BMV-712 Smart and the Charge Efficiency Factor

Question

question

gravely-headstone asked May 7, '22

BMV-712 Smart and the Charge Efficiency Factor

Hi,

I am new to Victron, having only recently acquired a small system ( Solar controller, BMV-712 and Volt/Temp sensor ).

My question relates to the Battery Monitor - BMV-712 Smart - and its Charge Efficiency Factor ( CEF ) setting.

I recently conducted the following test. I set the Battery Amp Hours to 100Ah ( it makes the maths easier ), and the CEF to 100%, and then changed the SOC to a level less than 100% - as it would be if I had used some battery power. I then monitored the SOC rate of increase as current went through the shunt to recharge the battery.

The results were what I expected. Every 1Ah of charge that went into my battery ( according to the BMV ) caused a 1% increase in SOC. That is what I would understand as 100% efficient.

I then changed the CEF to 50% and set the SOC to a level less that 100%, as above, then monitored the SOC rate of increase as the current went through the shunt. The results were NOT what I expected.

Instead, for every 1Ah that went in to the battery ( according to the BMV ), the SOC went up by 2%. This what I would understand as 200% efficiency - not 50% efficiency.

For 50% efficiency I would expect that for every 1Ah that I put into the battery the SOC would go up by 0.5% - because of heat losses, battery chemistry or other factors.

I am perplexed, because nothing is 100% efficient when it comes to charging batteries, or converting sunlight to power, so why would there be any reason to move the SOC towards 100% at 2X the rate that the amp hours are being put into the battery.

Am I perhaps not understanding the true meaning of % Efficiency as it is defined in the Northern Hemisphere ? Or is my BMV-712 getting confused because it is located South of the Equator, and is therefore hanging upside down ? ( But it could not be that, because we all know that the earth is flat ! :-) )

I would appreciate any input, opinion or words to set me straight.

TIA

BMV Battery Monitor

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Modifications Space

Node Red Space

Answer 1 · 2022-05-07T06:00:35Z

Alexandra answered · May 07, 2022 at 06:00 AM

@Gravely Headstone

Usually the Charge Efficiency is a known figure from the battery manufacturers. So on a datasheet somewhere it will be tucked in there by the charge/discharge specs.

For the most part CEF on lithiums can be affected by new or second life but not by much. Usually are around 96% to 99% efficient. Also affected by rate of charge and temperature.

The bmv manual has a section on it. Factors such as age on some battery chemistry affect this figure.

Anytime you alter a setting on the bmv you are altering its calculating algorithm. So for it to be accurate you also need to get it right. And at the end of it will synchronise to 100% SOC using the tail current. So during charge for the first cycle it will be inaccurate.

1

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Alexandra ♦ commented · May 07, 2022 at 06:02 AM

An example from my babies at home.

0 ·

img-20220507-080104.jpg (3.6 MiB)

Answer 2 · 2022-05-07T07:06:24Z

gravely-headstone answered · May 07, 2022 at 07:06 AM

Hi @Alexandra - thanks for your very quick response - but I think that you are missing my point.

My point is that since you can not create something from nothing in the real world, then you can not put one Ah worth of charge in your batteries and have the SOC go up by 2 Ah worth of charge - well, you can, as it is just a mathematical calculation, but it does not make real world sense.

I hastily drew up this crude chart.

The blue line ( to the right of the red area ) represents 100% - as it is in the BMV712 now, where 1Ah charged will raise the SOC by 1% ( in my example above ).

The green line represents 50% efficiency - one Ah charged raises the SOC by 0.5%. This is where I 50% should be, because I may well be ignorant.

The red area ( very roughly ) represents all other CEF settings from 51% to 99%, where 1 Ah charged will raise the SOC by a factor between 1 and 2. It has nothing to do with battery chemistry or whatever - it is simple maths and I would like to understand the reason for doing it this way.

The way it works now does not help someone like me with an old battery, which consumes far more amps in for each real life SOC increase ( the green line ).

Somebody out there must know the answer :-)

I hope.

1651905782566.png (18.7 KiB)

3

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Alexandra ♦ commented · May 07, 2022 at 09:11 AM

@Gravely Headstone

But you are trying to only look at one part of the sum. Battery charging is not linear.

It is a calculation so if incorrect variables are added incorrectly it will calculate incorrectly. (Every setting in Victron Connect is a variable).

What peukerts do you have programmed? Since every parameter set affects the calculation all the settings including tail current and charge detection time have an effect.

I realised how difficult this is to explain. Have tried a few times now. The BMV measures different to what is programmed the CEF and peukerts are tied together. It will think it need 10 amps to raise to certain level, but the CEF says it will absorp 5 amps but your peukerts is different it will assume a capacity change of sorts, almost as if there is degradation, and so the SOC will rise faster.

It will synchronise to the correct 100% once voltage and tail current have ben reached (but will only be correct if those are programmed correctly).

What battery do you have?

0 ·

gravely-headstone Alexandra ♦ commented · May 07, 2022 at 10:24 AM

@Alexandra Again - thank you for taking up your valuable time to put me on the correct track.

I should have stated in my original post that for the tests that I ran, Peukert was switched off because it was one variable too many. Once I have got my head around this other issue I will start playing with Peukert - presuming that I do not give up and stop wasting everyone's time.

In the few tests that I did with Peukert switched on, it did not move the data points below the blue line, which I what I was trying to achieve, and I am fairly confident that even if I ran a test with every Peukert parameter and every CEF parameter, none of them would produce a curve that was not above the 1:1 ( 100% efficiency ) line - ie they would all advance the SOC faster than the current going in.

But I am very happy to be proven wrong !

FWIW, my battery is a 15 year old 12V FLA 1024Ah - but it is not relevant for the purpose of this exercise. It is just a load for me to put charge into.

Anyway, we will probably remain at cross purposes if we continue, so thanks again for your time - and cross this one off your list :-)

0 ·

Alexandra ♦ gravely-headstone commented · May 07, 2022 at 03:49 PM

@Gravely Headstone

Not at all we are all here to learn. It is easy to set up a bmv if you know all the variables.

And you do not so are in a unique position (and possibly frustrating depending on the kind of person you are). Old FLA batteries would prove a challenge as they are unknown in capacity and charge efficiency unless you have a starting point like a datasheet.

There is a peukerts calculator available linked on this page. So to be honest you dont need to know the math the whole way.

I have been inspired to mess with mine and see what happens ;)

0 ·

Answer 3 · 2022-05-07T11:21:41Z

JohnC answered · May 07, 2022 at 11:21 AM

Hi @Gravely Headstone

Where are you getting this 'Ah' charged amount from? If you're counting back the 'Consumed Ah' then that's where the issue lies.

The BMV doesn't count Ah towards SOC because it's not a measure of energy. Add Voltage and it becomes Wh, then it is. And I suspect when charging, deductions from the Consumed Ah are actually 'concocted' somehow, so as to show a realistic figure. Bear in mind too that an Ah-In can't hold the same energy as an Ah-Out, as charge V's are invariably higher under charge than discharge.

Unfair!, unscientific!, you might say. Nah, a concession to consumerism. Ah itself is a empirical battery rating, how many A you can get from a batt down to an unholy V. Less scientific than Wh.

If you really want to test this, use the cumulative Charged Energy, but it only reports in blocks of 100Wh.

2

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

gravely-headstone commented · May 08, 2022 at 07:36 AM

Hi @JohnC - thanks for responding.

I will start off by saying that even from my current position of mainly being ignorant of facts, I ( respectfully, of course ) dispute your assertion that the BMV does not count Ah towards SOC.

It seems to me that the ONLY things that the BMV does to earn its keep is to measure current passing through the shunt in either direction, and distribute voltage and temperature information to the Solar Controller ( in my small system anyway ).

It must then process the Amps and Voltage information, and come up with the good stuff, including SOC, that it displays on its internal screen, and on my phone. Therefore even the charged energy must be derived from the current and voltage information that it measures, as does all the other information that it displays, because it does not have another source.

I would also be a bit p'ed off if I thought that it was not providing me with information that was not as close to reality as was possible - as I may be wanting to use that information for other purposes.

However, I do agree with you that Amps out can not equal Amps in, and I would wager that "Amps In" would be on the low side of the equation. Hence the reason for my original thesis - that I ( still ) can not understand how one "unit" of charge can produce a change in the SOC that is greater than one. And they are related, if only by mathematics, because a change in one brings about a change in the other.

0 ·

JohnC ♦ gravely-headstone commented · May 08, 2022 at 10:47 AM

@Gravely Headstone

All fine. But the BMV nowhere reports Ah charged. There's references to 'Consumed', 'Discharge' and 'Drawn'. But nothing applies to charging. And I contend that the charge credits applied to Ah are rigged to show realistic figures for Consumed. Else they'd drift excessively when not synched to 100% SOC. Whoever coded this did a helluva job too.

So I say it's useless using counted-back Ah figures to do anything useful with. Ah is for rough stuff anyway.

Just for fun, contemplate this too, my overall cumulative figures..

That's 80.5%, fla's, such is life. But I use a Charge Efficiency of either 90% or 91% (seasonal), only because I can't select anything in between. Coincidentally?, that's half of my losses.

I love these devices though, accurate and reproducible results when well tuned. But I'm not an Ah person, never look at it..

0 ·

1652005655054.png (22.6 KiB)

Answer 4 · 2022-05-07T12:14:26Z

Trevor Bird answered · May 07, 2022 at 12:14 PM

Hello @Gravely Headstone , I understand what you have done and completely agree that what you observe is the exact opposite of what you would expect.

I have adjusted this parameter in the past to improve tracking so I know it works the way we normally expect. It is possible the shunt is in backwards? I know it is a strange question but it is the only way I could see any charge /discharge efficiency the reciprocal of what is expected.

2

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

gravely-headstone commented · May 08, 2022 at 06:14 AM

Hi @Trevor Bird - I did ask myself the same question a while ago - so I checked physically - and it is installed correctly. Besides which, when I use the battery the Consumed Ah accumulates in the -ve direction, and when I charge the battery, the Consumed Ah accumulates in the +ve direction, so it can not be both wrong and right.

Thanks for responding.

0 ·

Trevor Bird gravely-headstone commented · May 08, 2022 at 10:37 AM

I figured that was a long shot. If you are making this post you haven’t put the shunt in backwards. I have no idea what is going on with that issue. It makes no sense to me.

0 ·

question