The 11.6, which is definitely for the hour not minute, is definitely an artefact, caused by the erroneously high energy out figures from 13:53 to 13:55 inclusive when the compressor was still running, so I can't use if(amps_in = zero) ie compressor isn't running then kWh_out = 0 to remove those high readings (the amps_in zeros that do then appear are because it's cycling at the moment because of the mild weather).
Unless the maestro (@derek-m) can think of a solution, I think I am going to have to identify those minutes by the valve position pattern, eg something like:
if htg_on_off = 1 and dhw_on_off 5mins ago = 1 then kWh_out = 0
There is a setting for how long after the DHW reaches the required temperature to run the pump (defaults to 5 minutes but can be read). The DHW temperature usually rises a further degree C in this time.
The simple way to model all real time systems is as a Finite State Machine .
@filipe - this is the key data at the transition time with rather meaningless minute COPs added but they do show how the abberation occurs:
At the beginning of the space heating period, the high residual delta t from the dhw heating period causes the apparent energy out to be far higher than it should be given the actual energy in.
The simple way to model all real time systems is as a Finite State Machine
When anything in computing is described as simple, it usually means rather the opposite! A quick google and I am pretty sure I was looking into another infinite rabbit hole, descent without end, world without end.
I've just spent another morning drilling holes in my head courtesy of HA, as it looked as if it might have a convenient way of running my scripts. But of course it doesn't, courtesy of HA bizarre world view that neither the past nor the future exist. The scripts fail because HA exterminates them after 60s on the grounds that HA doesn't manage long running process. In HA's world, the future isn't a foreign country, the future doesn't even exist. The is no such time as next hour, tomorrow or the week after next in HA's world, instead, the world ends in 60s. Not with a bang, not even with a whimper, but with a cryptic entry in HA's logs.
Midea 14kW (for now...) ASHP heating both building and DHW
Here is a rather cryptic table of htg and dhw COPs with the amps_in shifted downwards one cell at a time:
Shift 3 sort of works initially, but the later htg COPs are too low (>1 with even bigger shifts). Works best by just over-writing the first three htg amps in with the last three dhw amps in:
Still not perfect, but may be good enough, bearing in mind these are minute data and extremes will be ironed out a bit in the hour data.
PS should add this is in effect a slightly tweaked version of what @derek-m did.
When anything in computing is described as simple, it usually means rather the opposite! A quick google and I am pretty sure I was looking into another infinite rabbit hole, descent without end, world without end.
Well, I beg to differ. The important thing is to understand how code is designed. It seems to me the COP is irrelevant once the compressor stops but there will be residual heat transferred (energy out) after the compressor stops.
I really don’t understand what you are displaying above. The observations I make show the DHW COP gradually dropping from the maximum quoted by Midea for my model as the DHW sensor rises. It is also affected by outside air temperature quite significantly, but that is only relevant when comparing different plots of the DHW heating cycle, which also vary depending on the initial DHW temperature.
Your readings from Modbus are effectively events some of which indicate fundamental state changes (compressor on/off, pump speed change, compressor frequency change etc.). Sometimes people write code full of conditions instead processing events according to different states. The later is far more elegant and simple to understand albeit that some extra code is needed to implement the finite state machine. The ASHP hardware is itself a finite state machine.
Mine is about 10 and a half months old and has consumed about 2800 kWh for a 300m2 home. I haven’t heated 24/7 though.
@filipe - the problem I am trying to deal with is the lag: at any one reading (all the data is collected in one go once a minute), the energy out has come from an earlier energy in, if only because the water takes time to move from the heat pump to the inside of the house. Normally this doesn't matter, it gets ironed out, but the exception is the combination of milder weather, giving a low space heating RWT, and the end of a DHW heating cycle, which gives a high RWT. Once I have enough data (this bit of the system has only been running for a couple of days) I am going to see if this can be dealt with by using longer time periods (in effect smoothing out the aberrations).
Bear in mind this thread is in its way part of the Beginner's Guide thread. I'm not trying to get perfection, just a 'good enough' monitoring system that allows me to keep an eye on how the system is performing. Writing code full of conditions is simply too complicated, and overkill. I'm still not sure what 'finite state machine' actually means, or what one would add. The writer in me also asks why the qualifier (finite), because it's use implies there are times when it may not be finite, which means it must be infinite, and an infinite state machine sounds rather improbable. Without additional coverage of how it works, and what advantages it brings, it remains just incomprehensible jargon to the layman.
The other aberration I have to deal with is all the different ways of monitoring energy in give different results. Usually, but not always, the external meter shows the largest kWH in, then comes the Midea kWh now minus kWh some time (hour/day/week) ago, while the calculated (amps x volts) value is always the smallest. At the moment, the metered value is about 10 to 12% higher than the calculated value. As I have said before, this may well be because, as the Midea manual implies, the amps value (which is also always an integer - rounding errors possibly, depending on how founding is done?) only reports amps supplied to the compressor. If (I need more obs to confirm) the calculated value is always a certain percentage below the meter value, then I can manage that by using a correction factor.
I didn't comment on the charts you posted earlier because they don't really make much sense to me. At 0722 monthly peak 15 min demand is 1.79kw, and March usage is 3.43kWh; at 0818 the 15 min peak is 2.15kw, and March usage is 5.02kWh??? Charts (and numbers on/near them) need explanations unless they are brilliantly clear.
2800kWh over ten and a half months which includes most of the current heating season for a 300m2 home sounds on the low side, unless your home is very well insulated. It also needs context: did that amount of energy keep the house at design temps? Anyone can run their system so it consumes say 3000kWh per annum, but that doesn't mean anything unless we know the outcome as well as the input.
I didn't comment on the charts you posted earlier because they don't really make much sense to me.
I would have thought that in the context posted it was obviously the power shown on the graphs rather than the surrounding summary data that I was discussing (the Emporia Energy Monitor had only been installed for a day or so!). This one is after the start of the compressor coming down to a steady state.
I don’t understand why you don’t use an external electric meter on the circuit. I thought you wanted independent verification as you don’t trust Midea!
Seeing your graphs covering a compressor cycle would help. Perhaps your data used to produce your graphs is incorrect. As I collect better energy input now using the new CT clamp meter I find the Midea data to be reasonable accurate. Even your system should be working at a COP of 4.5 when it starts from cold.
I think you are very mischievous twisting words like finite. Plenty on the web about finite state machines. Reading about how heat pumps works as machines is worthwhile.
I didn't comment on the charts you posted earlier because they don't really make much sense to me.
I would have thought that in the context posted it was obviously the power shown on the graphs rather than the surrounding summary data that I was discussing (the Emporia Energy Monitor had only been installed for a day or so!). This one is after the start of the compressor coming down to a steady state.
I don’t understand why you don’t use an external electric meter on the circuit. I thought you wanted independent verification as you don’t trust Midea!
Seeing your graphs covering a compressor cycle would help. Perhaps your data used to produce your graphs is incorrect. As I collect better energy input now using the new CT clamp meter I find the Midea data to be reasonable accurate. Even your system should be working at a COP of 4.5 when it starts from cold.
I think you are very mischievous twisting words like finite. Plenty on the web about finite state machines. Reading about how heat pumps works as machines is worthwhile.
Phil
Whilst I try not to be critical of other's posts, I feel that the use of the description Finite State Machine (FSM) or Finite State Automation (FSA) is not that productive in the great scheme of things. As an Engineer I am quite familiar with many engineering terms, but would not use them on the forum unless necessary, and without giving a suitable explanation if required.
It is extremely unlikely that I would state that 'one can obtain a bar of chocolate from a finite state machine', but I would probably state that 'one can obtain a bar of chocolate from a suitable vending machine'.
I would respectfully suggest that the use of a systems correct description, is much more productive than using a coverall term, that few, if any, fully understand.
I would have thought that in the context posted it was obviously the power shown on the graphs rather than the surrounding summary data that I was discussing (the Emporia Energy Monitor had only been installed for a day or so!). This one is after the start of the compressor coming down to a steady state.
I'm sorry, but I still don't understand what that image shows. Frankly, it's verging on dreadful. I guess it is what happens when you try to do things on a ridiculously small screen. At least you have now underlined the fact it is only a day or so's worth of data, which to me is not long enough to get a reasonable picture, all the more so as it has been mild, and most of our heat pumps will have been idling.
I could go on at length, but here are a few problems with that image:
(a) the 'title': Walnut View > ASHP? Is this the 'nut' case?
(b) the numbers at the top are a mix of kW and kWh. Most of the time, we are not interested in kWs per se, but in kWh (kW-hours), ie how much energy we have used, not what the moment by moment power is. Yes, we need kW data, but only as a means to get kWhs.
(c) the left hand figure, Monthly peak 15 min demand (in kW) is almost meaningless. If min means minutes, then I have no real concept of what a monthly peak 15 minute demand is, let alone why it matters enough to use up valuable screen space displaying it. Because time is included, it is really an energy (kWh) value. Perhaps it means the highest monthly energy use was 2.15 kW over 15 minutes, but does that also mean 8.6 kWh? The use of the 15 min period is baffling. The more I think about this figure, the more it confuses me.
(d) the middle figure is OK as an concept if it means what I think it means, total kWh use over the entire period (as shown, March, as explained by you, only a couple of so of days in March), but the value shown seems extremely low. When it is very cold, I use getting on for 5kWh per hour.
(e) the right hand figure (confusing uses watts not kW) says at 8:16am the average rate was 1549 watts. I have no idea what this means. Average of what? Over how long?
(f) the chart itself I think is meant show minute kW values, in simple terms how 'hot' the heat pump was running at each minute. As noted before, this figure is only a means to an end, ie kWs (how hot my engine is running/bulb is burning ie power) are a step on the way to determining energy use. It is the latter that matters. I cannot really see the point of minute by minute plots of kWs. It is much more useful to know how much energy we have used over much longer periods (hours perhaps, days, weeks, months, seasons).
I don’t understand why you don’t use an external electric meter on the circuit. I thought you wanted independent verification as you don’t trust Midea!
Perhaps you haven't had the time to read some of my posts. I often mention that I do have an external third party meter that supplies only the heat pump, and that is one of the reasons why I question the Midea data. To recap, I have three ways of getting energy in data: (1) external meter (2) the Midea app method (current lifetime use minus previous lifetime value eg if the value now is 11030 and 24 hours ago it was 11000 then I have used 30kWh in the last 24 hours and (3) summing collected amps x volts collected over the wired modbus connection. The problem is none of them agree, so either they are all wrong, or perhaps one is right, and the others are wrong. I strongly incline towards believing the external meter is likely to be most accurate one, and I am currently trying to determine whether there is a reasonably constant variation between the values, in which case I can apply a correction factor, or the variations are random. This will take a little while, as I need more data than I currently have for the calculated (method (3) above) values.
Even your system should be working at a COP of 4.5 when it starts from cold.
I very much doubt this. Heat pumps have to work harder when starting from cold (that's one reason why we run them 24/7). Not even Midea's own data claim COPs this high. At lower ambient temps, around zero, and at radiator LWTs (40 degrees plus), their data has COPs in the 2-3 range (source: page 44-45 of the Engineering Data manual). According to the same data, to get a COP of over 4 at a LWT of 50 degrees, the ambient temp needs to be 20 degrees or more, when I respectfully suggest only a berk would be running their heating.
This is a fundamental problem with heat pumps, and especially heat pump marketing: they do absolutely brilliantly, sky high COPS - but only when you don't need them.
I think you are very mischievous twisting words like finite.
Maybe, but I detest redundant words. Why do we always swap out these days? I don't, I just swap, because I know swapping includes movement, both in and out, and so I do not need that extra word. Likewise, the word 'finite' in 'finite state machine' on literal meaning is redundant (because apart from God or whatever your deity is, everything is finite), unless the word forms part of a phrase 'finite state machine', in which case it is jargon, and needs both justification and explanation by the person who has chosen to use it - exactly as @derek-m says.
Don't even think about getting me started on reaching out and sharing.
PS I should perhaps add in that I myself am not immune to redundant word situations, which on rare occasions may spill over into ongoing redundant word situations, though so far I have managed to avoid an infinite ongoing redundant word situation.
@cathodeRay Thanks for your much work getting the data of the pump.
Last year I installed mechanical and electrical a midea clone named Kaisai with 8.4 KWh. Also assambled a heatcounter in the 28 copper pipe and an electrical 3-phase-counter. I bought the new counters with 2% and 1% accuracy. They works fine and I red their data daily at the same time since the february 11. The data show very small differences with the data of the midea clone daily and over the month or all time. The resulting COP from the two measurement systems since 11. february is exactly 4,43. I mean that this is a good result. I only bought and assembled the counters because I did not know before that also two counters working in the heatpump.
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.
