Let me first of all say that both methods work and produce results and the process has been highly useful in the need for participants to burrow deep into the inner bowels of a home heating system.
I don't doubt that your graphs and calculations have been correctly carried out based on the available data.
Since I don't like giving up on a problem that I feel can be solved with a little extra effort, I therefore had a closer look at the complete raw data, to see why there may be differences between the two methods.
I agreed to the comparison exercise knowing full well that there would be differences between the two methods, but hoping that the differences would point us in the correct direction so that the differences can be minimised to acceptable levels. I started by taking the 'minute' raw data and averaging or summing the various values to create hourly data, but rather than applying the 1.18 correction factor (something that has always concerned me) I kept the hourly totals to show just V x I. Comparing these V x I values with the spreadsheet values for the 12th to 13th Dec results, showed a variation of 856W over the full day, with the V x I coming in at 29935W and the spreadsheet value being 30791W. 856W over a 24 hour period equates to approximately 36Wh, which could be the hourly power used by a water pump. The 12th to 13th data is particularly good for comparison, because there is no setback applied, but also no DHW production, so the system was only supplying heating.
Even the most rudimentary measurement system should be able to measure voltage and current and produce a power value with greater accuracy than 18%, so why should it be necessary for such a large correction factor? This may also answer the question of the 'vampire' load.
Consider the following:- The manufacturer's data tables contain the results of test carried out at an international testing facility, which I would hope could measure all the variables with greater accuracy than 18%. I would assume that the 'power in' values shown within these tables would indicate the power used by the heat pump itself, and possibly one external water pump, which would be the minimum requirements for the system to operate.
So that now bodes the question of what additional equipment may be being powered through the heat pump dedicated power meter, such that the V x I value needs to be increased by 18% for the values to match?
Since the Power In (PI) value used within the spreadsheet is derived from the manufacturers data tables, it does not benefit from an 18% increase in its value, it would therefore be interesting to see a graph of PI x 1.18 plotted against the observed input power.
You are correct it is time to wind down the discussion until after the festive season, so I will save my remaining concerns until later.
I'm posting these charts for discussion, not as a conclusion. They have not been easy to produce. I wanted to get all the data on one chart, but it ended up far too busy, so I have divided it into two charts. What they show is the observed energy in (blue line) with a setback, and then the predicted energy in without a setback. I have also added the OAT (ambient temp) as it is the major determinant of energy in. Recall my observed minus expected methodology uses these two variables, and determines the energy saved (or not saved) as the cumulative difference between the two over a 24 hour period. In the upper chart, the expected (predicted) values come from @derek-m's model, in the lower chart, the expected (predicted) values come from my regression line (regression of observed energy in on OAT). In the upper chart, 'the information' (that which determines the values) comes from theory, in the lower, 'the information' comes from observations. All the usual caveats (small sample, one building etc etc) apply, as ever.
First, the things that make sense. The observed use (blue line) makes sense, goes inversely up and down with the changing OAT as expected. You can also see the recovery boosts immediately after the two setbacks (where the blue line drops to zero). Both the expected (predicted, red) lines follow the observed blue line fairly closely in the pre-setback period (R squared values are 0.84 for the model based prediction, 0.91 for the observations based predictions). The model based line is a bit spikier, the observations based line a bit smoother, but that is probably neither here not there.
The differences, and they are marked, occur during the setback days, when the model based predictions (upper charts) fall clearly below the predictions based on the observed data (lower chart). This is how and why the model predicts trivial or no savings, or even worse, extra cost, when using a setback, the (possible/likely) implication being that the heat pump has to play significant catch up during the all the time when it is running. But what if for some reason the model predictions for running without setback are artificially low? The model predictions will then wipe out any actual savings achieved by running with setback.
Equally, of course, it is possible the observations based predictions are artificially high. If that is the case, then any savings will be artificially inflated. The only thing we do know with reasonable certainty is is the actual energy in when running with setback: what we do not know is which (if either) of the predicted series is closest to what would have been the case if running without setback.
These charts cannot answer these questions on their own, a longer, ideally much longer, run of data is needed. But they do contain some intriguing hints from the setback days. First, the observations based predictions (lower chart) do follow the actual energy in closely when the pump is running, although they miss the recovery boost (which is OK, without a setback, there is no recovery boost). On the other hand, the model based predictions (upper chart) do seem low, all the more so whenever comparisons can be (very roughly) made between similar OATs. For example, in the period before the first setback, the OAT stays around 11 degrees, and the energy in is pretty stable at around 1 kWh per hour (and both the model and observations based predictions agree). However, in the period between the setbacks, the OAT drops to around 10 degrees, but the model predictions remain at or even slightly below the 11 degree OAT predictions, while the observations based predictions increase slightly (as expected). Even more curious, in the period after the second setback, the model predictions fall even lower (to their lowest point), despite the fact the average OAT during that period being around 11 degrees - as it was in the period just before the first setback.
As I said, and now repeat, this very short run on one system cannot provide definitive answers, but, at least for me, it does raise interesting questions.
Holidays over, back to the 'grindstone'.
Let me first of all correct one of your statements. The spreadsheet simulation is not based upon 'theory', it is based upon the manufacturer's data tables and the Laws of Physics and Thermodynamics.
Please correct me if I am misinterpreting the graphs, but the upper one appears to indicate that the Energy In would appear to be reduced, by NOT initiating a 6 hour setback.
The lower chart based upon your regression model appears to show little or no improvement either way.
Charts can indeed be quite revealing, but not as revealing as the underlying data.
I can’t compete with the detail, complexity and (I hope) the 40 plus page discussion of the previous analysis but this is my attempt at answering @cathoderay’s question; does setback save money without compromising comfort, with a really important addition, in my house, occupied by me. Before anyone asks, I am well aware what a rigorous, scientifically sound physics/engineering experiment looks like and it’s nothing at all like this 😆. Instead what we have here is an unashamed combination of anecdote and informed guesswork.
I have my whole house at c. 21 deg with a setback to 17 deg of 7-8 hours at night. Effectively the ASHP is off during the setback because the house almost never drops below 17 deg. I sometimes use Ecodan AA and sometimes WC. I do limit the LWT in AA to 43 deg so the recovery is quite gentle; similar to WC. The house heats up to 21 deg, always by 12 noon; sometimes sooner, whichever method I use.
My method of calculating savings is as follows: (1) Measure energy consumption actuals with a setback. (2) Estimate energy consumption without a setback. (3) Subtract (1) from (2). My setback always starts at 10pm and my recovery always ends at midday. So 10pm to 12 noon the next day is always the estimate; 12 noon to 10pm is always measured. I have a pseudo scientific way of calculating the estimates that is a little (but not much!) more sophisticated than eyeballing it.
These are the three days I looked at showing energy, LWT and OAT against time. The green line is (1), the red line is (2) and the blue one is OAT.
This table has the analysis.
Date
Mode
Setback
Actual kWh
Estimate kWh
Saved kWh
% setback
% saved
05/11/23
AA
10-5:30
17.52
21.59
4.07
31%
19%
26/11/23
WC
10-5
35.04
43.72
8.68
29%
20%
04/12/23
WC
10-6
29.30
37.92
8.62
33%
23%
In summary, I think setting back saves me about 20%. But remember that’s in my house, where I (and more importantly Mrs M) am warm enough.
We now have a few people doing some modelling and experiments on setback for an ASHP. IMO modelling a real ASHP in real conditions is quite difficult. I would like to see a model of a much simpler system as a starting point to understanding.
Can anyone explain how the energy required to heat air inside a hollow cube made of a homogeneous material of known mass, thermal conductivity and SHC would be affected by setbacks? If we all understood that we could start to add the complexities of the real world.
I feel like we're trying to figure out how a bumble bee flies without knowing how a simple glider works...
We now have a few people doing some modelling and experiments on setback for an ASHP. IMO modelling a real ASHP in real conditions is quite difficult. I would like to see a model of a much simpler system as a starting point to understanding.
Can anyone explain how the energy required to heat air inside a hollow cube made of a homogeneous material of known mass, thermal conductivity and SHC would be affected by setbacks? If we all understood that we could start to add the complexities of the real world.
I feel like we're trying to figure out how a bumble bee flies without knowing how a simple glider works...
You are quite correct that simulating a home heating system is a difficult problem to solve, particularly since there are so many different variables involved, many of which will insist upon 'varying'. 🙄
The original spreadsheet was developed to calculate the probable heat loss of a home at different OAT's, and then with reference to the available manufacturers data tables, to calculate the probable changes in LWT and COP as the OAT varied.
The spreadsheet has now evolved and been improved to include the effect of using different WC settings, the effect of a home's thermal mass, also the effect of changes in IAT, along with selection of fixed water pump speed or fixed DT setting. It can now even cope with the likely temperature changes due to having a PHE within the system.
As with any simulation, the accuracy of the results will be dependent upon the accuracy of the raw data used, otherwise it will be 'garbage in, garbage out'.
A further factor that may need to be assessed is how well the heating system has been designed and is actually operating, since the manufacturer's data will have been obtained from a system operating correctly. Nevertheless the simulator can help highlight problems with the raw data that may be due to the heating system design and/or operation.
The actual mathematical formula's used within the spreadsheet are not particularly complex, the complexity comes when selecting the appropriate data from the manufacturer's data tables, since it is necessary to not only select Energy In, Energy Out and COP values, at a particular OAT, but also at the relevant LWT, and the loading at which the heat pump is probably operating. Not quite Simples. ☹
We now have a few people doing some modelling and experiments on setback for an ASHP. IMO modelling a real ASHP in real conditions is quite difficult. I would like to see a model of a much simpler system as a starting point to understanding.
My post on the subject a few months back was a crude attempt at just such a model. Basically it considered the house as a single mass, leant heavily on the principle of conservation of energy (which is very useful indeed as a tool to circumvent having to think about detail, at the expense of 'smoothing over' some time variations in energy flows), and used some rough and ready approximations for how COP varies with flow temperature. For the parameters assumed it showed a modest saving at the mild temperatures we are currently experiencing. The method is explained in the first post and I subsequently refined it a bit.
Can anyone explain how the energy required to heat air inside a hollow cube made of a homogeneous material of known mass, thermal conductivity and SHC would be affected by setbacks? If we all understood that we could start to add the complexities of the real world.
I cant exactly (and I have a degree in physics), but if you google it you will get some hairy formulae. That's why I used a single mass model in the above.
In general terms I think its obvious the house (once its fairly warm) heats air-first and cools fabric-first, which is helpful to us humans in smoothing out perceived variations, but makes things more difficult for experimentalists, modellers and theoreticians. From an experimental point of view the key impact is that it will take the house a long while fully to catch up with both changes in ambient and changes in heat input, and in dynamic conditions its entirely possible it never reaches equilibrium. Experimentally my house (which I would guess has a fairly typical thermal mass, not dissimilar to some reported here) takes perhaps 2-3 days fully to stabilise. I have a feeling that modelling a house as two thermal masses (air and fabric) as opposed to one might be a very good approximation, but rather challenging I fear to implement in excel (although a few custom formulae might do the job and the basic physics is no more difficult than O level (I don't know about GCSE).
I dont think there is any doubt that answering the question generically is difficult/impossible because there are so many variations. Any model is only as good as the assumptions/theory/simplifications and any experiment suffers from the fact that the external conditions cant be controlled and thus doing the 'control experiment' (ie working out what would have happened without setback) is nigh on impossible. Thus many end up resorting to simulating the control experiment by ... modelling (of a different kind, but modelling nevertheless). Furthermore, as the results are likely to be dependent on the specific conditions, extrapolating it to others is dangerous and thus an experiment is arguably only really valid for the specific set up.
The best we can hope for IMHO therefore is probably some experimental data points from specific systems, plus some general guidance of how varying particular parameters are likely to change the behaviour, gleaned from models. That however would be a major step forward in itself.
I don't just spend my time manipulating data within spreadsheets, I also observe, monitor and even record some of the changes occurring within our home. I have found that with the correct control system and using the correct control philosophy, it is possible to keep our indoor air temperature, certainly in the central hallway, at a constant 21C +/- 0.1C, with the occasional deviation to +/- 0.2C when the OAT is changing more rapidly than normal. All of this without measuring and using an OAT value.
I agree with your assessment about the rate and flow of thermal energy, but I would say that when in heating cycle, the heat emitters are heated first, followed by the indoor air, followed by the inner fabric of the building, followed by the outer fabric of the building, and eventually the outside air. Hence the cause of global warming. 😋
When not in heating mode, the heat emitters cool first, thus helping to maintain the indoor air temperature and inner fabric temperature, but eventually the outer building fabric cools which in turn cools the inner fabric, which therefore cools the indoor air and any humans that may be inside.
As you have stated, to simulate all of that would be quite complex even for a fairly straightforward system. I think the best that can be achieved is to represent the home, and everything within, as a thermal mass, to which energy is added or subtracted, with the result being observed as a change in the indoor air temperature as the energy flows re-balance.
We now have a few people doing some modelling and experiments on setback for an ASHP. IMO modelling a real ASHP in real conditions is quite difficult. I would like to see a model of a much simpler system as a starting point to understanding.
My post on the subject a few months back was a crude attempt at just such a model. Basically it considered the house as a single mass, leant heavily on the principle of conservation of energy (which is very useful indeed as a tool to circumvent having to think about detail, at the expense of 'smoothing over' some time variations in energy flows), and used some rough and ready approximations for how COP varies with flow temperature. For the parameters assumed it showed a modest saving at the mild temperatures we are currently experiencing. The method is explained in the first post and I subsequently refined it a bit.
Can anyone explain how the energy required to heat air inside a hollow cube made of a homogeneous material of known mass, thermal conductivity and SHC would be affected by setbacks? If we all understood that we could start to add the complexities of the real world.
I cant exactly (and I have a degree in physics), but if you google it you will get some hairy formulae. That's why I used a single mass model in the above.
In general terms I think its obvious the house (once its fairly warm) heats air-first and cools fabric-first, which is helpful to us humans in smoothing out perceived variations, but makes things more difficult for experimentalists, modellers and theoreticians. From an experimental point of view the key impact is that it will take the house a long while fully to catch up with both changes in ambient and changes in heat input, and in dynamic conditions its entirely possible it never reaches equilibrium. Experimentally my house (which I would guess has a fairly typical thermal mass, not dissimilar to some reported here) takes perhaps 2-3 days fully to stabilise. I have a feeling that modelling a house as two thermal masses (air and fabric) as opposed to one might be a very good approximation, but rather challenging I fear to implement in excel (although a few custom formulae might do the job and the basic physics is no more difficult than O level (I don't know about GCSE).
I dont think there is any doubt that answering the question generically is difficult/impossible because there are so many variations. Any model is only as good as the assumptions/theory/simplifications and any experiment suffers from the fact that the external conditions cant be controlled and thus doing the 'control experiment' (ie working out what would have happened without setback) is nigh on impossible. Thus many end up resorting to simulating the control experiment by ... modelling (of a different kind, but modelling nevertheless). Furthermore, as the results are likely to be dependent on the specific conditions, extrapolating it to others is dangerous and thus an experiment is arguably only really valid for the specific set up.
The best we can hope for IMHO therefore is probably some experimental data points from specific systems, plus some general guidance of how varying particular parameters are likely to change the behaviour, gleaned from models. That however would be a major step forward in itself.
Neither can I and so do I! I'll have a look but I will give up very quickly if it looks too hard!
I think you are right in that the air/house are not always in equilibrium. I've never been in one but we were always told at school that it's toasty warm inside an igloo (where thermal mass and OAT are both less than 0 degrees C). Although I suppose the walls melt rather than warm up.
I also think that the factor that has most impact wrt setback in an ASHP system is how quickly you reheat, to the point that almost nothing else matters. I have some data from when I allowed the Ecodan AA to unleash most of its power to heat the house up and I think the savings I've mentioned here were all but wiped out if not worse. I'll post the data when I get a minute.
We now have a few people doing some modelling and experiments on setback for an ASHP. IMO modelling a real ASHP in real conditions is quite difficult. I would like to see a model of a much simpler system as a starting point to understanding.
Can anyone explain how the energy required to heat air inside a hollow cube made of a homogeneous material of known mass, thermal conductivity and SHC would be affected by setbacks? If we all understood that we could start to add the complexities of the real world.
I feel like we're trying to figure out how a bumble bee flies without knowing how a simple glider works...
You are quite correct that simulating a home heating system is a difficult problem to solve, particularly since there are so many different variables involved, many of which will insist upon 'varying'. 🙄
The original spreadsheet was developed to calculate the probable heat loss of a home at different OAT's, and then with reference to the available manufacturers data tables, to calculate the probable changes in LWT and COP as the OAT varied.
The spreadsheet has now evolved and been improved to include the effect of using different WC settings, the effect of a home's thermal mass, also the effect of changes in IAT, along with selection of fixed water pump speed or fixed DT setting. It can now even cope with the likely temperature changes due to having a PHE within the system.
As with any simulation, the accuracy of the results will be dependent upon the accuracy of the raw data used, otherwise it will be 'garbage in, garbage out'.
A further factor that may need to be assessed is how well the heating system has been designed and is actually operating, since the manufacturer's data will have been obtained from a system operating correctly. Nevertheless the simulator can help highlight problems with the raw data that may be due to the heating system design and/or operation.
The actual mathematical formula's used within the spreadsheet are not particularly complex, the complexity comes when selecting the appropriate data from the manufacturer's data tables, since it is necessary to not only select Energy In, Energy Out and COP values, at a particular OAT, but also at the relevant LWT, and the loading at which the heat pump is probably operating. Not quite Simples. ☹
I'll be interested in what the model makes of my data. I think my ASHP runs reasonably close to the data book; it's a simple set up and it's never over stressed. You should be able to tell as the energy delivered is there too.
