16s6p 48v lifepo4 battery configuration

[drummerboy1864]

I know im a new user and might have some amature questions so I apologize. But I love the support on this forum from reading through past posts.
However reading through posts, I am unable to come up with a configuration for a 16s6p system on a single BMS. Im working with batrium and they have assured me I only need 16 monitors with a single watchmoncore bms if I parallel first, then series and put everything in 1 string.
However with these prismatic cells im not coming up with a good battery layout for how to make this work. and I realize this kind of setup has its drawbacks as well. Im using 230AH 3.2v prismatic CALB cells. I have 96 of them, and would love for it to be one big 70Kwh battery bank for my solar system.
Is this really possible to set these cells up in a 16s6p configuration on only 16 cell monitors and 1 BMS? or do I need to set up 6 strings with 6 BMSs. Or do I pick a happy medium of 3 battery banks of 16s2p? They are all grade A cells, so balancing I realize is difficult with this many, but will be easier with grade A cells.

Any advice is greatly apprechiated!

 

[time2roll]

Make 16 blocks of 6 cells each in parallel. Now you have 16x 3.2v cells with 1380 ah.

Arrange these blocks in series. Straight line, box them in a rectangle, place on different shelves. Just so they are electrically in series.

Connect BMS. Connect to system.

 

[MrM1]

Issue would seem to be with so many cells in a single parallel block, and one bms, the BMS will read those 6 cells as one single cell. Looks like a lot of potential for the BMS not catching a bad cell. Yes they're all in parallel so the 6 would kinda self balance. But 6 cells to a single bms lead seems pushing things for accurate monitoring.

 

[drummerboy1864]

Make 16 blocks of 6 cells each in parallel. Now you have 16x 3.2v cells with 1380 ah.

Arrange these blocks in series. Straight line, box them in a rectangle, place on different shelves. Just so they are electrically in series.

Connect BMS. Connect to system.

Yes this is kind of the arrangement I came up with as well. However.. I sent this is batrium and they said this was not recommended. Not sure why though?

 

[drummerboy1864]

Issue would seem to be with so many cells in a single parallel block, and one bms, the BMS will read those 6 cells as one single cell. Looks like a lot of potential for the BMS not catching a bad cell. Yes they're all in parallel so the 6 would kinda self balance. But 6 cells to a single bms lead seems pushing things for accurate monitoring.

and this is the downfall I figured. The BMS would not catch a bad cell within the block of 6. However.. my thoughts were, if one goes bad, the BMS is going to think its all bad right? cause they will all dip to the low voltage of the single bad cell, and therefore at that point, I can pull the block of 6 and do a top balance and see which is bad right?

 

[MrM1]

A lot off people do it. I'm sure you are right. It would probably be OK. It just depends on how you've designed things and what you want and want your comfortable with.

Pros
- costs
- only 1st bms
- fewer connections to manage
- fewer parts and fuses

Cons
- lessons precise monitoring
- lack of redundancy
- less fail safe. If one cell goes bad whole battery bank is down
- current handling limited to a single bms

For me I like at least 2 bms and 2 complete individual batteries in parallel . Can handle higher current and I can easily take a battery down for service without shutting down the system.

But also, doesn't batrium have "modes" or something to add on more cells without buying more individual BMSs?

 

[time2roll]

Yes this is kind of the arrangement I came up with as well. However.. I sent this is batrium and they said this was not recommended. Not sure why though?
View attachment 80294

Those blocks of cells do not look wired in series. Each block + should connect to the next block -
The end block - will go to the B- terminal on the BMS and the C- BMS terminal to the shunt.
The opposite end block will have an open + to go to a main distribution bus and fuse.

In the diagram shown it looks like all cells are parallel, no series.

 

[Rocketman]

I think you have four choices on a battery: The single Batrium can handle any of these arrangements.

1. 6p16s. One big battery each supercell has 6 cells in parallel- then all 16 supercells create the battery - one class T fuse at the battery positive.

2. 3p16s2p two batteries are made up of sixteen supercells - each supercell has three cells in parallel. You will have two class T fuses on each positive of each battery. You will need two Batrium K9’s to monitor.

3. 2p16s3p three batteries are made up of sixteen supercells - each supercell has two cells in parallel. You will have three class T fuses on each positive of each battery. You will need three Batrium K9’s to monitor.

4. 16s6p (I don’t know if this one is possible to link 6 Batrium K9’s). You will have 6 batteries and EVERY cell will be monitored. You will need 6 class T fuses and 6 K9’s.

There will be challenges in making sure when you pull a battery out of the bank - work on it- then put it back in - it should be close to the same state of charge as the rest of the batteries.

With one Batrium core you could monitor all of the above configurations, or you could have a Batrium core (and contactor or breaker) for each battery - probably only makes sense for the 3p16s2p battery.

You may want to order four extra cells - for backup cells and to make a 12v battery for use and practice.

 

[drummerboy1864]

I think you have four choices on a battery: The single Batrium can handle any of these arrangements.

1. 6p16s. One big battery each supercell has 6 cells in parallel- then all 16 supercells create the battery - one class T fuse at the battery positive.

2. 3p16s2p two batteries are made up of sixteen supercells - each supercell has three cells in parallel. You will have two class T fuses on each positive of each battery. You will need two Batrium K9’s to monitor.

3. 2p16s3p three batteries are made up of sixteen supercells - each supercell has two cells in parallel. You will have three class T fuses on each positive of each battery. You will need three Batrium K9’s to monitor.

4. 16s6p (I don’t know if this one is possible to link 6 Batrium K9’s). You will have 6 batteries and EVERY cell will be monitored. You will need 6 class T fuses and 6 K9’s.

There will be challenges in making sure when you pull a battery out of the bank - work on it- then put it back in - it should be close to the same state of charge as the rest of the batteries.

With one Batrium core you could monitor all of the above configurations, or you could have a Batrium core (and contactor or breaker) for each battery - probably only makes sense for the 3p16s2p battery.

You may want to order four extra cells - for backup cells and to make a 12v battery for use and practice.

This is really good information, I apprechiate it!. And your right I did also order an extra 4 cells for that reason, so I have 100 in my possession. I think i may go with the 3p16s2p, I also have two class T fuses sitting in my basement for this setup.
I like the idea of having two batteries in case one goes down, but im not sure how I can run that on just one batrium core? wouldnt I need one for each 16s battery?
Although note, I am using the newer M8 BlockMons not the K9 setup. It was more recommended for lifepo4 cells and they are much easier to setup. they do allow all 96 batteries to have a blockmon on it if needed, but its much more expensive that way. The blockmons connect to the watchmon core bypassing the need for a K9, but I do believe this would mean I need two watchmon cores. Might be better that way anyways

 

[drummerboy1864]

Those blocks of cells do not look wired in series. Each block + should connect to the next block -
The end block - will go to the B- terminal on the BMS and the C- BMS terminal to the shunt.
The opposite end block will have an open + to go to a main distribution bus and fuse.

In the diagram shown it looks like all cells are parallel, no series.

lol your right, each block is not in series. thats what I get for my late night diagrams. Ill re-write this up, I apprechiate the catch!

 

[jtvt]

With parallel first you don't have much for balancing capability.

I use 4 separate 16s strings with 64 blockmons with one watchmon 4. Works fine.

 

[drummerboy1864]

With parallel first you don't have much for balancing capability.

I use 4 separate 16s strings with 64 blockmons with one watchmon 4. Works fine.

Its great to know your doing this. My original design was like this 96 blockmons in seperate strings. But its just quite a bit more cost, if things get complicated I may go this route lol the individual cell monitoring would be nice. Im not as worried with the balance, If I have 3 or 6 cells in parallel, and they are all grade A and top balanced, they should stay in sync decently

 

[jtvt]

Its great to know your doing this. My original design was like this 96 blockmons in seperate strings. But its just quite a bit more cost, if things get complicated I may go this route lol the individual cell monitoring would be nice. Im not as worried with the balance, If I have 3 or 6 cells in parallel, and they are all grade A and top balanced, they should stay in sync decently

I went this route for various reasons which have been beat back and forth on this site multiple times (easier to find bad cell, can disconnect one string and still have working battery, simple bus bar setup, etc, etc).

I would get some extra blockmons. I've had two that flashed zero volts while balancing which causes a critical threshold fault and system disconnect. There are some posts indicating that the cause might be essentially a timing bug in the blockmons, a problem with heat causing bad readings, just bad hardware, or the wiring and not the actual blockmon. The jury is out to actual cause. I replaced the blockmons and left the wiring, so I don't think It's always wiring. I scaled back the bypass temp limit and haven't had another problem. But anyway, I have 8 spares on hand because they do have problems.

 

[drummerboy1864]

Ill definitely get some extra block mons, I just ordered a second batrium bms setup with extra blockmons. So Ill plan on having two sets of 3p16s batteries.
each battery is going to look like this, anyone see any issues with this setup? 3 cells per blockmon, I dont think is too bad for monitoring. better than a 6p pack at least.

 

[Rocketman]

That shape may not be the best choice.

It may be tough to get some compression on that shape of a battery (not saying it can’t be done).

A 2x8 arraignment may be better - or a 4x4.
Easier to compress.
The main positive and negative end up near each other.
For the blockmons - you won’t need to extend one wire a lot (because starting and ending points are near each other.

 

[x4dow]

i would also worry with the bottom cells being crushed under the weight

 

[drummerboy1864]

That shape may not be the best choice.

It may be tough to get some compression on that shape of a battery (not saying it can’t be done).

A 2x8 arraignment may be better - or a 4x4.
Easier to compress.
The main positive and negative end up near each other.
For the blockmons - you won’t need to extend one wire a lot (because starting and ending points are near each other.

as far as it getting crushed under the weight, Ill probably end up doing it horizontally anyways.
Good point on the compression. Im not 100% convinced I should compress, but I know thats an age old debate. I would probably go 2x8 arrangement, each battery on its own shelf with a box to the side of it for the BMS on dim rails. a 2x8 arrangement would fit almost perfectly on my 53"x17" shelves. (https://www.menards.com/main/storag...-72h-3-shelf-rack/1994062/p-1444422057172.htm)

Im not sure if I should compress or how much. But what Im thinking... is because there will only be about a 2.5 inch gap on the sides of the shelf, I may just add two squares of 5/8" plywood on each side. this would not really compress it, but it would cause compression as the batteries bulge, perhapes extending life of the cells as well as the busbars (which will be flexible). Is this a dumb idea?

 

[jtvt]

Is the bus bar sized for the current between each 3p?

 

[time2roll]

What supplier puts those terminals on?

 

[drummerboy1864]

Is the bus bar sized for the current between each 3p?

bus bars are 3mm, and are rated for 400a continuous.
each 3p battery will go to a 1000A shut (i should have used the 500a but I already have the 1000a) and combine into one string via 600A terminal bar into a single 500MCM wire where it will travel to the inverters and split between three 6kw inverters, probably on 2awg wire (not sure I can fit larger than that on the inverter terminals)
continuous current from inverters would not exceed 375amps @48v