry to say, currently I have no control over the Flow, nevertheless - I am sure the math add up for what you say, but the missing part in that calculation is the flow temp.
I would prefer to have 20 l/min at a temp of 38 degC, rather than 48 degC at 10 l/min - I would problably have the same heat put into every room, but should at least use less power to make 38 degC than 48 degC - That is why, at least I, prefer as much flow as possible - the high flow will make the hole emitter 38 degC with large flow, rather than 2/3 is 48degC - need more energy to make those 48 degC.
And I really do not need these 48 degC to get heat into my rooms.
Lets say I have an emitter that produces 2200 watts at 70/40/20 - the same emitter will produce around 800 watts at flow temp of 38 degC - the hole emitter is hot around 38 in the top and 34 in the bottom, best way to transfer the heat from emitter to the air - and the dt is 4.
The same calculation with 48 degC and dt of 8 shows the same emitter will throw out 1350 watts - this is actually around 550 watts I do not need.
In my mind I say this - It should be cheaper to heat water from 34 to 38 degC compared to heat from 40 to 48 DegC - If I am wrong about this, I would prefer to remount my RA-N valves to slow down flow, set the ASHP to 45-50 degC and everything should be fine?
What am I missing here?? LOL
I think, If I had the correct sized ASHP for the heat I need in my house - yes I truely get your point, that it does not matter so much about the flow, allthough I belive there still will be an upper and lower limit to what suits an ASHP best - On my installation I produce today at 8 degC outside - Input is around 900 watts and I get about 4kw of heat produced - this is way over what my emitters can transfer to the air, so thats why I experince cycling. @Timbones This is where I have a theory about my emitters - If I get 3 layers instead of a 1 layer, which I have 3 of, I will get the same effects as if I had a high temperature flow - the emitter would be hot on 2/3 and cold in the bottom where the return flow is - so this will result in a better dt for the ASHP to work with.
Example: I have recently installed a 3 layer 800x900 mm. rad in my Bathroom, before there where a 2-layer 800 x 500 mm. So a little jump to oversized Recently I run at 24 hour COP around 2,88-2,95 pretty stable when temperature is 5-8 degC outside. We have had stable temps, so I tried to turn the new installed 3-layer completely off for 24 hours - COP in that 24 hour timeframe was 2,5 - turned it back on and I got it back to 2,88-2,95 COP. So a positive effect on COP directly made by slightly increased dt, for the hole system (The 3-layer rad is warm at about 2/3 and cold in the bottom, so dt on this rad is around 10-12 - I have not measured it, since I do not have the equibment for that.)
In a perfect setup, all rads would be oversized, and then it would be easy to control dt for whatever you would like to have as dt - controlled by flow rate (Allthough I have no control, the ASHP have this built in)
Then yes, it would not matter whetever there are a high, medium or low flow.
Regarding buffer - I get your point allthough it could have a purpose, allthough it would cost a bit higher flow temps. Adding 300 liters to the system could be a factor to flatten out the short cycles that I experince (8-10 minutes) at least it is what I reaand hear fro local folks
Kill me if you find my post to long LOL I know I like to write, to make folks understand my complx mind… LOL
@Ian_Calderbank
Guess you are right - @SarahH wrote it is in fact a part of the controlling system - I was not aware of that factor… My bad…
This is the one…
