Samsung ASHP (Joule) poor performance - help needed!

Hi Christian, yes that is very helpful. I’ve started going down a similar route - following the suggestion from @Old_Scientist I’ve set my system to a flat WL curve of 35c and am I am seeing much better performance with little difference in the indoor temperature (not enough for my wife to notice yet)!

The last 6 hours at approx 5c outside temp has given me a COP of 3.35 which might not sound much but is definitely an improvement. This also includes a period when the rads were calling for heat where the COP was less than 2.5.

I’ll continue to refine the flow temps and chase the design documents as I need to know whether the system has been implemented to the spec and what my system volume is.

Apart from this I need to work out if changing my fixed speed pump to PWM is going to make a material difference. Thoughts?

Regarding what turns things on/off:
a) any of the thermostats calling for heat
b) DHW

My Gen 6 seems to make some use of the PWM feature whether set to 100% or 70% max pump speed. The control strategy in the Gen 6 tries to maintain a deltaT of 5degC. If it finds it cannot maintain the deltaT it may well use the PWM feature. For instance, here is an example …

https://emoncms.org/app/view?name=SamsungHeatpump&readkey=88a6c522550d5fd6ffece80a646b4952&mode=power&start=1738226535&end=1738248210&cop=1&flow=1&carnot=0.47

It initially starts up and settles back to 70% flow (19.7l/m), then overshoots and restarts. After the restart the desired flow temperature is slowly falling and now it adjusts the flow rate using PWM and settles on 17.8l/m while maintaining a deltaT of 5degC-ish. It does a pretty good job of this, why it couldn’t do that from the initial start, who knows …

Many sources on the interweb state the 5degC deltaT gives best efficiency but I haven’t found anything that gives a good thermodynamic reason why this should be so. It seems to be dependant on the refrigerant, synthetic refrigerants are usually 5degC, whereas CO2 and Propane are higher.

Years ago, when I still had 2 LPG boilers I had a brilliant idea; get WiFi TRVs for almost all of the 23 radiators in my house and let them call for heat when they need it.

I must have been drunk at the time as it was truly one of the worst things I have ever done in terms of the performance of my heating. (I still have 4-5 of them which I will probably put on Fleabay next autumn)

Especially with a heat pump you don’t want lots of zones calling for a bit of heat.
You want to deliver a nice constant load and if an area gets too hot you want it to close down gradually; maybe reduce the flow but don’t cut it off.

If any zone just stops, its water will cool down then when it comes back on you will use loads of energy to get it back to where it was.

I’m guessing that your 5 thermostats each have a schedule for when the zone is in use and when it is not.
I’m also guessing that there is a big difference between the target temperatures of these time periods (maybe 21c on, 7c off).
If that’s correct, my first suggestion would be to change the off periods to be just 1 degree cooler than the target.
Run it like this for 2 days to see how your energy consumption changes and how your CoP changes. I say 2 days because when I was determining my curve, 1 day was not enough to see the full effect of a significant change.

Getting the best from a heat pump can be really counter-intuitive as we have been used to heating with boilers, double the flow temperature and speedy changes in the living areas.

I’m pretty happy with my weather compensation now, but I realise that I now have to decide whether I want a comfy house all the time and great CoP when it’s above 5c, or do I want to save a few quid by switching it off for a few hours overnight? I could to that as I can see that if I turn it off for 3 hours or more, it takes less than the energy saved to get the system going again when it comes back on.
Lucky me, I get elec for 7p/kWh 23.30-05.30 so it’s not really worth turning it off before we get up in the morning. Also when I get up for a wee it’s not unpleasantly cold.

I don’t think you should worry about delta T or the PWM capability of your pump until you are basically happy with what the heat pump is doing, that’s all fine tuning.

Edit: I noticed earlier that your DHW seems to be coming on at random times, so I think the HP is keeping the whole tank at target temperature all the time. Depending on your elec tariff it’s probably best to schedule the DHW to run soon before you expect to need hot water and let it cool outside those times.

3 Likes

I don’t use any schedules, just let the system run in “manual” mode. The UFH thermostats are set to same value and the upstairs a little cooler. We don’t set back the UFH but do manually set back the upstairs by 0.5c when we go to bed. I also have a cheap rate overnight so don’t mind heating the house then. You can see from this morning’s data how when only the radiators are calling for heat the DeltaT drops, the heat pump uses less energy and the COP also drops. HA shows the PWM signal is dropping as well but as I don’t have a PWM pump…

I’ll have a think about DHW when I am happy with sheathing as I’m trying hard to resist the temptation to change too many things at once! It’s interesting that you see the defrost cycles kicking in which as discussed in Samsung R32 ASHPs - when is a defrost not a defrost? does have the effect of pulling my DWH temp down.

Keep working on the WC curve and bringing the flow temperature down :+1:

I have an 8kw Samsung heat pump, I’ve noticed a large improvement when PWM is working compared to when it isn’t, as much as 1.0COP loss when dt was 2, compared to dt of 4-5.
Same flow temps and same outdoor temp.

I’m curious as to why you have so many zones. The general concensus is to run everything open all of the time, only using TRVs in bedrooms to limit overheating. If you have 5 zones but only a couple are open, your emitters wont be able to get rid of the heat and your dt will reduce. (PWM will help somewhat)

Also, balancing is extremely important, even if emitter sizes are corrsct for the rooms, the water will take the path of least resistance, and different pipework to different emitters will affect it. Balancing sets the correct resistance across each emitter, so each emitter has the correct flow rate. You could end up with all the water cycling around a small number of emitters, again, limiting how much heat can be emitted from the water.

1 Like

I wasn’t involved in the design of the system so not sure why it is like it is.

Earlier today the radiators were heating and despite having a flat WL curve at 35c, the flow temp get rising which based on earlier advice may week indicate exactly that the emitters are not releasing the heat. Would a lower deltaT help the emitters release heat or is that down to balancing?

I’d suggest doing another text with all of your thermostats set max to open up the full heating circuit. This will enable your circuit to emit more heat and, raise your dt, and theoretically raise your COP. Your heatpump is too big to heat fractions of your house. This isn’t how heatpumps work efficiently.
If you are happy with the results, keep your system open all the time and balance the emitters correctly.

Thanks. If I’ve understood I think you are saying that I need to leave every thermostat open and rely on WL? I’ll experiment today especially as I only have one zone and I think the UFH and Rads really need to be running at different flow temps for this to work properly.

I think I need to call Joule again tomorrow to discuss what I’ve learnt this week.

Fitting fans to the radators may allow them to work at same flow temperature as UFH. DIY Controller for radiator fans

I had an interesting! Call with Joule today. I wanted to know:
Why my 12kw system seems to be limited to 24-27 ppm flow rate when Samsung say 35lpm is needed to be most efficient.
The answer was to ask Samsung tell me how to achieve 35lpm when heat exchanger plate has 1” pipework!

Should I balance the radiators to deltaT 5
It’s better to balance by flow rate rather than temperature.

I then asked if I would benefit from a PWM pump rather than fixed speed when I can see the ASHP seemingly reduce the PWM signal
I was told that I couldn’t replace one of the primary pumps with one PWM enabled. If I wanted a PWM pump it would need to be fitted as a secondary pump feeding the buffer tank. I didn’t really follow the logic of why this was the case!

He ended up advising me to set 2091 to 3 so that I didn’t exceed the WL curve target temp, however this resulting in short cycling so I’ve set it back to 1 again.

All in all, no further forward! I welcome your views on any of this particularly to comments re PWM pumps.

This is not how I read the Samsung documents. It needs 35l/m to reach rated maximum output, at lower flow rates the maximum output will be reduced pro-rata.

1 Like

The Freedom Heatpump heat loss calculator suggests that the 12kW Gen 6 Samsung will produce a maximum of 10.5 kW at design temperature, so with no glycol this equates to a flow rate of approximately 30l/min at a DT of 5C. I presume that if the flow rate is only 25l/min it can still output 10.5 kW but at a higher DT (6C in this example). I don’t know the ideal DT for the Samsung, so a DT of 6C may produce a better COP than a DT of 5C or it may not. However, I am not convinced that high flow rates are as important as some people seem to think, especially when OAT is several degrees above design temperature.

Thanks for clarifying re flow rates - that makes much more sense.

Any advice on whether I must have a buffer tank to make use of a PWM pump?

Hi @MikeJH,
In my experience, a higher water circulation rate (thus lower dT) can actually improve CoP slightly, since you also get a lower dT in your radiators, which allows them to run at a slightly lower inlet temperature for the same average temperature thus heat output, which allows you to drop your LWT a little, which directionally improves CoP. You have to trade off the extra circulating pump power requirement with the reduced compressor power requirement, but within limits (say down to a water dT of about 2degC) increasing the circulation rate can be economically beneficial (though there may be other controller constraints that prevent this of course…).

1 Like

This is rubbish. I have no buffer, 1 pump and it’s a PWM pump. The Samsung training material shows that this is a perfectly valid installation option. My installer initially installed a non-PWM pump, and I got him to change it.

I had another issue with getting PWM to work, but Samsung sent me a new PCB for the indoor unit, and it that fixed the issue.

One of the main issues I have is pipe noise, so I prefer to have PWM to lower the flow rate when it deems it necessary. I suspect there’s a restriction/s in my pipework, because it makes a hell of a racket under the bedroom floor when at full speed on space heating mode, which reaches around 14.8lpm. On DHW mode, it’s pretty much silent and and reaches 19lpm. The main feeds from/to the heat pump run under my bedroom floor, but also the main feeds to/from the radiator circuit and a lot of tee offs.
This does limit my heat output at dt5 to around 4.8 KW, but my home’s heat loss is around 4.5KW, so this isn’t an issue in of itself.
I’ll probably look getting some of the system re-piped in the summer, to reduce run lengths and increase pipe diameter in some places, but I’ll settle with it how it is for the time being.

Hi @SarahH ,
That is a very good point. It is one reason I allow my kitchen radiators to run at low dTs as the kitchen is one of my coldest rooms. However, some ‘experts’ seem to say that for the best efficiency, overall dT needs to be controlled - hence the discussion on PWM pumps. I imagine that for some heat pump models, a dT of 2 is less efficient than a dT of 5 and I have even seen a dT of 8 being quoted as the most efficient dT.

I’ve finally got hold of the design documentation from Joule and it appears what they specified and quoted for was not installed. The most notable difference is aluminium radiators have been replaced by steel (1 K22, 3 towel rads and the rest type 21).

From what I can work out from google it seems the radiator outputs of what the builder has installed may be similar to that specified by Joule though most of the inter web talks about dt30 and dt50 and the design dt is either 20.5 or 24.5 (at 3=45C flow). Is there a simple rule of thumb that allows me to convert from dt20(ish) to dt30?

Google also says aluminium is recommended for AHSP because:

  • Heat transfer:
    Aluminium conducts heat significantly better than steel, meaning it can quickly transfer heat from the water flowing through the radiator to the room.
  • Low water content:
    Aluminium radiators have a lower water volume, allowing them to react faster to temperature changes and providing better control over your heating.
  • Energy efficiency:
    Due to their fast heating capabilities, aluminium radiators can potentially help reduce energy consumption when paired with a heat pump.

So, I need more of your expertise please:

  • Is the difference between teen and aluminium material and do I have a case for insisting the builder installs the “correct” radiators?
  • Would the difference also explain some of the issues I am having with the performance of the radiator circuit?

THere are various web pages giving conversion factors. This page has a table and calculator https://www.radiatorsdirect.co.uk/delta-t-conversion.

I rather think those advantages of aluminium radiators are pretty specious in terms of a heating system. Yes, aluminium is a slightly better conductor of heat than steel, but given the low thickness of the radiator walls that’s not going to produce a discernible difference. Low water volume is also unimportant for heat pump systems where they tend to be run for long periods, so fast response isn’t needed. It could even be a disadvantage if you want to have a system with a large water volume to keep the heat pump happy. The efficiency line is complete rubbish. That’s obviously marketing material written by the same sort of people who try and sell some resistive heaters as more efficient than others.

(Low mass and low water volume could be a minor asset in systems that stop and start often, but those aren’t going to be heat pump systems.)

If the nominal output is about the same I can’t see that it’s worth pursuing. If the aluminium radiators were chosen specifically for their appearance, then I would complain.

3 Likes