ReedRichards

The cylinder came from Telford. As best as I can tell my installer bought the cylinder and (LG) heat pump as a package through Telford (as you can now, although a different heat pump brand https://www.telford-group.com/product/clivet-air-source-heat-pump-packages/ ) and Telford included a booklet of instructions on how to set it up. These included some very unambitious settings for the weather compensation parameters which I changed subsequently. I am perfectly satisfied with my heat pump and it cannot be far off the promised SCOP of 3.2 for heating, although I have no way of knowing exactly. But electricity has trebled in price since I bought my heat pump so any measures I could take to improve the performance are worth considering (as indeed for any heating system at present). I replaced an oil boiler without expecting much change in heating costs and as oil has only doubled in price it's a bit galling that I am now paying more for heating than I would have if I had just sat on my hands.

Maybe we are talking at cross-purposes? I was referring specifically to Caleffi's 3-port buffer tank configuration in my last two posts. Also, my buffer tank is integrated with, and sits directly below, the DHW tank. It's short and squat, which I imagine are the worst dimensions to avoid stratification. Here is a picture. That's the heating flow on the left and return on the right. The input/output ports to/from the heat pump are round the corner at the same heights. I don't think this corresponds at all to the buffer tank configuration you have in mind.

Yes, I understand that @IanR. But if you want to minimise the potential mixing that goes on in the buffer tank or minimise mixing water from the heat pump with water from the buffer tank then you do need to match the flow rates. For example, if the flow rate in the heating system is, say, twice that from the heat pump then the system will draw half its water from the buffer tank and the buffer tank will be replenished entirely from water that has passed through the heating system. So initially, say, water at 40 C from the heat pump would mix with water at, say, 30 C in the buffer to make water at 35 C. If that loses, say, 5 C going though the heating system then it will be at 30 C coming back so the buffer remains at a constant 30 C. This is very probably an extreme and unrealistic example but it illustrate what, potentially, could "go wrong". Is the ideal to make the flow rate from the heat pump just a bit greater than that around the heating system so the buffer tank warms up gradually? It seems that way but maybe there is a flaw in my reasoning?

My heat pump seems to ramp up its power output until it reaches a plateau. Most of the time it then stays at this plateau level until something causes it to stop. This is certainly the return water getting too hot when it is heating up just the smaller zone in the morning. Or perhaps when both room thermostats are satisfied later in the day. The plateau levels it stops at appear to be quantised and to depend on the outside temperature. So when it is mild out it goes to power level 1, when it is around zero outside it goes straight to power level 3. This is the mode of operation with Weather Compensation but it seems similar to what you describe; the compression level and hence the input power is selected on the basis of the outside temperature. My Weather Compensation is not the fine-tuned type but something that gives me enough headroom to slowly raise the inside temperature during the day, as I have always liked to do.

I have a good deal of respect for Caleffi as I have previously read some of their other reports (and they make some useful plumbing parts that you can't get in the UK). I looked at Caleffi's 2 & 3 port configurations and what I don't know for sure is how the pump on the heating system manages to take the hot water as it arrives from the heat pump but not the cooler water in the buffer tank. It's got to depend on how well-matched the flow rates are between the heat pump pump and pump on the heating side. I'm not sure I how I would go about matching flow rates if I wanted to try.

Does anyone know how to achieve logic operations with mains voltage? When my heating is running I only (perhaps) need a volumiser, so I could close a valve to remove access to the buffer ports on the flow or return. I would only open the valve for a 4-port configuration when both of the two heating zones are closed. So that's A NOR function, isn't it? I guess that could be difficult. Alternatively I could open a valve to bypass the buffer with an OR function, in fact my internal pump works that way so maybe I would just need a connection to the bypass-valve in parallel with the pump? But that would rely on water preferring to bypass the buffer rather than go through it even though it could do either. I don't have the plumbing experience to know where water will want to go of its own accord.

I think Load Compensation might free the heat pump from the constraint to maintain a 5 C flow differential between flow and return and that might well be beneficial in real-world circumstances, although I'm not so sure about the test chamber; such a pity that Brendon squandered the opportunity to find out. My second pump is within the heated fabric of the building and only runs when the thermostat calls for heating so the issue with the running cost would be that it is £50+ a year at a COP of 1 rather than £16.66 at a COP of 3 (or whatever). I think it's probably running at 50 W on the medium setting. It won't run for as much as 20 hours a day but let's say that it does because that's 1 kWh = 34p per day during the heating season. So £50 per year is certainly of the right order of magnitude. I don't need mixing but I do need (I believe) a volume of warm water to defrost the heat pump, even when the valves to my radiators are closed.

Well I believe everything here is based on the Brendon Uys study and the inference some of us have drawn from it: The actual change was from a COP of 2.7 in the Test 2 configuration to 3.4 in the Test 3 configuration where the buffer was removed. Another change was that in Test 3 the heat pump was allowed to do Load Compensation (or something akin thereto) rather than just Weather Compensation. The "missing link" in the study would have been a test where only one of the two changes made between Test 2 and Test 3 was made but that didn't happen.

Although on/off at 55 C may not be quite what it seems. My heat pump appears to target a 5 C temperature difference between flow and return as its top priority. If the heating has been off for a while the water inside the system will have cooled and since most of this resides inside the heated fabric of the building, most of the cooling provided useful heat. Suppose it has cooled to 30 C. In my case the heat pump starts off at low power so gently raises the temperature. Although it is trying to get back to 55 C it may never get there if it's mild outside and the heating demand from the building is low and the room thermostat reaches temperature. So maybe the average water temperature is 40 C during a cycle and you get a "natural" weather compensation even when this is not engaged.

Mine doesn't. The manual states that a minimum water volume of 50 l is required and suggests alternatives of a volumiser, a low-loss header or a buffer tank (a 4-port buffer is illustrated). if you have a 4-port buffer then the input water to the heat pump has come straight out of the buffer. And either the water gets completely mixed in the buffer or a probe is only measuring the temperature of that part of the buffer in which it is positioned. Sorry about the confusion over the temperature data I reported. One thing I am thinking is that weather compensation alters the target temperature of the water leaving the heat pump. But perhaps superior weather compensation would reduce the target temperature differential across the heat pump or alternatively reduce the pump speed as the outside temperature increases.

I'm not measuring anything, I'm going by what the heat pump reports at A and K but I hope that A and B are at almost the same temperature, as should be L and K. The drawing of my heating system is accurate except insofar as "Manifold" is two 2-port valves, one for each zone. Points B, C, J & K are well-insulated and the insulation is fixed in place so measurements there would be difficult. I have tried to make measurements at E and F across radiators but I only have a single thermometer and the water temperature never holds constant so it's difficult. @PhilTMy pump is a UPM3 Flex 25-75 130 AZA. The pump itself is not noisy and makes a low humming sound but water in nearby pipes and radiators makes a rushing sound when water is flowing through them. I tried it on setting 1 for a while and that makes the system a little quieter but only a little.

Thanks @JamesPa. My internal pump is a Grundfos 3-speed pump. It's normally on setting 2 because setting 3 increases the system noise level and I don't like the noise; and because the installer set it up to be on setting 2.

Sorry, I should have mentioned that I have only radiators. Their nominal target drop across them is 5 C at the maximum rated flow temperature of 50 C but it's extremely difficult to check that they are well balanced because the output temperature from the heat pump is rarely constant because it is very rarely cold enough outside to require this. And the flow rates are so large that the flow is quite audible in those radiators immediately downstream of the pump (which is how I knew which zones were on). The audible flow is something I like least about my heat pump heating; the external oil boiler that preceded it gave me completely silent heating inside the house. Buffer or not, my heat pump seems to be trying to maintain a 5 degree differential between flow and return without knowing that it could do better using less power to maintain a smaller differential (assuming it is capable of using less power). My reading of the Brendon test was that Load Compensation was not invoked until part 3 of the test, although I'm not sure how much difference it would make in that case. My heat pump looks as if it has the potential to do better with Load Compensation although I don't know quite how that works with multiple zones. Reducing the temperature differential the heat pump aims for (by eliminating that in favour of a better means of control) would also have the effect of reducing any adverse effects of mixing. Even if there was a setting I could change to say: "aim for a 3 degree differential instead of 5" looks like it would be beneficial at the moment. I have not looked at @JamesPa's simulation yet; I'll do that now.

Here you are, @JamesPa. I have a 4-port buffer integrated with and underneath my hot water cylinder. My LG therma V 12 kW heat pump is set to do Weather Compensation but a Drayton Wiser thierd party controller overrides the LG one. I have two zones, one for most of the house and one for the 3 smallish bathrooms. The heat pump reports its output and input flow temperatures. I have personally insulated the pipes between the heat pump and the buffer using Kingspan phenolic insulation and that appears to be effective. It also makes measuring the temperature actually at the buffer inputs and outputs rather difficult, will you accept the heat pump temperature data? This only updates about every 30 seconds so I think some of the details about what happens are lost. At the moment it is warm out and my heat pump operates in 20-minute cycles. At the start of each cycle the inlet and outlet temperatures are typically around 35 C. When the room thermostat calls for heat the internal heating pump circulates water from the buffer for a approximately two minutes. The external pump then comes on. This runs for another two minutes before the compressor turns one during which time the inlet and outlet temperatures fall by a few degrees as the cooler water from the heating system reaches the heat pump. In the next stage the outlet temperature starts to increase, initially on its own. The outlet temperature lags about 5 degrees behind. This morning it got to 40 degrees out 34 in and the main zone cut out as warm enough. This left the smaller bathrooms zone which is required to be warmer at this time of day. This got to 45 in 40 out before the heat pump cut out even though the thermostat was still calling for heat and the pump was circulating water around that zone. The maximum power drawn was about 3.5 kW for a 12 kW heat pump. In principle the maximum output temperature should be 41 C at 5 C outside. A few minutes later and both zones now back on, both pumps running but compressor still off. Flow temperatures have dropped to 36, 36. now 37, 33 as compressor comes back on, then 37, 32, then 37, 33 main zone off, then 39, 35, then 40, 36 , then 41, 37, then 42, 38, then 43, 39, then 43, 40, then 44, 40 and compressor off. maximum power about 2.5 kW. Internal pump on bathrooms zone runs for a couple of minutes longer. Now no call for heat although the bathroom thermostat is 1 degree short of its target temperature (main zone at target). The second cycle, which I recorded completely, ran from 07:25 to 07:45. So at the moment the heat pump has the difficult task of maintaining the temperature in most of the house whilst trying to warm up three smallish bathrooms. But this is the real world.

I agree. It seems such a pity that the opportunity to do some good science by exploring other configurations was wasted. It looks to me is that there are two differences between tests 2 and 3; apart from the removal of the buffer the heat pump uses a different control method rather than using weather compensation. And obviously the volume of water in the system between test 2 and 3 is changed whereas it could easily have been maintained.

I think you might struggle to find someone who will install an air-to-air heat pump for you. If anything, they are more commonly found in commercial properties and the commercial installers don't necessarily want to do domestic work. But perhaps somebody here could make a recommendation?

Does it? I wonder if @JamesPacould model this as part of his weather compensation spreadsheet?

You are absolutely correct, @JamesPa, I was just thinking that myself and realising I had done this incorrectly. However the pump that circulates the water around my heating system operates at a fixed speed so I should be able to scale DT according to Delta T, I think. I will look at that and see if it makes a difference.

@JamesPa In the calculation I made to derive a weather compensation curve for my radiators, the temperature drop across the radiator was a parameter. I cannot see that as a listed parameter; is a particular value used?

I think @JohnMo is citing a scientific study that was discussed here recently (but which I can't find at the moment).

This is exactly the action of Load Compensation.

MCS calculations are building-specific and do not take into account the occupancy. Fill your house with enough people (or enough power-hungry appliances) and you don't need a heating system.

There are too many words for me to take it all in but I did a search on "weather" and got no results so I gather you don't explicitly say if you are using Weather Compensation or not. Failure to do so would reduce your COP in milder weather. What is the (maximum) output water flow for heating that your heat pump is targeting?

I think it probably is Load Compensation but Heat Pump vendors seem extremely reluctant to use the same terminology that is applied to gas boilers. If you are trying to maintain a steady state (and the response time constant is not too long, so radiators probably) then Load Compensation should compensate for inaccurate Weather Compensation settings. What does Homely do that a combination of Weather Compensation and Load Compensation does not?

I would say there are three ways you can configure a heat pump: Like an oil boiler with a fixed output water temperature and just an on/off control to maintain your room temperature (probably with a limit to the number of starts per hour to prevent short-cycling). This configuration results in the highest running cost. Like a (more sophisticated) gas boiler where Weather Compensation is used to reduce the flow temperature as the outside temperature gets warmer. You might also get Load Compensation but few heat pumps offer that feature and you have to use the manufacturer's own controller/room thermostat. Advanced ("hard core") where you abandon the use of thermostatic controls and tweak your Weather Compensation settings and radiator/UFH flow rates to keep your house at a constant temperature as the outside temperature varies. This could take a lot of tweaking to perfect and not all heat pump controllers allow you to make a night-time setback if you don't want the same internal temperature 24/7. But this should give you the most economical operation. There's another separate issue with how you ensure your heat pump operates with a minimum volume of water (as they require).