Posted by: @ivanhoew

↑

Sampling rate is every 30 seconds ray , if i zoom in i can see that rate .

i probably have not explained myself vey well ,  i have been using weather compensation all the time ,  i am just using a fixed weather state to try to compare eff.

yes the data can be downlaoded to excel , i am hoping to use this to calulatec my cop .

All that is good news, and I now understand what you are trying to do.

I did try something similar, fixed flow temp instead of WC, not long after I had my heat pump installed. It turned out to be more expensive, at least on a Mk 1 Eyeball assessment.

You may know how to calculate energy in and energy out to calculate COP, if not, for each hour (this is a 60min trailing COP)

energy in over the hour in kWh = (mean amps over the hour x mean volts over the hour (or 240 if you don't have volts data)) / 1000

mean amps = sum for the hour, then divide by number of readings (120 if every 30s), ditto for volts

Energy out is a bit more complicated:

basic formula is energy out over the hour in kWh = mean flow rate x (mean LWT - mean RWT) x specific heat capacity of circulating fluid (4.2 for water, adjust down if you have glycol)

note the importance of 'over the hour' - strictly speaking this is power (kW) but because it is over an hour, it is also energy (kWh)

watch out for units, especially flow rate, above formula is for flow rate in litres/sec, adjust as necessary if you have different units 

I do all this in python scripts and then log it in csv files. You can probably set up a spreadsheet with a couple of formulas an hour apart, and then drag then to cover other hours. Be careful, mistakes are easy to make! 

Once you have hourly data, you can easily sum to get weekly data, monthly data, seasonal data, whatever interval you want data. 

Funnily enough i was just doing this yesterday ..

I assumed 20% anti freeze for a 3965 specific heat .   i couldn't get a rsponse from the installers to find out the exact % of coolant .

its all a work in progress graphs wise .

ill do some colder weather later .

regard

robert

@ivanhoew - yes, I agree, still a work in progress on the graphs! Try to make them clearer eg clearer x axis labels, delta what? etc. It's also not clear for example what the COP is. Momentary (spot, at that instant which isn't actually very meaningful at all)? Last minute (ditto)? It seems to me an hour is about the shortest interval for which a COP makes sense.

My standard monitoring chart has two plots, one for the temperature variables (minute data), and one for the energy variables, each bar being the trailing hour's values ie 1200 bar is the hour from 1100 to 1200. Strictly speaking the COP should be bars too but a line makes it much easier to read.  This is what my chart for the last 24 hours looks like. The data is in csv files (one for minute data, another for hourly data) and the plots are produced using plotly, a python module: 

It shows typical low frequency cycling at moderate OATs. The steps in the OAT are due to the cooling effect of the heat pump on the local air when it is running. Your first chart appears to suggest the same thing, the dips in OAT coincide with the heat pump running. 

still gathering data to help me decide . a few more graphs i have compiled ..

Looking at these i can see best cop,so far , is about 29c  FT , and evidence cop is worse as it gets colder , which i know is a given , but i like to start with the basics  🙂

I also noticed my worst cop is above feezing ,which is interesting , and prob due to me experimenting with a higher FT.

in the three day graph i am generally hovering in the 3- 3.5 range , over some pretly cold amb temps , so thats not too horrific .

regards

robert

@ivanhoew - these charts are difficult to interpret because of inadequate labelling, and insufficient accounts of methodology. For example, in the first chart the COP is very volatile, and regularly dips to -7 or below. A COP of -7 is quite an achievement, even during a defrost cycle (when a COP might appear to be negative, because of the reverse energy flow, but even that is in a way an artefact of how we do the calculations, insofar as a negative COP doesn't really make sense)! How do you calculate the COP, and over what interval?

Posted by: @ivanhoew

↑

Looking at these i can see best cop,so far , is about 29c  FT

I am not sure I can see any evidence for this. Can you point it out?

@ivanhoew  

I must say I echo what @cathoderay is saying here. 

I hope what Im about to add doesn't offend you  (and apologise if it does - its not my intent) - but is there any possibility you are overthinking this?  Fundamentally heat pumps are pretty simple, at least in principle,  to optimise (provided they dont have a buffer/PHE/LLH which complicates matters a lot):

• remove/disable all or most external controls and certainly any external thermostats, unless they have been specifically designed for heat pumps (only Homely at present SOFAIK)
• balance the emitters
• get the weather compensation adjusted as low as you can consistent with the house heating up to your required temperature, with the heat pump operated 24*7
• leave largely alone unless you are experiencing identifiable problems or want to tweak eg for a ToU tarif

To a very large extent you have to trust this, because making reliable comparative measurements of consumption, with a sufficient degree of accuracy to be meaningful, over a period less than several weeks (you possibly even need a whole season) is nigh-on impossible. 

You also have to trust that the manufacturers of the heat pump know their system reasonably well, and if it cycles or displays other behaviours, its because it has to to satisfy the demand. 

All of course subject to the assumption that there is no identifiable problem. 

Of course its your system and you are entirely entitled to do whatever you like.  There is no doubt this stuff is fascinating (if you like that sort of thing) and offers endless potential for playing around with scripting etc.  There is also the possibility of playing 'top of the scops' on openenergymonitor, if you are that way inclined!  That said there is also no doubt that it can become addictive.

Incidentally its entirely possible that worst COP is just above freezing.  The wet climate plus cold temperature increases the need for defrost cycles (which sap energy and thus COP) and this is particularly true between 4 and 0.  Once it gets below zero the air often becomes drier and, as the temperature falls further, quite noticeably so.  Conditions don't always follow this pattern, but do so more often than not.  Personally I find -4 more comfortable than +4 due to the (usual) difference in humidity!

(@ivanhoew)

Hi Robert

I have a very similar house to yours, and am considering installing a 12kW Harnitek combo.
Would love to compare notes on your install and costs - tried to PM you but as a new member I cannot until I have made 10 posts.
Can you either PM me your email, or can we start a topic based on the Harnitek combo?

Thanks
Jonathan

Posted by: @ivanhoew

↑

I have been trying to understand and achieve the best settings for my radiator water temperature on my heat pump.

My setup is , Harnitek 12 kw ashp .with internal 250 tank setup .this is called the Harnitek combo.

1865 farmhouse , shrewsbury, welshpool area , 180m elevation ...130m3  4 on 4 , shwiff system insulated all exterior walls ,270mm loft insulation ,pvc double glazing , with argon , one cavity wall full of poystyrene  balls ,6 kw of panels and 4.9 kw battery . new heating radiators and piping to suit heat pump .

I tend to run the hs at 16c most of the time with 15c at night .

My question is , if  i set the heat curve at 32c @ 3c ambient , i get one waveform , and if i set it at 31c ,i get a  diferent one .and different cycling rates ..

Posted by: @jamespa

↑

Incidentally its entirely possible that worst COP is just above freezing.  The wet climate plus cold temperature increases the need for defrost cycles (which sap energy and thus COP) and this is particularly true between 4 and 0.  Once it gets below zero the air often becomes drier and, as the temperature falls further, quite noticeably so.  Conditions don't always follow this pattern, but do so more often than not.  Personally I find -4 more comfortable than +4 due to the (usual) difference in humidity!

Ah ! thats fascinating James thank you , that fills in the blanks as to why its working like that .

so !  i have done a little bit ....oh yes i have !

i found the areas of my logs that had ambient temps within a degree ish , and had a section of two hours each , one at 32c flow temps,  (NOTE these are the inlet temps , so the high side of the average ). and one at 29c .

you can see the 29c one cycling more often , around 20 minutes gap .

cop for these is like this ..

i then picked out the smoothest hour for each period ..

which gave these cop numbers ..

now i could go and look at my energy usage for those periods ..

These work out,for the two hour period , to 2.8 kw for the 32c Flow temps ,and 2.2 kw for the 29c flow temps .

for the single hour  its 1.5 kw and 1.1 kw .

lets see how that correlates to the average amps used ..

2 hour period , 32c amps average is 4.47 amps , 29c amps average is 3.11 amps .

so 32c is 1.44 X 29c  ...

1,437 X 2.2kw = 3.16 kw ,miles away from our 2.8 kw i measured .. so thats doesnt work out ,

lets try the 1 hour test period ..

averages are ..

2.94 amps for the 29cFT  ,

and 5.01amps  for the 32c FT .

power used was ..1.5kw and 1.1 kw .

so  1.7 x 1.1kw = 1,87 kw ,again not  correct , so the average amps way of working it out doesn't seem a winner .

so , so far , it appears that the lower flow temps of 29c over 32 c , gives a 27-36% drop in kw used .

i have probably made loads of mistakes in my maths here , and will come back and edit it furiously later ! 🙂

regards

robert