Thanks for your feed back. What else/ what topics do you think would be useful?

⁠The issue of maintaining a high flow rate for heat pump efficiency seems to come up a lot in discussion groups, whether that’s expressed as pros and cons of buffers, zoning, or TRVs.

If it's cycling more than 3 times/hours, you may need a small buffer tank

@@hvacdesignsolutions Hi. That's possible. I do have a Low Loss Header, and I have been experimenting with changing the power of the pump which re-circulates water from the Low Loss Header through the radiators. At current temperatures,(

I'm in the exact same situation. My WORST scenario in actual heating usage (calculatede on 6 years bills) is 4,5Kw, they quote a 8Kw heat pump that give 7,28 Kw @ -7/35 ... and my external project temperature is -5 so .... WHY???

Indeed, HP compressors 'soft start' these days, so wear & tear should not a worry

Thanks Adrian. I suggest your first enquiry about configuration should be with the manufacturer. They will have worked through buffer options, and how it relates to their particular controls and they will have sensor positions in mind. I have seen a few very bad buffer tanks where a lot of unwanted mixing happens inside them, so there are pitfalls to avoid. e.g. where to get 35C at your radiators, you need 40C at the heat pump. In this case, dropping the COP considerably. If it's only system volume for defrost that is the worry, then a small 2-pipe volumiser seems a good idea to me. I have never understood why manufacturers don't start defrost in the normal way, and if there isnt enough heat in the radiator circuit, it briefly diverts to DHW. This should 'steal' only a tiny bit of DHW heat, and only in certain conditions. Anyhow, each heat pump has different requirements, so you need to refer to the manufacturers recommendations.

Mmm.. well, whilst that Ecoforest graph is interesting, maybe it could also be a bit misleading here since there are other factors (like water temperature) that are probably more important. e.g. whilst a 'small' unit might run at high revs a lot mid winter, it should also spend a lot of hours at mid speed, and it could probably run at lower temperature (less rest time). A big unit is likely to 'stray' into higher flow temperatures at times, and may spend a lot of the year cycling. How it's operated also has a bearing on things. e.g. if its fairly constant, or if it has long-ish rest times. Finding the best efficiency relating to unit size (wrt heat demand) seems hard to assess, and no doubt its the extremes that are the worry, and a region in the middle with similar performance. I think you are barking up a similar tree. Maybe theory and practice don't quite match since heating a house is a dynamic thing... (day/night), and all buildings and occupants are a bit different. Not sure if that helps!

@@johncantor4056 thanks yes that is helpful - as you say it's not clear cut and I was sort of comparing it to very "1 dimensional" controls on a gas boiler without weather compensation so hadn't considered water temperature aside from the assumption that if an "average" property has an 'over-size' condensing gas boiler, it can be dialled down to reduce the overall output, provided the rads can emit the heat, then it can happily output constantly at its lowest modulation, but if like one of the examples you showed, where little else was done and a 'like-for-like' HP replaces a gas boiler, it would be better to go smaller instead, but would that have a detrimental impact on the efficiency? I guess inverting HPs have a wider working range and so as you say avoiding those extremes is the key. Plus if done as a retro-fit without other changes like bigger rads and more insulation, those things can be done afterwards and thus erring on the smaller side would still be a better bet than worrying about those few days where the 'calculated' heat losses at -3C make people think they need to 'go large' in order not to feel cold...on those days, a smaller HP could easily just run continuously in a more efficient manner rather than the on/off manner of heating that so many people are used to in this country!

Thanks for your comments.. I wish I could answer your question, but I dont actually know how the air-air control works. I would expect the system to register how many degrees the actual room temperature varies from the set-point , and respond accordingly... surely that is how it must work. i.e. as the room setpoint is approached, the compressor slows down, along with the fan speed. this gives the best comfoort. But does it also vary with respect to the outside temperature? I would expect it does. Given that a heat pump could give out twice the heat in mild weather compared to what it can give out in cold weather, then some sort of limitation for mild weather is surely part of the strategy. I would be interested to know.

Thanks, I think you’re right. Outside temp and temp differential (outside to set point) should be what the computer is most interested in. The heat pump must work harder if outside temp is cold and if your set point is a large jump up from current room temp; hence my theory of inching my thermostat up a couple degrees at a time as room gradually warms. Hopefully keeps compressor work light and air flow temp lower. My rough estimate is my heat pump electricity consumption goes up about 3% for every degree I turn up the thermostat and also goes up 3-4% for every degree the outside temperature goes down. Either way it makes sense as I’m increasing the differential from outside temp to thermostat set point. My electricity use was doubled with our recent week of 16 deg temp. House is old, could use better insulation, and Seattle homes not built for that. Thanks again for interesting discussion Steve

@@yt551217 In colder weather, the compressor will need to spin faster to get the same heat output, and in cold weather, more heat is needed, so surely it manages this intelligently in some way. Sounds like you are doing the right thing by increasing it slowly

Yes I agree that COP is sometimes used too much (given too much 'weight'). Its actual energy use that's important. But lets say you improve all journey types by correct tyre pressures, then all MPGs might improve by say 10% for all... But, skewed as you say because diesels are not so good at short journeys. Its a matter of testing in relevant condition.

@@johncantor4056I would say the cost of energy to achieve comfort is most important since prices of power can vary significantly

Hello. I'm wondering if this 'cycling' is actually defrosting. This can happen about as you describe, and will a lot in cold damp conditions. Actually worse when say 0C to +5C. Nothing to worry about. If you catch it stopping, you should see the white frost on the fins melt over a few minutes. Otherwise, in this weather, the unit is likely to run steadily. They generally vary (modulate) between 33% and 100%. When the heat demand is lower, (milder weather) it will need to cycle, and cycling on/off mid range can sometimes be better than running steadily at 33%. difference between the two is often slight.

I think I am getting at point that its generally better to keep the heat pump running in cold weather, so you dont really want the room thermosts turning it off... e.g. lets say it's running very happpily and steadily in cold weather (though might be interrupted by defrosts every say 40 mins). Anyhow, if the thermost turns it off, then there can be some catching up to do which may not be easy. Ideally the compressor will modulate down and keep running. The manufacturers thermosts is usually best, and this is where 'room target' or 'auto adapt' can work very well. In your case, I guess you are trying to vary/reduce the water temperure to see if say 40C is enough etc. Maybe your smart stats are trying to be too helpfull! can you simply turn the TRVs to max for your experiments?

Yes, its not clear in Ecodan info, but I think the 10-60 is related to time response of room sensor. I guess some 1/2hr rests should result in some better COP due to cold start, BUT it will 'rev up'. Shame there is not an 'economy' setting, maybe watching outside temperature, that would stop it going to maximum speed so eagerly.

@@zlmdragon. I have never quite understood this setting. There seem to be other factors at play. Of course the actual compressor speed is all important, and I found the 6kW 'revs up' unnecessarily even when its quite mild out. I expect some 'rev ups' are there to help compressor lubrication, but I still find it straying to high output, unlike the older models. I guess there are 2 main considerations a) ensure that enough heat is being delivered, b) minimise running cost. It seems to be programmed mainly for 'a'

You are right that COP at start (after a long rest) is poor as it warms everything up. Compressor would be warm though if starting when cycling is frequent. There is however another thing at play... at start-up, the compressor pressures are equal, so the compressor has a very 'easy ride' at first, until the pressure difference across the compressor builds up, so at start, if it were not for various inefficienies, the COP would be much higher at start up

Hello Roman. You can look at my calculator at heatpumps.co.uk/calculators/ to help assess buffer sizes. It might be useful. It's not that straightforward because lets say your 9kW Daikin actually only drops to say 4kW in milder condition, and all is in steady-state and happy, then if you need to cycle, as the radiators rise above the temperature you need, then unfortunately, it wont re-start at 4kW again. Instead, it will 'rev-up' a bit, possibly to 9kW for a short while before it slows down again. (I actually don't have much experience with Daikin) So, to be clear, the buffer volume will slow down the rate of water temperure rise, so it should help, but there are more than one way to configure a buffer tank, of which small details like the sensor positions is important. You are adding complexity. I suggest you go with Daikin's recommendations because they will have developed their controls to work in a certain way.

Unless you have ground heating or a large / complex installation, it's likely that you don't actually need any buffer tanks or secondary pumps. Find out whether your installation meets the water volume requirement for your HP on its own, without the 25 litre buffer tank... if it does, get rid of it and get rid of that secondary pump too. Your COP is likely to improve.

Yes, heat meters too expensive for most people. I think a lot can be gleaned from sensors on pipes. The heat (kW) is proportional to flow-return dt. IF flowrate constant. e.g. if you look at my graph at 2:40 in, and measure (with a rule) the dt at 27C point and 29.5C point, I make it 16mm and 15.5mm.. so the dt has reduced a tiny bit, but input power has increased, so COP a bit worse. This is just an example of what you might be able to glean. However, if water flow changes, this is meaningless, so you need to do specific experiments, and repeat a few times.

@@johncantor4056 Thanks John I will try and put something together. Its a shame that heat meters are so expensive.

@@timcx I’ve got a hitachi yutaki I’m trying to get to grips with in terms of flows and cycling. It doesn’t have good data output either, I’ve recently found home assistant and Shelly iot devices. You need a bit of IT skills but you can get a lot of info from Shelly em and some temp sensors.

I'm doing exactly this- flow, return and a current clamp on the HP power. Very useful to set up the system indeed.

@@simoncanfer5030 Hi Simon how are you collecting the temperatures, what equipment are you using? I was thinking of using EmonTx but open to other ideas.

There are too many dts! If the flow-return dt is large, then the condensing temperure is likely to be closer to the flow temperure, BUT the system is worse off because the mean temperature to the heating system is worse... I think this needs spelling out with a sketch... thanks for highlighting it. The COP estimator is fairly simple. I tried to make a version to include flow-return dts, but it became unnecessarily complicated!

​@@johncantor4056 Yeah, with lower flow rate dt we can get higher medium temperature and it means that we can use lower LWT. But on other side if lets say we apply dt = 5 for cold weather, than for moderate weather dt can be 2 or even 3 and heat pump can start cycling. I am just playing with configs of my heating system and trying to get the best SCOP. My heat pump can't work constant with dt < 3.5(that's for moderate weather that I have 90% of time). In this case LWT starts to raise as heat pump starts to increase frequency. And it can just work with higher LWT(1-2 higher than needed) or starts to cycling. Can't find any info if it happens due to algo on my heat pump, or it's physical limitation of process that happens inside condenser. Maybe if dt is too low, gas can't transfer to liquid on 100% and that's why heat pump starts to increase output (some kind of protection). For now seems I need to make dt 4-5 for moderate weather and 8-10 for really cold weather. Or make a variable flow rate.

Maybe it's also the limitation of compressor speed that you mentioned in the video. I see that it avoids to work with frequency less than 35-40% and power consumption less than 20% of max.

The estimator assumes a certain flow temperature, and if the flow-return dt changes, (assuming same heat (kW), then the flowrate must be changing, and return temperature is changing too. An alternative would have been to give a 'mean' flow-return temperature, and this might be more relevant to water-to-refrigerant dt. I somehow need to spell this out better because as it stands, it looks like you can change the 'offset' and change the COP, however, the 'offset' is more a function of heat-exchanger size/compressor speed than anything else. I am pondering it.

Its a bit complicated since many systems have a fixed flow circ pump, so typically, if flow-return dt is 5 at high power, then it will naturally become 2.5 when the system is 1/2 power. Other systems simply 'track' the flow and return temperatures, and they adjust the circ pump automatically so the flow-return dt stays at 5 degrees (kelvin/Celsius). Furthermore, some systems might assume a fixed flow, and the control uses the dt value as control in some way. That might seem all a bit daunting, but I dont think fine-tuning of dt will actually affect the COP much. However, if the dt is 6 or more then the return is getting cold, so the flow temperure may need to be warmer to compensate, so for that reason, a large dt is rarely desirable.

Hello Anthony.. i look forward to watching your video through ( no time right now!!).. One initial question.. how accurate do you think your heat assessment is when dt is less than 2? A small sensor error will make a big difference. with no heat, bit pump running, what dt is showing on the heat meter? Is it zero? I will watch your video soon,

@@johncantor4056 Hello John I find the flow sensor minimum reading is 6.5l/min (which is also what the manual says) and the next lowest reading is 7.1l/min so that’s just over 10% error. Likewise when it comes to the primary current reading for the compressor the lowest readings are 0.5, 1.3, 1.8, 2.3 Amps etc… There’s absolutely no variation in between those discrete readings. What this means is very choppy current and flowrates and therefore very choppy instantaneous coefficients of performance. However when you average it out over the course of 10 minutes then you get some quite revealing trends happening, I tend to use 10 minute averages to determine what’s going on with my heat pump. DeltaT temperatures are accurate to within 0.1 degrees on both the flow and return temperatures. I am able to determine the accuracy of my calculated power consumption by sampling the readings against the power meter feeding the heat pump and the averages are very close but instantaneous readings can be off by quite a bit.