opinion on best way to fix a factory made Powerwall

[gonmag8]

Hi all.

A friend asked me to see what is up with his Powerwall.
He says he has it for about 18 months and a few months ago it throws an error code and wont start.
So, took my tools and got the lid out....
Found, 2 big battery modules and a victron inverter, the battery modules had stickers saying LifePo4 48V 200Ah ... so far so good... I might be able to work with this ...
Thought about accessing the BMS but the company that sells these battery's don't give BMS access to the user, as such I decided to have a look at the battery modules.
One of the big modules measured 26.32v and the other 0V, so I assumed the BMS must be on the module that measures 0 and the other one are just 8 raw cells measuring 3.29v each … so far so good.
Opened the module that measured 0V and found a BMS, Relay, Fuse, etc and 2 sub-modules of 4 cells.
1st sub-module measures 13.16v , this is good (4 cells 3.29each) 2nd sub-module measures 10.43V … no good....
Tracked the BMS wires of that sub-module to measure the cells individually and found out that 3 cells are at 3.29 and one at 0.56V.

Guess the problem has been found then... one cell died and the BMS refuses to start the battery … fair enough …
I am surprised that one cell is dead. This being a factory made power wall and supposedly using grade A cells, how is one dead in about 18 months ?

Now, to try and fix this battery …
Do you think its worth charging that cell individually ? In my limited experience I would say that cell is no good anymore (at least not as good as the others) and even if it takes charge it will always be the limiting factor for the 16s pack.
What do you guys think ?

Would it be an idea to buy a new cell … top balance all the cells and try to start the battery again ? or buy a new BMS and make a 15s pack ? other ideas ?

the 4 cell sub-modules are all wrapped together in heat shrink and I have not butchered them to see if the terminals are welded or screwed but I suspect they are welded and it might be a pain to salvage the 3 cells that seem to still be good.

Give me your 2 cents.
Thank you all.

 

[sunshine_eggo]

Do you think its worth charging that cell individually ? In my limited experience I would say that cell is no good anymore (at least not as good as the others) and even if it takes charge it will always be the limiting factor for the 16s pack.
What do you guys think ?

Hell no.

Would it be an idea to buy a new cell … top balance all the cells and try to start the battery again ? or buy a new BMS and make a 15s pack ? other ideas ?

the 4 cell sub-modules are all wrapped together in heat shrink and I have not butchered them to see if the terminals are welded or screwed but I suspect they are welded and it might be a pain to salvage the 3 cells that seem to still be good.

Give me your 2 cents.
Thank you all.

You've listed several viable solutions. As long as you engage in best practices and ensure that all cells in series are the same spec or similar spec and perform within 10% of each other (capacity, internal resistance in particular), it's probably worth pursuing.

Do the BMS connect in any way to the inverter?

Is there a Victron GX device?

I would convey to your friend that he's lucky to have a friend willing to tackle it. Even more so that you have a clue about this stuff... :D

 

[gonmag8]

Thanks for your reply, it is reassuring.

There is a small victron box under the main invertor, I cant remember if it says GX on it or not. It does however have an Ethernet cable conecting it to the bms.

 

[sunshine_eggo]

Thanks for your reply, it is reassuring.

There is a small victron box under the main invertor, I cant remember if it says GX on it or not. It does however have an Ethernet cable conecting it to the bms.

This?



or:



It could also just be a custom Raspberry Pi running Victron Venus OS.

That does complicate things. You'll likely need to retain the existing BMS and replace the cell. The system may not work without Victron BMS integration.

There might also be a way to access the system from the GX device.

 

[gonmag8]

Said box is coloured victron blue but can't recall how it looks like... Will have to drop by and take a pic

I can't seem to find 200Ah cells but gwl has 100Ah ones so i can get 2 of them in parallel.
Will have to butcher the sub-pack of 4 cells soon to see what can be done ?