What 282 kWh CATL LFP look like

[yabert]

Parallel BMS is easy enough.

I'm not a huge fan of BMS in parallel, but I trust physical busbars in parallel.
Here a 45Ah battery I transformed in a 90Ah with weld busbars 10 years ago for an EV.
The plan with those CALB is the same. One BMS for one giant 1368Ah 51V battery.

 

[yabert]

Well then, saw it.

OMG.... this idea of carbide metal saw do is way in my mind ?

 

[SparkyJJO]

I'm not a huge fan of BMS in parallel, but I trust physical busbars in parallel.

Why not?

 

[yabert]

Why not?

Reliability and cost.
It's not a hard no for this project, but generally I will prefer the reliability and the low cost of a busbar + screws or welds instead of the potentially problematic BMS (thousand parts/cost $$) + wires + connectors.

 

[SparkyJJO]

Um You need both bus bars AND BMS. Busbars to carry the current, BMS to protect the cells against over/under voltage. Running no BMS is strongly discouraged.

 

[yabert]

Um You need both bus bars AND BMS. Busbars to carry the current, BMS to protect the cells against over/under voltage. Running no BMS is strongly discouraged.

Of course. I'm not arguing against BMS here, I'm arguing 6 BMS vs 1 BMS.

 

[yabert]

I'm clearly not a fan to destroy this awesome piece of engineering... with a circular saw, but wow, this aluminum was so easy to cut.
It seem like I could try to cut this aluminum chill plate in six to have nice 11.7 kWh 16S packs ?

 

[SparkyJJO]

Of course. I'm not arguing against BMS here, I'm arguing 6 BMS vs 1 BMS.

What BMS can handle 96 cells in a 16S6P 48V configuration? I suppose some commercial grade stuff somewhere does.

 

[yabert]

What BMS can handle 96 cells in a 16S6P 48V configuration? I suppose some commercial grade stuff somewhere does.

That exactly the point. It will not handle 96 cells, but only 16 cells because I plan take six 228Ah cells and create one 1368Ah cell.

 

[labeeman]

Of course. I'm not arguing against BMS here, I'm arguing 6 BMS vs 1 BMS.

I have 20 BMS in parallel what is the problem with that each one only has to handle 1/20 of the load or charge?

 

[Skypower]

I'm clearly not a fan to destroy this awesome piece of engineering... with a circular saw, but wow, this aluminum was so easy to cut.
It seem like I could try to cut this aluminum chill plate in six to have nice 11.7 kWh 16S packs ?
View attachment 177343

A standard circular saw is a bit faster and may gum up the blade or wear out faster even on aluminum. Aluminum circular saw blades are cheaper than a metal cutting tool though.

 

[yabert]

I have 20 BMS in parallel what is the problem with that each one only has to handle 1/20 of the load or charge?

Nothing wrong with it. I simply prefer the reliability of non active components (busbars vs Fets, drivers, current measurement, etc).

 

[labeeman]

A standard circular saw is a bit faster and may gum up the blade or wear out faster even on aluminum. Aluminum circular saw blades are cheaper than a metal cutting tool though.

use a little lubricant like WD40 or kerosine on aluminum

 

[labeeman]

Nothing wrong with it. I simply prefer the reliability of non active components (busbars vs Fets, drivers, current measurement, etc).

I know of no BMS with nonactive components.

 

[yabert]

I know of no BMS with nonactive components.

Playing with cells/BMS since 15 years now, I've seen way to horror stories of BMS who destruc cells.
So, to me, minimum of BMS and I plan to add a monitor device who verified the job of the BMS is right.

 

[SparkyJJO]

So wait, you plan on doing 6S16P? 6 cells in parallel to each other, each set of 6 on a single BMS channel?

What BMS have you used over the years that were so bad?

 

[yabert]

So wait, you plan on doing 6S16P? 6 cells in parallel to each other, each set of 6 on a single BMS channel?

Yes, well 16S 6P. One big cells. Exactly the same as when LG chem, CATL, SK innovation or car manufacturers decide to connect cells in parallel in EV battery packs.

What BMS have you used over the years that were so bad?

Not only me, but friends, coworkers, members on forums.
It's not always related to the electronics stuff, some time is the programing/control and regularly it's user.
Still, to me:
BMS = Extra cost, more consumption, more prone to failure, extra wires/connections.
Busbar = Reliability at low cost.

 

[yabert]

First test with the saw done.
I'm amazed how well it cut in this chill plate
The fact that chill plate is almost empty to allow the cooling to flow result that all the wall have small thickness.
The other axe will be way more tricky.

 

[SparkyJJO]

Yes, well 16S 6P. One big cells. Exactly the same as when LG chem, CATL, SK innovation or manufacturers decide to connect cells in parallel in EV battery packs.

Not only me, but friends, coworkers, members on forums.
It's not always related to the electronics stuff, some time is the programing/control and regularly it's user.
Still, to me:
BMS = Extra cost, more consumption, more prone to failure, extra wires/connections.
Busbar = Reliability at low cost.

Well I guess the biggest difference is you're using true automotive grade cells that have already been matched. I wouldn't attempt that with the run of the mill so called "A" grade we normally get, partly because even if they were premium cells the odds of getting perfectly matched cells are next to nil and those variances may play out on undesirable ways with multiple cells in parallel.

Though flipside is, if you have multiple packs each with its own BMS and one BMS fails, you're only down one battery and can still operate off the rest. If all your cells rely on only one BMS and it fails, well now you're out everything (unless you have a spare BMS just sitting on standby).

 

[Skypower]

First test with the saw done.
I'm amazed how well it cut in this chill plate
The fact that chill plate is almost empty to allow the cooling to flow result that all the wall have small thickness.
The other axe will be way more tricky.View attachment 177435

You’ll be the expert soon. ?