I find John Cantor’s video most useful for explaining efficient running of a heat pump.
He stresses that cycling is not necessarily a bad thing and gives examples of what is good and what is not. The thing to ensure is that all TVRs are keep open while the system is set to run as otherwise the pump will have to adjust its settings as it will also for changes in outside temperature. It is never possible to cover the whole year continuously. Some heat pumps will modulate better than others but some like Midea cut out and when they restart they make adjustments to increase or decrease output. But it is my experience that the Midea is less efficient (lower COP) at lower output. Now I just calculate an instantaneous COP manually, but my conclusions are equally valid. Now Midea may have decided that it is more efficient to stop for a while rather than continue less efficiently.
It seems that the separation of primary and secondary circuits is problematic. Low Loss Headers and Plate Heat Exchanges need to be setup properly as well as matched to the system. These can be a source of cycling. It seems to me that a few degrees C drop from primary to secondary is about the limit. Anything more is indicative of a problem. My installer has agree to work with me to improve mine. He was very interested to see the curves I have posted here of the flow and return each side of the PHE. It certainly will be quick to tell what works and what doesn’t. So we will reverse the pump temporarily (swapping flow and return on one side is better), and also he will bring a less powerful one with more speed variation. It is not clear whether one secondary pump speed is ok as the primary is variable. Being critical of my installer would not have been a good idea. I think you have to dangle a carrot rather than use to whip! Maybe there is a market for a portable monitor. It didn’t take me long to fit the HA kit I built.
My main goal is to cope better with -9c outside next winter without incurring a massive electric bill. I will be disappointed if I need the LWT higher than 42C assuming the PHE and pump is sorted out. As to whether I go with WC, that will depend on lots of things. It seems to me that I don’t like the bedroom heated at night but I don’t want it to get too cold! One solution is to reduce the LWT enough to maintain a lower temperature. This is the opposite of WC though! If Homely worked that might be answer.
@filipe Im not sure that there was much 'good' cycling, just some where the difference was negligible. The key messages for me were
1. Cycling means that the average temperature seen by the emitters is less than average temp seen by the HP, and since the HP is less efficient at high temperatures this is not good, abeit that the difference, in come circumstances, is small. This is entirely logical and a direct result of the Carnot efficiency, ie the basic physics.
2. Oversizing of the heat pump, resulting (unless it can modulate deeply) in cycling for much of the year is bad (because its means that (1) is happening all year)
All this makes sense and shows why right-sizing of heat pumps is so important.
Unfortunately the evidence is that heat pumps are all too frequently oversized, quite often substantially. This does of course protect the installer from service calls, but may not be good for the customer. The more I think about it the less I understand why whole house sizing is not checked against experimental measurements. There is somuch uncertainty in the theoretical calculations (because of the uncertainty in the fabric, particularly if you ignore what the customer tells you about the construction) and the sizing is so important that it feels almost delinquent not to do this, generally easy, sense check.
Unfortunately the evidence is that heat pumps are all too frequently oversized.
Unfortunately it is next to impossible to modulate efficiently over the entire range of outputs. However, I agree that unnecessary oversizing is not good practice. Most of the problems come from not getting the emitters properly matched to the rooms and the HP. Some home owners will also not like radiators operating at 30C and will go back to gas if they don’t feel warm enough. There also comes a point when it doesn’t really matter if there is some cycling as long as it doesn’t happen at times of high demand for heat. My system only runs for a few hours a day now before the Solar gain clicks in.
I’m sure the Octopus approach of targeting standard modern houses with systems that through experience become designed correctly will have a big impact on the transition to renewable technologies. It doesn’t help the older housing stock, but maybe in the big picture that doesn’t matter and price pressures will ease and those houses stay on fossil fuel.
There also comes a point when it doesn’t really matter if there is some cycling as long as it doesn’t happen at times of high demand for heat. My system only runs for a few hours a day now before the Solar gain clicks in.
Agreed. Without doing all the sums my guess is that the critical temperatures are 4-8 (perhaps 12)C where the energy demand is high and is a temperature at which our climate, at least in Southern England, spends a considerable amount of time. If you have designed for, say -2, it doesn't take much over-specification to get into the realm where the HP wont modulate down far enough.
In my own case (which I admit I quote frequently, but I think it may be not unusual), two MCS surveyors got to 16kW, my calculations (based on MCS assumptions but adjusting for fabric upgrades) get to 10.5kW, and the actual measured demand at -2 is 7.5kW. Most of the time the demand sits just over 4kW. So my intention is to fit an 11kW (most likely Mitsubishi) pump, but even that is marginal at 4kW. Had the MCS brigade got their way with their 16kW+ offerings, I would be well into cycling for much of the heating season which is really quite atrocious.
I’m sure the Octopus approach of targeting standard modern houses with systems that through experience become designed correctly will have a big impact on the transition to renewable technologies. It doesn’t help the older housing stock, but maybe in the big picture that doesn’t matter and price pressures will ease and those houses stay on fossil fuel.
I completely agree that this makes sense in the start up phase, but it wont do long term if the older houses remain on fossil fuel. Here is a plot of the age of UK housing stock, ~70% pre 1980 and its a fair bet that many of these have had either fabric upgrades which are not readily visible, or more recent extensions built to a better standard. Thus making the assumption that the whole house is built to the original standards (which both the MCS surveyors who did a full survey of my house did, even though I told them about the fabric improvements) leads to significant oversizing.
it is next to impossible to modulate efficiently over the entire range of outputs
Why?
Please expand on this.
Yes I understand that an (AC mains) electric pump requires a certain amount of electricity to get it started and then keep it turning.
I am told (not verified) that the limiting factor is the compressor. Compressors are leaky so turning them down too far (ie going to slowly) results in significant losses. This makes sense, although its not fundamental physics like the Carnot equation so can probably be overcome. @filipe may have a better explanation.
This post was modified 2 years ago 5 times by JamesPa
4kW peak of solar PV since 2011; EV and a 1930s house which has been partially renovated to improve its efficiency. 7kW Vaillant heat pump.
it is next to impossible to modulate efficiently over the entire range of outputs
Why?
Please expand on this.
Yes I understand that an (AC mains) electric pump requires a certain amount of electricity to get it started and then keep it turning.
And I appreciate the difference between;
a standard mains pump (Grundfoss 15/60) being supplied with an AC waveform that's chopped using MOSFETs/Triacs
an AC pump which accepts a logic-level PWM signal to control its speed (Alpha1 or UPM3)
But if there's a new type of device required, then what is the problem which it needs to overcome?
First of all the entire operating range of required output from a heat pump, can go from no heating required at all, to possibly over 100% of maximum output in adverse weather conditions. So no system can modulate over the full operating range.
I'm not certain if the actual compressor has a minimum and maximum operating speed, but the electric motor driving the compressor will certainly have limitations. If the speed of the compressor motor is being varied by changes in frequency, then I think the lower limit for an induction motor is 10hz, or 20% if 50hz is 100%. This is more to do with preventing the motor from overheating rather than anything else, since the normal cooling fan attached to the motor shaft operates at the speed of the motor.
Industrial systems can operate electric motors down to 5hz, but require an external cooling fan to be fitted to the motor. The cooling fan is supplied from a separate electrical supply at 50hz, so operates at full speed whatever the speed of the main motor.
I do believe that there are some present heat pumps which control the speed of the compressor, the speed of the internal water pump and also the speed of the heat pump fan, so should offer the greatest continuous operating range.
In an ideal World maybe a heat pump should contain two compressors, a small one for 33% duty and a large one for 67% duty. During milder weather conditions only the small compressor would be required and should offer quite a degree of turn down, as well as improved efficiency. When the heat pump loading gets to the point where the small compressor is operating at say 80% capacity, the small compressor is stopped and the larger compressor is started. During adverse weather conditions, if the larger compressor cannot meet demand, the smaller compressor should also run. Such an arrangement could also help alleviate the undersize or oversize problem that can occur with heat pump selection.
For the time being, to help overcome the possibility of too frequent cycling during milder weather, the trusted room thermostat should be utilised to control the starting and stopping of the heat pump.
I think the lower limit for an induction motor is 10hz, or 20% if 50hz is 100%
Ah... that's not how I thought a (domestic-sized) 240v AC pump was operated using PWM.
I expected the basic 50Hz to still be applied, but that it would be 'chopped' into on/off sections like this
Coincidentally I was asked to 'look at' a faulty washing machine for someone else a few days ago.
The 240v mains was fed through a rectifier before it went anywhere else in the machine. It hadn't occurred to me that modern washing machines were no longer using motors/pumps which were driven from the 50Hz mains.
I think the lower limit for an induction motor is 10hz, or 20% if 50hz is 100%
Ah... that's not how I thought a (domestic-sized) 240v AC pump was operated using PWM.
I expected the basic 50Hz to still be applied, but that it would be 'chopped' into on/off sections like this
Coincidentally I was asked to 'look at' a faulty washing machine for someone else a few days ago.
The 240v mains was fed through a rectifier before it went anywhere else in the machine. It hadn't occurred to me that modern washing machines were no longer using motors/pumps which were driven from the 50Hz mains.
To be quite honest, without looking inside a heat pump unit to see what motor is actually fitted, I can only assume that it will be an AC induction motor.
I don't know what control is actually used on water pumps, again without looking inside, but the original phase angle control is not as per your diagram. During each half cycle, the switch on point can be delayed, but once switched on, it remains on until the zero crossing point is reached. I do believe one of the problems with phase angle control is it can induce harmonics back into the electricity supply.
I think the lower limit for an induction motor is 10hz, or 20% if 50hz is 100%
Ah... that's not how I thought a (domestic-sized) 240v AC pump was operated using PWM.
I expected the basic 50Hz to still be applied, but that it would be 'chopped' into on/off sections like this
Coincidentally I was asked to 'look at' a faulty washing machine for someone else a few days ago.
The 240v mains was fed through a rectifier before it went anywhere else in the machine. It hadn't occurred to me that modern washing machines were no longer using motors/pumps which were driven from the 50Hz mains.
I have just found the info linked below, which would seem to indicate that PWM uses Zero Crossing control rather than Phase Angle.
I do believe one of the problems with phase angle control is it can induce harmonics back into the electricity supply.
Ah... well that seems to prove the case.
In discussions with senior engineering managers at NGED, they have cited harmonics from heat-pump installations as one of the two significant factors which contribute to losses.
Whereas phase-imbalance only creates losses at the local sub-station, harmonics propagate through the network and are detected at the higher-voltage transformers.
I've just plotted this from voltage measurements on a 33kV line (Bulk Supply Point transformer output):
The graph shows the voltage imbalance (plus or minus) as a percentage over a 20hour period. The area beneath the red line shows the amount of energy which gets lost as heat.
So that's a negative outcome from the way we control motor speed on heat-pumps. ☹️
Impact of the heat pump performance The heating COP of inverter driven heat pumps is mainly affected by the primary inverter losses. With state of the art PWM inverters an average COP reduction between 8% and 15% can be expected. This disadvantage however is compensated by a better adaption of the heating capacity to the system needs. Practical experiences show that an overall improvement in seasonal performance factor {SPF) of 17% and more can be achieved due to reduced storage and cycling losses [Henderson 90]. CONCLUSIONS The inverter efficiency of commonly used pulse width modulation inverters can be calculated with a simple model using three empirical constants. Experiments show that the efficiency varies between 80 and 96 %, where the maximum occurs at high speed and load. Secondary inverter losses in the induction motor can also be calculated by considering the skin effect and the distortion currents as main dissipation sources. For inverter switching frequencies higher than 1 kHz these additional losses become negligible {<1 %). It is crucial to adjust the inverter parameters precisely, such as the voltage/frequency ratio to the motor characteristics. The reduced heat pump COP is compensated by a better system efficiency. Altogether an improvement in SPF of 17% and more can be achieved with inverter driven heat pumps.
My observation is that part of what makes a HP having a lower COP at lower output, is mathematics.
Example: A HP that worked for an hour in one day has a different COP for the day, if you take the whole day power consumption vs just the hour that it worked on.
For a duration of time a HP+fixed speed water pump will have a fixed amount of energy used(taking out the actual energy used by the compressor motor).
The water pump, circuit boards, inverter board, plus the afferent energy losses of the electrical and thermal parts of the HP(outdoor), all of these added up being more or less constant, will be a percentage of the total energy used, which in the equation of the COP it will drive it up and down depending on the energy output amount.
For the rest of the difference, I'm not looking to search and understand why, but an experiment would be to have a water to water HP, with equal heat exchangers/flow and notice the energy used vs energy moved from one tank to another at different power levels......
For who wants to see some data, in these files are part load data of vaillant HP, toshiba AC and toshiba split A/W HP which is derived/having the external unit from the AC product.
It looks like at some temperatures, the highest COP it's in the middle range, but the percentage difference differs depending on the LWT from 35 to 55/65.
In general I've observed AC datasheets not having this conundrum, with the highest COP at lowest power output, but some do have the same issue, with COP best in midrange.
For the time being, to help overcome the possibility of too frequent cycling during milder weather, the trusted room thermostat should be utilised to control the starting and stopping of the heat pump.
Its not so much frequency that causes the loss in efficiency, its the fact that, to deliver the same energy from the emitters, the heat pump must operate at a higher temperature if operating intermittently than is required if operating continuously. So Carnot inevitably gets you - its unavoidable so far as I can see and is dependent, to first order, on mark-space ratio.
Actually there may be an argument for faster on-off switching rather than slower (same mark-space, higher frequency) because then the temperature variations will be less and so the increase in flow temperature required to deliver any given amount of heat also less. In the limit a heat source that modulates on and off once per minute (say) with a 50% mark-space ratio is indistinguishable, as far as the rest of the system is concerned, with one operating at half the power, so the required elevation in flow temp versus continuous operation would be negligible. Obviously, at this point, wear and tear and start up 'costs' take over' but I need to think on this one a bit more!
This post was modified 2 years ago 5 times by JamesPa
4kW peak of solar PV since 2011; EV and a 1930s house which has been partially renovated to improve its efficiency. 7kW Vaillant heat pump.
Thinking about installing a heat pump but unsure where to start? Already have one but it’s not performing as expected? Or are you locked in a frustrating dispute with an installer or manufacturer? We’re here to help.
