BMS an unnecessary waste of money and energy?

[Abe53]

Could someone point me to an authorative (sp?) test to show the value of BMS? I run several battery systems and have yet to see negative consequences of cell imbalance. The constant drawing down of high Voltage cells and pumping up of low voltage ones must waste energy? My cells each seem to find their happy place until they give up the ghost, at which time I just replace the dead cell. What possibly can go wrong, go wrong, go wrong....

[tetrasection]

For simplicity's sake say you have 4x 12V batteries that are supposed to be charged at 14V each, which is 56V in total.

Ideally your batteries would look like this: 14V 14V 14V 14V

If you apply a 56V charge to batteries that are unbalanced you might get something like this: 13.5V 14V 13.5V 15V

That 15V battery will fail very quickly.

Edited June 3, 2021 by tetrasection

[YellowTapemeasure]

That's a pretty good analogy @87 Dream, thank you.

[Energy-Jason]

Very cool.

[wolfandy]

12 hours ago, Abe53 said:

show the value of BMS

The see post by @Achmat here about the Pylon BMS forcing grid charge when his SOC got down to 8%. I'd say that a BMS that overrides inverter settings to prevent damage to the cells is definitely value

[ALP]

Is there a module or control unit that can read and understand different BMS's of different makes.

This to add different makes of BMS / batteries to one inverter.

[Bobster]

I understand then technical reasoning for having a BMS, but it does seem to me that when it comes to assessing the state of a system there are two versions of the truth

1) The truth

2) What your BMS says

I have a recurring problem where my SOC drops below 40% then starts charging again. The BMS reports a linear charge to about 72% then it jumps to 100%. I'm told, and I believe, that the 100% can be relied upon. Fine, but then what was the SOC the rest of the time and given that my system is set up to start drawing from grid when SOC gets down to 40% (so 40% should always be available in case of an outage), was I drawing from grid unnecessarily?

Recently I was asked to take part in a beta test of a new, locally designed, BMS. OK... so nothing else changes, but it showed me using almost no power over night. OK... maybe the original BMS was misreporting, but a rough calculation based on average load from midnight to 6am  showed that the original BMS was in the ballpark. So the new BMS was under reporting hugely. Could my batteries have run nearly flat but the BMS be reporting plenty of charge available? I don't know. The other thing was that reported SOC NEVER showed 100%. I was told that it would take a day or 2 for the BMS to "learn" my batteries and it would eventually synchronise. But this didn't happen in 4 days and I got fed up of waking up to being told that I had 85+ % SOC when I knew that couldn't be true.

OK.... this was a test. The unit was in beta. I understand what that means, and the manufacturers were modifying the software on the BMS. I'm in IT. I know that things are seldom right the first time and thorough testing is prudent.

But my point is that the perceived (by me) state of the system varied hugely according to which BMS was hooked up. I am now back on the original BMS (BYD) but I still have that problem of the SOC being underreported if the system has discharged to the 40% mark.

So why is it so difficult to have something that actually reports the state of the system to you? I understand that doing that EXACTLY is difficult, but if it's showing me a steady climb up to 70% and then 5 minutes later it is 100% then that's not being not quite exactly accurate, that's not even being in the ballpark. Same with the new system that showed me using very little power overnight - again that was not even in the ballpark.

 

[YellowTapemeasure]

8 minutes ago, Bobster said:

I have a recurring problem where my SOC drops below 40% then starts charging again. The BMS reports a linear charge to about 72% then it jumps to 100%. I'm told, and I believe, that the 100% can be relied upon.

 

Pylontech US3000Bs have a similar charging "signature", fairly linear charge rate up top 88% and then it sits at 88% for quite a while, and the quickly shoots to 100%. The US3000Cs are reported to have a similar "signature", just different values.

With an integrated manufacturer supplied BMS, one would hope that they know what's best for the cells, and that it was tested under all sorts of different atmospheric conditions (temperature, air pressure, humidity etc.) because their warranty is dependent on it. But this kind of testing would probably not be be economically viable, and as you have pointed out, we do see mistakes being made, and we also know that no software / firmware code is ever perfect, that would be impossible.

[Bobster]

14 minutes ago, YellowTapemeasure said:

Pylontech US3000Bs have a similar charging "signature", fairly linear charge rate up top 88% and then it sits at 88% for quite a while, and the quickly shoots to 100%. The US3000Cs are reported to have a similar "signature", just different values.

With an integrated manufacturer supplied BMS, one would hope that they know what's best for the cells, and that it was tested under all sorts of different atmospheric conditions (temperature, air pressure, humidity etc.) because their warranty is dependent on it. But this kind of testing would probably not be be economically viable, and as you have pointed out, we do see mistakes being made, and we also know that no software / firmware code is ever perfect, that would be impossible.

I am not alone! 
 

[Abe53]

Hi Bobster, I also have a Victron - Pylontech system and the Venus GX seems to manage it well.

However, my powerwall consist of 1802 x 18650 lithium-ion cells, (factor that, how many strings of how many cells  

On this system I use a manual battery management system. Each string can me quickly isolated and tested in situ under 1C load. A string under performing is isolated and cells individually tested 10 at a time on my bench tester. Cells under performing are then replaced, they only cost R1 apiece. To avoid the dreaded overcharge, solar charges to max 4V per cell. (Specs say max 4.25V or such) The powerwall feeds a UPS which switches off when batteries are at about 50% soc. Since batteries are so cheap I need never use full capacity. So far so good!

[Tim003]

18 hours ago, Abe53 said:

Cells under performing are then replaced, they only cost R1 apiece. 

Where do you get them so cheaply? 

[stok]

On 2021/06/04 at 2:11 PM, Abe53 said:

Hi Bobster, I also have a Victron - Pylontech system and the Venus GX seems to manage it well.

However, my powerwall consist of 1802 x 18650 lithium-ion cells, (factor that, how many strings of how many cells  

On this system I use a manual battery management system. Each string can me quickly isolated and tested in situ under 1C load. A string under performing is isolated and cells individually tested 10 at a time on my bench tester. Cells under performing are then replaced, they only cost R1 apiece. To avoid the dreaded overcharge, solar charges to max 4V per cell. (Specs say max 4.25V or such) The powerwall feeds a UPS which switches off when batteries are at about 50% soc. Since batteries are so cheap I need never use full capacity. So far so good!

Abe53, please share you source of 18650 cells with the rest of us???