Again, not quite. All that my (non-smart) roomstat settings do is to ensure that there is not more than 1 “on” and 1 “off” signal every 15 minutes (i.e. 4 cycles per hour). I’m not insisting on 1 “off” period in the 15 minutes - I leave the roomstat internal programming to decide whether an “off” signal is required within that period. Cycling appears to improve my CoP (see above), so as long as it doesn’t exceed 3-4 times per hour (compressor mechanical wear considerations) then I don’t foresee too many problems.
Yes (except that you mean #2091/2/3 I think). I have #2091 set to “3” so either WL or roomstat are allowed to control - whichever hits its setpoint first gets first dibs… (ditto on WL hysteresis or roomstat TPI setting).
Sort of. Programme intervention is rather different to controller intervention. Your “shifts” are defining the times that the controller is active/inactive. You have a controller setpoint that it cannot be expected to meet (24degC) so it stays “on” until the end of shift. With this setting, you aren’t giving Nest a chance to “learn” what you want. You’ve paid a lot of money for your Nest roomstat - why not let it repay your investment and allow it take over, at least for a few days?
Because what the best does is garbage. It registers when you activate it by passing in front and registers it as a moment when the house needs to be warm. I was seeing the boiler (before the HP) bring triggered at some stupid times because of this. I’ve switched the learning piece off immediately. Again, total garbage.
Yes sorry I meant 2091/92/93. I see what you mean regarding the cycling but so far I haven’t done any. My system runs constantly. I may try the other options on the FSVs but still don’t see why people keep saying the wired thermostat is better than the nest. To me they look exactly the same
Most wireless roomstats are not well suited to heat pump operation. As you can see from my plots above, I’m getting what would be called short cycling (every 15 minutes) because my roomstat cycle rate is set to 4 per hour, and I don’t think I can set my particular roomstat (a Honeywell T3R) much lower than this (i.e. longer cycle times). The T3R gives a wonderfully constant room temperature, but at the cost of short cycling.
On the other hand, a wired (electro-mechanical) roomstat with a hysteresis of say 1degC wouldn’t switch off the heat pump for an hour or more - it takes a long time for a well-insulated room to cool down by 1degC. But this swing in room temperature isn’t to everyone’s taste.
Nowadays you can buy wired thermostats with adjustable hysteresis, so you can set one that balances your comfort needs with cycle frequency. Maybe your Nest can be taught to emulate a wired controller and give long cycle times with an acceptably low room temperature swing?
That is easy to achieve, even with home assistant. The question is if I want this stop and go.
Currently, I set my thermostat to 24 degrees which I know it’s impossible to achieve so the system will just run until I send the off signal. I was told by the installer I should have the pump running as long as possible. I keep reading instead to have it cycling would be better but I cant understand the why as every time it starts it requires more power.
If I set my temp to 21, it will reach it eventually, then stop, then will restart after the house drops by 0.5 or 1 degree I can remember. That will create cycling for sure.
On the cop side, last night with roughly 10 degrees outside and a very low flow temp of 34 degrees circa, 3.2 cop. Unbelievable. Same weather the day before 3.9. I really don’t get it
There has been a lot of discussion on the Buildhub forum about running your ASHP low and slow. In fact 24/7 using WC and the heat pump temperature as the target, so no thermostats at all. Obviously you have to have a system designed to do this, i.e. larger radiators etc.
We don’t do this because we have a massive insulated slab, so batch charge on the Intelligent Octopus Go rate. We only have UFH, no upstairs heating we’re almost almost passivehaus. So we use our thermostats like you Frederico, i.e. as limiters and to kick off the heating and stop it at the end of the cheap rate, the ASHP runs continuously unless it’s so warm that one of the thermostats hits the limit (24C). The thermostats are fancy smart ones with all the ‘smarts’ turned off. The advantage of smart ones for us is simply that we can see the 3 zone temperatures remotely as well as in the shoulder season, turn the heating to come on, if it looks like it’s going to be cold and we are away from home.
We had to replace the original ‘smart’ thermostats with the current ones, because we couldn’t turn off the ‘smart’ feature to start getting the house warm before the actually scheduled start time. So the old ones would start the ASHP before the cheap rate. So a warning to anyone wanting to do this, make sure you can disable the ‘smart’ features.
The table you shared is for SCOP values, so seasonally adjusted averages, not at a specific OAT of 2C as is currently the case for you. Of course efficiency will be significantly improved as OAT increases.
I’m looking in the Samsung Technical Data book, at the heating capacity table on page 29.
For your 5kW model, an OAT temp of 2C and flow temp of 40C, it shows a heating capacity output (HC) of 4.54kW for an an input power (HI) of 1.54kW. Dividing one by the other gives an indicative COP of 2.95 under those specific conditions. (Sorry, I don’t have a direct link to the PDF any more)
I’m more concerned that today, 1:30 am to 6:30 AM, 10 degrees outside, LWT of circa 34 degrees, COP was on AVG 3.1 Vs 5.something from the table. This is what worries me. What is that I may be doing that wrong.
I believe that the Samsung published performance tables like the one @Old_Scientist included above:
Represent performance at a (synchronous) 50Hz compressor motor inverter frequency. At other compressor speeds the HP to P ratios may differ quite markedly - Samsung will probably have designed for a compressor Best (thermodynamic) Efficiency Point (BEP) at 50Hz. My own data suggest that this falls at lower inverter frequencies, maybe (I suspect) because the circulating seal oil flow doesn’t fall in proportion with the refrigerant, so the compressor has to work harder at lower speeds (per unit of refrigerant shifted) and so consumes more energy. (Data from my own setup suggests that the compressor efficiency at 20Hz may be as much as 20% less than that at 50Hz, though I’ve had to make some big assumptions in this assertion.)
The Samsung tabulated figures are (I am told by Tech Helpline) based on data gathered from their laboratory test rig, and thus represent “ideal” conditions, presumably with calibrated and tested instruments. If your “measured” water deltaT is just 0.5degC in error - one or both of Tw1 or Tw2 are a little bit out - that alone could explain an error of 25% or more in your CoP calculation.
If these potential errors (20% and 25%) are both working against you, you could be actually achieving a CoP of 3.1 x 1.2 x 1.25 = 4.7 which isn’t a country mile from the number in Samsung table.
If you are really worried that your CoP isn’t what it says on the tin, I’d urge you to consider installing a heat meter and a CT - they may cost £500 or more installed unless you do it yourself, but it might put your mind at rest.
Apologies if my reply is slightly off thread topic.
I think your observations align with a slide from John Cantor’s presentation (at around 20+mins) showing varying COP vs Compressor Speed, with a sharp fall off in COP below 25-30% which are more pronounced at milder OATs (e.g, reading from the chart, at A7W35 a COP of 5.75 at 40% compressor speed falls to around 4 at 15% compressor speed). Alos, from that chart, would that indicate that compressor has a BEP with a compressor speed of around 40% or am I confusing different things?
Are you able to expand a little on compressor speeds, BEP and how a heat pump modulates it’s output please (go easy on me - I’m only a chemist!)?
I don’t have access to anything other than the Samsung controller to read data, but can see a percentage value for the inverter pump under Indoor Zone Option (currently at 25% whilst ticking over at near minimum output).
As I understand from your comments, input power varies with inverter pump frequency (but is not linear). How does a heat pump modulate it’s output? Does it simply reduce the inverter pump frequency or can it also modulate output whilst maintaining a constant frequency?
What is the minimum / maximum frequency range of the inverter? Is 50Hz the nominal max and the heat pump can then modulate down to maybe 12.5Hz or so? Clearly I can hear the compressor ramp up and down, most commonly during startup or a DHW run, although mine is only at full throttle during the mid-end of an unrestricted hot water run.
I’m interested to better understand the effect of things like Quiet mode and RFC (#5051) on efficiency (for DHW cycles). Presumably there are pros and cons if reducing output maintains a narrower dT (pro - more efficient transfer of heat) but is less efficient due to reduced compressor frequency negating some of the benefits. This also helps explain why I (and maybe @bravoleader80) may never quite achieve the Samsung published values for COP if my heat pump is not operating at it’s optimum compressor frequency the majority of the time (my heat pump is oversized so rarely operates at anything other than minimum output until OATs fall well below 5C)
I have the 5kw, by looking at my data, when fullpower, the compressor pushes 47/50Hz, During the last few days, when trying to stay within the LWT of 35 degrees, it was pushing 16/20Hz. What I’m getting is a very low COP of 3.5-ish and I don’t understand why and/or what I should do to work on it. I’ve tried using AI but it drove me crazy with weird stuff and I ended up nowhere.
Things that I’ve seen is that when it starts, max flow rate (22lpm), max frequency (47Hz), Max pump speed 700+RPM and Inv duty 100%
Then it starts slowing down to 9.8lpm, 16Hz, 45% Inv duty.
No Matter what I do with the lockshields, the system comes back to this equilibrium. Even going full open does not change much the results. I dont get it.
@Old_Scientist I have tried using Frequency Ratio Control (fsv 5051) for DHW heating. It is difficult to get totally comparable conditions ie OAT and starting DHW temperature, but for what it is worth, here are two DHW runs. With FRC on, it takes longer, so less time for space heating during the cheap Cosy period, COP was higher, but electricity consumption was slightly greater (1.61 vs 1.5 kWh). I have the same Samsung model as you.
@bravoleader80 I would endorse what Sarah said about the real values for COP being up to 1.5 times greater due to inaccurate Samsung figures. I have seen posts from Samsung users who are convinced that the true formula for COP is (consumption + production)/consumption. Although this formula is incorrect, they say that the COP that they get from their proper heat meter only matches the COP from Samsung figures if they use this formula. In other words, when their heat meter was showing a COP of 4.5, the Samsung COP using the correct formula would have been 3.5, but using the incorrect formula it came out as 4.5 which confirmed their belief in the incorrect formula!
So you either have incorrect Samsung figures and not a problem with efficiency or ……. it could be even worse - your real COP could be 2.3 (3.5/(1.2x1.25))!!! The only way to find out is to install a heat meter and electricity meter as Sarah suggests. However, if you are comfortable and also happy with your electricity consumption as judged by your electricity meter or smart meter if you have one, then there isn’t a problem,
I agree. It should be possible to have a sense as to whether the COP figures are reasonable, or not. If you know your heat loss, or alternatively how many kW of heat per day the property previously required, and you know the approximate energy consumed, you should be able to tell if your COP is 2 or 4. I seem to be lucky/fortunate that the data from the Samsung controller is, for me, reporting COPs that are in the right ball park, but more importantly I’m very happy with the amount of energy I’m consuming given the heat produced (versus what we were using before the heat pump was installed).
No as well as I’d like. We have three dependent variables (WL target, minimised refrigerant subcool leaving the condenser, and minimised refrigerant superheat leaving the evaporator) but only two independent variables (compressor inverter speed and main EEV %opening), and I don’t know how the controller algorithm deals with this. Out of interest, today’s NASAmonitor nicely shows speed reducing as LWT approaches WL target, but unfortunately it doesn’t pick up EEV position so it’s not easy to see which does what:
The “inverter pump” is your primary circulating pump, not the Outdoor Unit (confusing Samsung terminology ), and that percentage is the PWM signal which is displayed even if you have a fixed speed pump like me (64% atm):
Oh, I am confused now. maybe the word “production” is the source of my confusion. My Samsung Gen6 reports heat shifted from the source, not total heat production. If you calculate COP from this (Heat shifted from source/consumption) it will give you the wrong answer. Most of the consumption is used in the compressor and ends up as heat in the working fluid, so the total heat production is (heat shifted + consumption). Therefore, for a Samsung Gen6 at least, COP = (consumption + heat shifted)/consumption. I think @glyn.hudson found this quirk years ago?
This covers a similar timespan to the NASAmonitor output above, but starting at 08:00. I can see that by that time, two of the three dependent variables had been sorted out, and only WL approach (the dark blue line) needed to be fixed. I’ve shown the two independent variables with thicker lines. But I still can’t really infer much about their control of WL approach. Can you?
If you want the raw data behind this chart, here they are…
Where are you getting your “heat shifted from source” figure from, @toadhall?
If you are reading NASA registers, presumably 0x4426/7 are based on Q = volumetric water flow x water density x water specific heat x (LWT - RWT). Personally I always make this calculation from its components (rather than reading 0x4426/7) so that I can correctly temperature-compensate density and sp ht (and for all I know, the NASA registers may be based on pure water so may give an error for glycol-based circulating fluid like mine).
As for power consumed, there seems to be some confusion between 0x8411 and 0x8413. Some folk believe that 0x8411 is Outdoor Unit only and 0x8413 is Outdoor + Indoor Units combined (which would be more realistic as the latter would include pump(s) power). If you have your own CT, have you ever compared its output with 0x8411/3?
), and that percentage is the PWM signal which is displayed even if you have a fixed speed pump like me (64% atm):
