Definitions of SCoP

[sharpener]

So I read that SCoP is formally defined as total heat output over a year divided by total electrical energy input.

 

As the instantaneous CoP is a function of Flow Temperature and Outside Air Temperature this leads to performance tables like this:

Are the OAT figures in the table then the design minimum figures, and so already take account of the variation through the seasons of the actual OAT. If so, what is the source of the OAT probability tables used for this (e.g. CIBSE?), which will vary from region to region?

 

Furthermore, with Weather Compensation (which is mandatory under MCS rules) the flow temperature is itself an inverse function of OAT. So do these tables already take that into account as well?

 

If not then they will underestimate the year-round average CoP as the flow temp needed to achieve a constant ambient temp inside the house will also vary with the seasons so will be less that the figure in the table.

 

Thanks for any enlightenment!

[S2D2]

I could be wrong, but I think you're looking for EN 14825. There's a decent explanation here: https://www.varmepumpsforum.com/vpforum/index.php?action=dlattach;ts=1505933467;topic=65119.0;attach=49952

 

An average year is binned into temperatures which gives a good agreement with lab test data, I was pleasantly surprised with how accurate it should be. There are quite sweeping climate classifications (average, cold, warm) but it seems to do the job.

 

It's calculated from lab testing points, so I believe it would be performing weather compensation in those conditions yes.

 

Caveat this entire post with I've never actually read EN 14825, I just thought it was interesting when looking at A2A manufacturer claims.

[JohnMo]

OAT are just the output kW at different outside temperatures.  SCoP is just taken as a function of flow temperature. It does not take account if WC, just the flow temp stated.

 

Biggest thing is get an ASHP good SCoP at the flow temp nearest what you expect to flow. Then the biggest change after that, is the install. Things like small buffers piped as a 4 pipe, will kill CoP anyway.  

[sharpener]

2 hours ago, JohnMo said:

Biggest thing is get an ASHP good SCoP at the flow temp nearest what you expect to flow.

 

My point exactly, what is it I "expect"?

 

If I design for 99% worst case OAT (-0.2C for the SW) then using weather comp I expect the flow temp to vary from say 30C in spring and early autumn to say 55C in the depths of winter. I don't anticipate spending a high proportion of the time at 55C but I think it is a reasonable corner case to design for. At this temp the Vaillant Arotherm + 12 is still producing more than the nominal 12kW output and the claimed SCoP (whatever that turns out to mean) of 3.63 is enough to beat the price I am currently paying for oil.

 

Problem is that the Vaillant installer I spoke to recently won't under any circumstances design for flow higher than 45, which seems not to take advantage of the machine's capbility on the limited number of days it might be needed. And chose a -3.0C worst case OAT which increases the required emitter sizes even further. 

 

IMO it should be my choice how to trade operating efficiency in extreme conditions against investment in 12 new radiators.

 

 

 

[sharpener]

3 hours ago, S2D2 said:

I could be wrong, but I think you're looking for EN 14825. There's a decent explanation here: https://www.varmepumpsforum.com/vpforum/index.php?action=dlattach;ts=1505933467;topic=65119.0;attach=49952

 

I think this might be an old link? It now redirects to what appears to be the landing page for a discussion forum in Swedish. I assume you meant a specific thread @S2D2 but it doesn't seem to be there any more.

[Nickfromwales]

33 minutes ago, sharpener said:

If I design for 99% worst case OAT (-0.2C for the SW) then using weather comp I expect the flow temp to vary from say 30C in spring and early autumn to say 55C in the depths of winter. I don't anticipate spending a high proportion of the time at 55C but I think it is a reasonable corner case to design for. At this temp the Vaillant Arotherm + 12 is still producing more than the nominal 12kW output and the claimed SCoP (whatever that turns out to mean) of 3.63 is enough to beat the price I am currently paying for oil.

If you are on oil, what improvements have you made to make a move to an ASHP sensible? You can pretty much forget calculations including maths from an assumed flow of 30^oC?! If you have not nearly doubled the size of the radiators, increased the size of all of the primary pipework, improved the insulation / 3G glazing, and more, then I think you are being VERY optimistic of an SCoP of over 3.5!

Retro-fit of an ASHP is often done poorly, or badly, or worse. Increasing the size of you roil tank x2 and strategizing when you buy oil (when it is at it's cheapest) may be a better solution for the next 3/5/7 years or so. Costs of electricity (increases of) must be inclusive of your overall decision making here per the same 3/5/7 year period  .

What standard is the dwelling at for airtightness and insulation / overall fabric quality, currently? If it's a passive house I'll stand down. 

[sharpener]

Purpose of this thread is to try and understand how SCoP is arrived at short of buying and wading through the whole of BS EN 14825. In particular it seems to me that the averaging process might not/does not take account of weather comp. It appears to be some kind of weighted average of the test points at -7, -2, +2 and +7 for two cases, low temperature operation and high temperature, which is not (now) really a high temperature anyway.

 

Maybe using the performance tables here https://www.vaillant.cz/downloads/projek-n-podklady/kl-06-e2-verze-01-18012023-2564719.pdf  and weather statistics from e.g. CIBSE I can build a more appropriate model but was hoping to avoid this.

 

Yes from Table 6.3.11 in this link the CoP is about 2.2 at A0/W55C but these conditions will prevail for a small part of daily cycle over a very small fraction of the year.

 

As mentioned elsethread we have a s/h barn conversion, it has D/G and MVHR but scope for wall insulation is virtually nil. The building is long and thin so zoning works well in it. Of course I am canny with oil purhasing, belong to a local buying group and buy nearly all I need in the summer, but the boiler is over 25yrs old and I would rather replace it at a time of my own choosing, ideally with an HP because of the 7kW of PV.

 

Have considered Veritherm assessment of heat loss coefficient but they cannot do it until August or September when the nights are again long and cool enough, and given the lead time for P/P that will be too late for W23/24.

 

There is a lot of free PV in the shoulder seasons so outright efficiency is not the main driver however and as always in these matters there is a balance to be struck between capital outlay and running costs.

 

Here the best is the enemy of the good and as @JamesPa has also discovered installers are extremely conservative and are wont to overdesign to avoid callouts e.g. chooosing -3C as a starting point in coastal Devon. Whereas I have a taste for elegance in design, have years of experience in engineering projects, and can afford to take some risks because I have an AGA and WBS as supplementary means of heating even if the benighted MCS process will take no account of them.

 

So all-in-all and with due thanks for the contributions so far I am still after a good explanation of how the SCoP figures are arrived at in the first place!

 

 

 

 

 

 

[DanDee]

2 hours ago, sharpener said:

 

I think this might be an old link? It now redirects to what appears to be the landing page for a discussion forum in Swedish. I assume you meant a specific thread @S2D2 but it doesn't seem to be there any more.

https://www.varmepumpsforum.com/vpforum/index.php?action=dlattach;ts=1505933467;topic=65119.0;attach=49952 link to the file in question

 

https://www.varmepumpsforum.com/vpforum/index.php?topic=65119.msg690076#msg690076 link to the post

[Marvin]

11 hours ago, JohnMo said:

 Things like small buffers piped as a 4 pipe, will kill CoP anyway.  

Hi @sharpener

 

Expanding @JohnMos comment further....

 

IMHO a CoP is only an  indicator of what is possible with a particular ASHP.

 

Similar to about as useful as how many miles per gallon a manufacturer says a car can travel! 

 

However in this case the manufacturer has no control over where it is installed and what it is connected to (and like a car how it is used.)

 

Simply put, referring to the CoP your measuring how much energy is put into an ASHP in relation to how much heat is produced by a given temperature difference between the air outside and the exiting water temperature, at exit, from the ASHP.

 

I am not surprised that companies are very conservative when sizing an ASHP.

 

I assume you are trying to work out what would be a reasonable size ASHP for your building.

 

To decide the suitable size of an ASHP for your building with more certainty than "an expected size and plus some" requires much, much, much, more information all the way down to how the end user wants or needs to use the system. Something not really practical or sometimes possible to do when quoting. 

 

If you want to take the long path of finding out with more certainty, it can be done as people here have, but don't expect a quick answer.

 

Good luck

 

Marvin

 

 

[SteamyTea]

Couple of technical points, the efficiency of a heat pump is proportional to the refrigerant temperature, usually around -25°C (248K), the outside air temperature AND the delivery temperature, between 35°C to 60°C (308K to 333K).

The flow rates of all the fluids (outside air, refrigerant, working fluids and inside air) makes a difference as well if calculating efficiency on temperature difference alone, rather than energy differences, which is the correct way to calculate it.

 

When looking at Outside Air Temperature (OAT) distribution, care has to be taken on which block of data is used.  Taking the last 120 years of data from the CET dataset can easily give different results.

Below is a Probability Density Function chart of all the data, data between 1960 to 1990 and 1990 to 2020.

The Skew, which is really just how lopsided the distribution is about the mean of each sample (or subset of samples) is same at -0.1°C difference (1990 - 2020). Similarly the Standard Deviation at 5.2°C.

But the mean temperature has increased by 0.8°C in the last 30 years (1990 - 2020) compared to the previous 30 years (1960 - 1990), which is significant, more for the design of the house than the heating system.

If you know your long term mean OAT and Standard Deviation, you can use a Normal Distribution to give you a good enough distribution at the 3 SD level (99.6%), medical research i.e. the treatment you get, is calculated at the 1 SD level, (66.2%), so our heating and cooling is calculated more rigorously than chemotherapy.

 

 

 

 

Looking at Minimum Temperatures, which is the main point of fitting a heating system, the Minimum Temperatures have changed from a Mean of 5.8°C (1960 - 1990) to 6.4°C (1990 - 2020) and the Standard Deviation by 0.1 (4.9 in 1960 - 1990 and 4.8 in 1990 - 2020), so not really significant.  Skew is also very similar -0.1°C (-0.3 in 1960 - 1990 and -0.2 in 1990 - 2020).

 

 

If you are worried about overheating, which features more on here than underheating (as we know we can just plug in a fan heater when it is cold) then Mean temperature has increased by 1°C (13°C 1960 - 1990, 14°C 1990 - 2020), Standard Deviation has a 0.1°C change (6°C between 1960 - 1990, 5.9°C 1990 - 2020).  Skew is similar (though they are not really related in the natural environment) 0.1 (0 between 1960 - 1990, 0.1 between 1990 - 2020).

 

 

The last two years we have had heatwaves, so here is the data for 2021 and 2022.

Mean Temperatures 14°C (1990 -2020) and 14.7°C (2021 and 2022), Standard Deviation 5.9°C (1990 - 2020) and 6.2°C (2021 and 2022), Skew 0.1 for both.

Two spikes on the 2021 and 2022 subset (blue line) (~10°C and ~20°C and a couple of smaller ones later on (~28°C and ~31°C, lovely).

 

 

Minimum Temperatures are not greatly different. Means 5.8°C (1990 - 2020), 6.4°C (2021 and 2022), Standard Deviation 4.8°C (1990 - 2020), 5.1°C (2021 and 2023), Skew -0.2 for both.

The Standard Deviation changing by 0.3°C is significant, but probably NOT a long term change, but like the effects of the French Revolution (1789 - 1799), it is too seen to tell.

 

 

Edited June 4, 2023 by SteamyTea

[JamesPa]

There is an extract with a discussion on how serr and scop are calculated here.  https://www.sciencedirect.com/topics/engineering/seasonal-energy-efficiency-ratio#:~:text=Basically%2C the SCOP calculation method,temperature over a heating season.

 

Still not exactly the exact definition, but closer.  

[JamesPa]

https://www.google.com/url?sa=t&source=web&rct=j&url=https://www.varmepumpsforum.com/vpforum/index.php%3Faction%3Ddlattach%3Bts%3D1505933467%3Btopic%3D65119.0%3Battach%3D49952&ved=2ahUKEwjkyLf-mKn_AhVDmFwKHX2oAaoQFnoECBMQAQ&usg=AOvVaw0p3Zn86KLVTlAcbi-xHp8j

 

Also referenced upthread, gives an example calculation.  So far as I can see weather compensation is not taken into account.  If this is correct theoretically one should do a bit better.  However real world losses are almost certain to counter this I would think.

 

I posted a few months ago a weather compensation calculator which estimates how much difference various WC algorithms make, using a bin analysis of OATs and published cop.  You could dial some figures into this if it helps.  Very roughly (from memory), using Mitsubishi COP figures, it's 25% at ft55, 15% at ft 45 and 5% at ft35.

 

Hope that helps somehow.

[sharpener]

10 hours ago, DanDee said:

https://www.varmepumpsforum.com/vpforum/index.php?action=dlattach;ts=1505933467;topic=65119.0;attach=49952 link to the file in question

 

Thanks @DanDee, that seems to work now and the article is very much the kind of explanation I was looking for. The method takes account of the reduced load at higher OATs by assuming the HP is part-loaded at the same flow temp, whereas with WC it would in practice be run at a lower flow to achieve the same end at a better CoP.

 

My climatic situation is a good deal warmer than the "Average" climate based on data for Strasbourg. And since mfrs typically don't publish figures for the optional "Warmer" (Athens) climate I can't easily correct for this with e.g. a weighted average of the two.

 

1 hour ago, JamesPa said:

I posted a few months ago a weather compensation calculator which estimates how much difference various WC algorithms make, using a bin analysis of OATs and published cop.  You could dial some figures into this if it helps.  Very roughly (from memory), using Mitsubishi COP figures, it's 25% at ft55, 15% at ft 45 and 5% at ft35.

 

Very much the kind of result I was anticipating, do you by chance still have a link to that thread @JamesPa?

 

Several possible installers have now converged on an overall heat loss of about 12.5 - 13.5 kW at -0.2 to -1.4C, my principal concern now is defining the correct rad surface area vs flow temp. In the past I have already more than doubled the rads in the living room and master bedroom, and I am planning to change the rads on landing and two other bedrooms and fit one in the en-suite. But I am not going for the gold-plated solution of installing a further five in utility, kitchen and shower room as suggested by one firm. If a subsequent buyer re-does the kitchen and rips out the AGA then fitting an extra rad will be small beer, the pipework is already there on the other side of the wall.

 

And it is not a disaster if I do not get them totally correct first time provided the HP is big enough to start with.

[DanDee]

12 minutes ago, sharpener said:

 

Very much the kind of result I was anticipating, do you by chance still have a link to that thread @JamesPa?

 

 

[SteamyTea]

@DanDee

I fully appreciate it is work in progress, but do you have lots of data? The current chart shows half the data points for non WC overall, and almost a third below 10⁰C.

[JamesPa]

1 hour ago, sharpener said:

Very much the kind of result I was anticipating, do you by chance still have a link to that thread @JamesPa?

@DanDeehas posted a link to the thread, an 'improved' model appears later in the thread here.  It was mostly built to test how important getting the WC curve 'right' is but can be pressed into service for other estimates

 

[sharpener]

@SteamyTea / @JamesPa can either of you suggest a ready source for temperature distribution by # of days in 1 deg bins for Plymouth?

 

I have found some data in the CIBSE Guide A but it is concentrated around and below freezing and does not cover the warmer days where the HP will I hope spend most of its time.

[SteamyTea]

7 hours ago, sharpener said:

@SteamyTea / @JamesPa can either of you suggest a ready source for temperature distribution by # of days in 1 deg bins for Plymouth?

 

I have found some data in the CIBSE Guide A but it is concentrated around and below freezing and does not cover the warmer days where the HP will I hope spend most of its time

The Met Office has some historical data. The nearest station will be close enough.

 

There is also Degreedays

 

https://www.degreedays.net/

Edited June 8, 2023 by SteamyTea

[JamesPa]

7 hours ago, sharpener said:

@SteamyTea / @JamesPa can either of you suggest a ready source for temperature distribution by # of days in 1 deg bins for Plymouth?

 

I have found some data in the CIBSE Guide A but it is concentrated around and below freezing and does not cover the warmer days where the HP will I hope spend most of its time.

This is where it gets difficult.  I used daily average temperature for the nearest weather station, this data is freely available.  This is going to introduce some error in the calculation the magnitude of which depends on how quickly your house reacts to external temperature changes. Mine is quite slow to react, taking a day or thereabouts to 'stabilise' so I was happy that average temperature was good enough for what I was trying to explore.  Degree days are good for some purposes, but probably not this one.  Hourly data is available by following links from the met office to another site (I can't recall where exactly).  You have to register and can then download about a month's worth for free, any more you have to pay.  Im not sure if this is a daily or all time limit.

 

[S2D2]

https://www.visualcrossing.com/ is quite good, also has hourly data. There's a daily limit to the amount of data downloadable for free accounts.

[sharpener]

9 hours ago, S2D2 said:

https://www.visualcrossing.com/ is quite good, also has hourly data. There's a daily limit to the amount of data downloadable for free accounts.

 

It's potentially all there, but AFAICS I would need to download the average temp for a whole year (ideally 5 years) and then sort the numbers into 1 degree bins to fit into @JamesPa's spreadsheet, is there a better way?

[JamesPa]

5 minutes ago, sharpener said:

 

It's potentially all there, but AFAICS I would need to download the average temp for a whole year (ideally 5 years) and then sort the numbers into 1 degree bins to fit into @JamesPa's spreadsheet, is there a better way?

That's what I did (I think for only 2 years). 

 

Sorting into bins is easy.  I cant remember exactly how I did it but round(t+0.5,0) rounds to the nearest whole number and then a pivot table on the result (to count) is one method, or you could use a series of countif's

Edited June 8, 2023 by JamesPa

[SteamyTea]

2 hours ago, sharpener said:

is there a better way

If you know the mean temperature, and the standard deviation (use 6⁰C if you don't) that will be close enough.