It looks like this forum has been the victim of dumbing down, a little way upthread you were prepared to accept an answer based on A level maths, now its O level.
I think you'll find the dumbing down comes not from my perfectly reasonable adjustment of the maths level for a simpler problem, but rather from the willingness of certain participants to jump in with snide remarks whenever they can. It does no one (me included when I do it) any credit, and is out of place on what is normally a very civilised forum. And don't forget, I can set the level at whatever level I want, for a comment that I write. Next time it might be nursery school maths.
A 8 degree F change in temperature is equivalent to a 4.44C temperature change, which for 8 hours would have the same affect as a 1.48C temperature change for 24 hours.
I can show the maths if you wish.
Even I can do that maths. But what neither of you appear to have picked up on, or if you have, commented on, is that this is a setback (and quite a sharp one at that, it is probably an on/off command given it is only for 8 hours). It complicated by the fact that it is in fact an 8 degree F change on the thermostat which is not the same as the 8 degree change in (room) temperature that you use, but nonetheless both of you seem content that savings of 10% are the order of the day for a 1 degree C thermostat/IAT drop.
Now as it happens my setback is similar in effect, ie the setback is large enough to mean the heat pump is usually off for the setback duration. I now have just about enough recent data to be able to get meaningful IAT for both during the setback, and after. Here is the period from the start of the setback to date for visual reference. Note that apart from the extreme events, the whole period has broadly similar OATs, with OATs hovering around something a little under ten degrees:
The average IAT during the stable part of the setback (6 - 24 Nov) was 18.3 degrees. The average IAT for the period 6 Dec to date, ie no setback, but excluding the early December cold snap, was 19.3 degrees. Using @derek-m's logic (which I agree with, in fact I was already applying it) my setback, very similar in effect to the US example (both sharp setbacks that in effect become on/off setbacks), means that my setback is equivalent to a 24 hour thermostat lowering of 1 degree.
Using @jamespa's simple maths, and OAT of around 8.5 degrees (eyeballed from the above chart, excluding extreme periods), then my savings will be
1/(18.8-8.5) = 9.6%
or rounded up, 10%. OK it is not the same as the 20% suggested by my recent assessment, but it is still a worthwhile saving.
This naturally leads, at least for me, to why isn't this a good enough assessment of setback savings? Given interval setbacks have a 24 hour equivalent (see @derek-m's comment above), take periods with broadly similar OATs (the more similar the better) and compare the average IATs with and without setback, and then apply the 10% saving for every 1 degree drop guideline. Simple, logical, and comprehensible, and you have a good enough answer for every day use.
Midea 14kW (for now...) ASHP heating both building and DHW
Like you, I prefer to keep the maths as simple as possible, so to take the US example and apply it to your situation.
If a temperature drop of 4.44C for an 8 hour period would probably produce a 10% reduction in energy supplied over a 24 hour period, then for you to achieve a similar 10% energy saving, your IAT would need to fall by 5.9C for the full 6 hour setback period.
The average IAT during the stable part of the setback (6 - 24 Nov) was 18.3 degrees. The average IAT for the period 6 Dec to date, ie no setback, but excluding the early December cold snap, was 19.3 degrees. Using @derek-m's logic (which I agree with, in fact I was already applying it) my setback, very similar in effect to the US example (both sharp setbacks that in effect become on/off setbacks), means that my setback is equivalent to a 24 hour thermostat lowering of 1 degree.
Using @jamespa's simple maths, and OAT of around 8.5 degrees (eyeballed from the above chart, excluding extreme periods), then my savings will be
1/(18.8-8.5) = 9.6%
or rounded up, 10%. OK it is not the same as the 20% suggested by my recent assessment, but it is still a worthwhile saving.
If you work out the 24 hour average IAT and do the simple math correctly you will indeed get a very crude estimate, one which is perhaps good enough for a fossil fuel system.
However it's definitely ( in many if not most cases) an over estimate, by quite a large margin if the drop is small, because of the things it doesn't take into account listed above. With a heat pump, typically having a capacity more finely tuned to the demand, the recovery strategy could easily negate the (energy) benefit and its absolutely certain that it does so with a house that has high thermal mass (unless the heat pump is poorly matched to the load or there is a high vampire load). Note that the previous sentence relates to energy consumed only not to charges where there are tou variations.
So, whilst in the case of a 24x7 reduction in thermostat setting, its a safe assumption that there is a saving even though the rough and ready calculation ignores several factors, in the case of a part time set back with a heat pump this is no longer the case. In this case, unlike a 24x7 reduction, the crude estimate is too crude, and the potential saving too small, even to guarantee it has the correct sign. To be good enough the other material factors must be accounted for, which is where the real world becomes more complex even though the physics is still very simple.
As I've said many times my gut feel based on all the figures, experimental, from modelling and from simple theory (mostly the last two, experimental results are yet to add much in reality) is that there are modest savings, order 5% (certainly nothing like 20%) available from setback in many cases. But I haven't seen enough yet to convince me that this is sufficiently generally applicable for it to be reasonable 'advice' to others, nor to put boundaries around the circumstances where the conclusions are safe. Others may of course be more gung ho, and @derek-m ,who is more familiar with the modelling, may be able to assert boundaries (albeit not tested by experiment so possibly of little interest to those who dont trust models).
Ps I've just seen @derek-m s post which crossed with mine. For the avoidance of doubt nothing I have written above is disputing his figures, which take into account the factors which the crudest calculation possible omits.
This post was modified 1 year ago 4 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.
A 8 degree F change in temperature is equivalent to a 4.44C temperature change, which for 8 hours would have the same affect as a 1.48C temperature change for 24 hours.
I can show the maths if you wish.
As James has shown, this would probably provide an approximately 10% reduction if heat loss from a building, whatever the heating system. A heat pump based system should actually provide an even greater saving, because of the improved efficiency of it not working so hard at higher LWT.
ie 4.44 drop over 8 hours is the same as 1.48 over 24 hours, which makes sense and is, according to you, legit (you just did it), an interval setback can be converted to a 24 hour equivalent (and presumably vice versa). In my system the 24 hour change was 1 degree less during setback, by the 10% saving per degree drop guideline, that equals a 10% saving.
Midea 14kW (for now...) ASHP heating both building and DHW
A 8 degree F change in temperature is equivalent to a 4.44C temperature change, which for 8 hours would have the same affect as a 1.48C temperature change for 24 hours.
I can show the maths if you wish.
As James has shown, this would probably provide an approximately 10% reduction if heat loss from a building, whatever the heating system. A heat pump based system should actually provide an even greater saving, because of the improved efficiency of it not working so hard at higher LWT.
ie 4.44 drop over 8 hours is the same as 1.48 over 24 hours, which makes sense and is, according to you, legit (you just did it), an interval setback can be converted to a 24 hour equivalent (and presumably vice versa). In my system the 24 hour change was 1 degree less during setback, by the 10% saving per degree drop guideline, that equals a 10% saving.
If the IAT fell from your desired 19C, to 13.1C and remained at that value for the full 6 hour setback period, then increased back to 19C, then I fully agree that you achieved a 10% reduction in energy consumption.
A 8 degree F change in temperature is equivalent to a 4.44C temperature change, which for 8 hours would have the same affect as a 1.48C temperature change for 24 hours.
I can show the maths if you wish.
As James has shown, this would probably provide an approximately 10% reduction if heat loss from a building, whatever the heating system. A heat pump based system should actually provide an even greater saving, because of the improved efficiency of it not working so hard at higher LWT.
In my system the 24 hour change was 1 degree less during setback, by the 10% saving per degree drop guideline, that equals a 10% saving.
As i say above not necessarily in your house which appears to have a high thermal mass and an unexplained high standing load (and other as yet unexplained features in its behaviour). But feel free to believe this if it makes you happier, the physics doesn't care what you or I believe, it will do what its laws dictate regardless.
10% is definitely more believable than 20%. There are arguments to adjust the crude calculation both ways:
The other heating sources will tend to reduce the percentage saving relative to the crude calculation
The vampire load (or whatever it is) will tend to increase the saving (for a setback only, not for a 24x7 reduction in thermostat setting)
So 10% might nit be too far out in your case. However there is currently no evidence to support thus (if that matters)
This post was modified 1 year ago 2 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.
But feel free to believe this if it makes you happier, the physics doesn't care what you or I believe, it will do what its laws dictate regardless.
This sort of abrasive remark does you no credit. There is no argument about whether the laws of physics apply, I accept them just as much as you do. Nor, frankly, is it about me achieving delusional happiness, which, equally frankly, is no business of yours. These ad hominem attacks create a totally unnecessary unpleasant tone, and as ever reveal more about those making the attacks than those they seek to attack. I don't know for certain, but I strongly suspect your tone (and @derek-m's) discourages comments by others.
The problem is that you say the laws of physics apply (on which point I agree), but you have so far failed to make those laws stick to my case, and by implication anyone else who is running a broadly similar heat pump setup, and who wants to know if they can get an energy saving by using setbacks. Even if the heat pump setup isn't broadly similar, the general principles, things like what does and doesn't need to be taken into account, are very likely to be widely applicable. It is a worthwhile exercise.
The latest row is that you have stated rather patronisingly that the energy saved by dropping the IAT one degree (which is not necessarily the same as dropping the thermostat one degree, especially in the case of cyclical setbacks) is down to simple maths, only expend paragraphs explaining why it doesn't apply in many circumstances, and then repeat them when I apply it. It may not have occurred to you that what I was doing was showing that there is a problem with the, what was it you called it, a safe general guideline? - apparently, it's not much use in my case, and I suspect many real world cases. @derek-m then made things even less clear by claiming:
As James has shown, this would probably provide an approximately 10% reduction if heat loss from a building, whatever the heating system. A heat pump based system should actually provide an even greater saving, because of the improved efficiency of it not working so hard at higher LWT.
I had to read that twice, but it appears to say a 10% saving is about right, regardless of heat source, and a heat pump should achieve 'an even greater saving'. Is it any wonder that simple O level minds like mine are confused?
Given that the weather forecast suggests the next few days will be relatively mild and have similar diurnal ranges, I have restarted using a setback, as from 2100 yesterday evening. Some nights I will have a setback, on others I won't, ideally, if I remember, on alternate nights. I will then have noon to noon records of hourly OATs, IATs and energy in for periods with and without setback. I will then make available the hourly OATs, IATs, setback periods and any other input variables used by @derek's model, and invite @derek-m to predict the daily energy in, by using his model. I will then post the energy in figures, and we can see how well the model predicts the observed values.
This sort of practical test is long overdue, having been sidelined by the other unproductive rows that we have allowed to dominate the recent discussion. I will say in advance that I will gladly accept the model if it demonstrates that it can predict energy in with reasonable accuracy (to within 1%? 5%? 10%? is any error random or systemic?), but equally, if it fails to predict the energy in, then I for one will not accept it in its current form, or indeed any form, until it can usefully predict the real world.
In the meantime, could I invite you (@jamespa) to expand on your notion of a vampire load, including the evidence for it, how it differs from a standing load, and what role, if any, you think the PHE has in creating this vampire load, including how it creates it. For ease of reference, here again is the plot of hourly data for the three control periods combined:
I don't see a vampire load, instead I see a tight correlation between OAT and energy in over most of the range of OATS, and then the energy in starting to flat line on the right hand side at an OAT of around 15 to 16 degrees (the ambient residuals plot also suggest something starts happening at about 15 degrees), which I suggest is the standing load. The value is around 200Wh, which seems plausible, equivalent to two 100W incandescent bulbs on for an hour.
Midea 14kW (for now...) ASHP heating both building and DHW
But feel free to believe this if it makes you happier, the physics doesn't care what you or I believe, it will do what its laws dictate regardless.
This sort of abrasive remark does you no credit. There is no argument about whether the laws of physics apply, I accept them just as much as you do. Nor, frankly, is it about me achieving delusional happiness, which, equally frankly, is no business of yours. These ad hominem attacks create a totally unnecessary unpleasant tone, and as ever reveal more about those making the attacks than those they seek to attack. I don't know for certain, but I strongly suspect your tone (and @derek-m's) discourages comments by others.
The problem is that you say the laws of physics apply (on which point I agree), but you have so far failed to make those laws stick to my case, and by implication anyone else who is running a broadly similar heat pump setup, and who wants to know if they can get an energy saving by using setbacks. Even if the heat pump setup isn't broadly similar, the general principles, things like what does and doesn't need to be taken into account, are very likely to be widely applicable. It is a worthwhile exercise.
The latest row is that you have stated rather patronisingly that the energy saved by dropping the IAT one degree (which is not necessarily the same as dropping the thermostat one degree, especially in the case of cyclical setbacks) is down to simple maths, only expend paragraphs explaining why it doesn't apply in many circumstances, and then repeat them when I apply it. It may not have occurred to you that what I was doing was showing that there is a problem with the, what was it you called it, a safe general guideline? - apparently, it's not much use in my case, and I suspect many real world cases. @derek-m then made things even less clear by claiming:
As James has shown, this would probably provide an approximately 10% reduction if heat loss from a building, whatever the heating system. A heat pump based system should actually provide an even greater saving, because of the improved efficiency of it not working so hard at higher LWT.
I had to read that twice, but it appears to say a 10% saving is about right, regardless of heat source, and a heat pump should achieve 'an even greater saving'. Is it any wonder that simple O level minds like mine are confused?
Given that the weather forecast suggests the next few days will be relatively mild and have similar diurnal ranges, I have restarted using a setback, as from 2100 yesterday evening. Some nights I will have a setback, on others I won't, ideally, if I remember, on alternate nights. I will then have noon to noon records of hourly OATs, IATs and energy in for periods with and without setback. I will then make available the hourly OATs, IATs, setback periods and any other input variables used by @derek's model, and invite @derek-m to predict the daily energy in, by using his model. I will then post the energy in figures, and we can see how well the model predicts the observed values.
This sort of practical test is long overdue, having been sidelined by the other unproductive rows that we have allowed to dominate the recent discussion. I will say in advance that I will gladly accept the model if it demonstrates that it can predict energy in with reasonable accuracy (to within 1%? 5%? 10%? is any error random or systemic?), but equally, if it fails to predict the energy in, then I for one will not accept it in its current form, or indeed any form, until it can usefully predict the real world.
In the meantime, could I invite you (@jamespa) to expand on your notion of a vampire load, including the evidence for it, how it differs from a standing load, and what role, if any, you think the PHE has in creating this vampire load, including how it creates it. For ease of reference, here again is the plot of hourly data for the three control periods combined:
I don't see a vampire load, instead I see a tight correlation between OAT and energy in over most of the range of OATS, and then the energy in starting to flat line on the right hand side at an OAT of around 15 to 16 degrees (the ambient residuals plot also suggest something starts happening at about 15 degrees), which I suggest is the standing load. The value is around 200Wh, which seems plausible, equivalent to two 100W incandescent bulbs on for an hour.
I didn't think the comment made by James could be misconstrued as 'abbrasive', he was merely stating that everyone can have their own opinion, but that the Laws of Physics cannot be ignored.
I fully agree that it would be far more productive if everyone could try to refrain from the use of inflammatory language.
It may also be useful that when re-posting other's comments, one should not use selective snippets, since it may be possible that these snippets may be taken out of context or misconstrued.
As an example, my previous statement:-
As James has shown, this would probably provide an approximately 10% reduction in heat loss from a building, whatever the heating system. A heat pump based system should actually provide an even greater saving, because of the improved efficiency of it not working so hard at higher LWT.
I would hope that readers would have realised, that to achieve a 10% reduction in overall energy consumption, it would be necessary to lower the IAT by 1.48C for the full 24 hour period each day. I apologise if anyone may not have drawn that conclusion.
I will be happy to run your data through my spreadsheet, and it would be helpful if you could also provide the following information.
Your other electrical energy consumption throughout each 24 hour period. Ideally this should align with the start of the setback period.
Both the hour and minute data in the same format as published previously.
Local weather station data covering the full test period, including OAT, Humidity, hours of sunshine etc.
Reasonably accurate temperature measurements on the pipework around your PHE, again for the full testing period.
Details of any physical changes that may have been made or variations in the method of operation.
Can you provide details of the probable concentration of anti-freeze within your system, and how this is used within your energy calculations, since your system is rather unique, in that only part of the system contains anti-freeze.
Your other electrical energy consumption throughout each 24 hour period. Ideally this should align with the start of the setback period. I can only do this crudely, for longer time periods, by comparing the main house kWH meter (which also includes the heat pump) and the heat pump only kWh meter. That said, my daily use probably doesn't normally vary that much day on day, though it will, obviously, between seasons.
Both the hour and minute data in the same format as published previously. Not a problem.
Local weather station data covering the full test period, including OAT, Humidity, hours of sunshine etc. There isn't an ideal local weather station with all the data. The nearest WOW (Weather Observations Website) hobbyist one - limited data, and not always complete - is about 5 miles distant, and the proper Met Office ones are fifteen or more miles distant. These do have current data, but again not for all variables. The MIDAS archive is more complete, but is historical data only, latest data usually a year or more old. I do have the 3 hourly data HA records, that is probably the best but far from perfect limited dataset, and it might be possible to 'triangulate' OATs, though I am not sure I would trust the results. Unfortunately, hours of sunshine aren't recorded, though weather is, as a text description. The last three days for one of the Met Office weather stations:
Reasonably accurate temperature measurements on the pipework around your PHE, again for the full testing period. Not practical because (a) I have to do it manually, head stuck in bottom of airing cupboard and (b) the IR thermometer on black insulating tape on the pipes is the only way of taking the measurements.
Details of any physical changes that may have been made or variations in the method of operation. No physical changes likely to be made over relevant periods, method of operation (setback or not) I can keep a record of, ditto for any WCC baseline values (what the WCC curve is when the auto-adapt script hasn't changed it).
Can you provide details of the probable concentration of anti-freeze within your system, and how this is used within your energy calculations. I think it is 10%, and it goes into the energy out (not in) calculation as an adjustment to the specific heat of the circulating fluid (4.05 instead of 4.2 for water).
It is going to take a little while to collect recent setback data, once I have some I will post it.
Midea 14kW (for now...) ASHP heating both building and DHW
I daresay this chart may be of some interest, because it covers a three day period of unusually stable outside air temps while running a six hour overnight setback. The big middle of the day spikes are DHW:
My comments:
(1) looking at the OAT, as measured by the sensor in the heat pump, and what happens during the setbacks, it seems to me the running heat pump may lower the OAT by one or two degrees in the prevailing conditions (mild steady OAT, no defrost cycles).
(2) the actual IAT recovers to within half a degree of the desired IAT by 0800, meaning the heat pump has satisfied the my comfort requirement (I can live with an actual IAT half a degree either side of the desired IAT)
(3) overall, the house is in steady state, insofar as the IAT shows no overall trend up or down
(4) the OAT is in broad agreement with a local Met Office WOW hobbyist record (about ten miles away, the nearest one has no current data: no discernable diurnal range, very slight upward trend over the period):
@derek-m - I suggest the relatively steady state conditions make this an opportune period in which to test your model's predictions of energy in against the observed values. The weather (this from an official Met Office website about 15 miles away) has also been rather constant:
I therefore attach the last week's minute and hourly data with energy in/out variables removed, making this a true blind test, and look forward to seeing the predictions your model makes. The outcome variable of interest is the energy in, in kWh per hour and day, and I suggest the period of interest is the last three days, from let's say 0600 on the 16th to 0600 today (19th Dec). I appreciate this is a short period, but it does have the useful feature of being relatively steady state, and thereby achieves some degree of control over the other predictor variables eg solar gain will not have varied significantly during the period, nor has the human behaviour element varied significantly during the period. My auto adapt script was running throughout the period.
Thank you for providing a redacted copy of the raw data, I am always amazed at how trusting humans can be.
Without giving away any state secrets, could you please identify when the compressor, primary water pump and secondary water pump, are actually operating?
Can you provide an estimate of how much of the DHW thermal energy may have 'leaked' out, or been 'released', within the thermal envelope, and how much was 'flushed' down the drain, so to speak.?
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.
