Suggestions for use of 24S (88V) modules for grid storage

[CyberBill]

I've come across a source for some large battery modules used in busses. 48 cells arranged as 2P24S, 88.7VDC nominal. 7.6kWh each. They weigh 140 pounds each with integrated BMS and cooling. It's looking like I'll be able to get at least 6 of them, maybe more, and might be able to get a trickle of them over time.

I have been planning a grid-tied battery storage system for a little while, and my plan was to use an EG4 GridBOSS and a FlexBOSS21 with some kind of 48V battery, but this kind of throws a wrench into that, as it (like other 48V battery inverters) only supports up to 60V. So I'm looking for some advice/recommendations. I know of at least two options:

Option 1: Open the modules, reconfigure for 16S. Either do 3P16S within each module, or 3x 2P16S out of each 2 modules. I'd have to open the pack first to see which would be more feasible. I'd also probably need to buy a 48v BMS for each module.

Option 2: Find an inverter that supports 24S packs. I really like the EG4 setup (gateway + multiple inverters), so in addition to finding an inverter I'd also need a compatible gateway.

What do you think?

 

[Hedges]

There are high-voltage inverters which support select HV batteries such as those from BYD.
There is a project with code on Git Hub to take EV packs and emulate supported HV inverter batteries.

R

Post in thread 'Seeking Advice: Low Voltage vs. High Voltage Batteries for Home Energy Storage'

Jun 18, 2024

It makes a lot of sense to me that the trend of ESS voltage increasing over time (12 -> 24 -> 48 -> ??) will continue. SMA's upcoming Sunny Boy smart energy hybrid inverter (just shipped this past week to SMA's distributors) only works with HV batteries (90v+ IIRC), which is really telling. But I'm kinda bummed, bc I really want to use that inverter but SMA said its currently only compatible with BYD HV battery systems, which are like $600/kWh, which is way more than some of the 48v ESS options out there (EG4/Goodwe/etc ~$300/kWh or less). Anyone know of any HV (>48) battery systems...

• Rational

• 
• 

maybe your integrated BMS is supported. I would think the bus uses a higher voltage configuration. Probably also has an additional box that talks to the modules and has contactors to disconnect.

What I know of people getting working so far is complete EV modules with complete BMS.

Alternatively, people do cut apart modules like you're thinking.
Instead of cutting, I've had the idea of making +24V and -12V batteries (relative to ground). 12V can be useful to backup telecom. But this is way more Ah than you're likely to need.

 

[CyberBill]

maybe your integrated BMS is supported. I would think the bus uses a higher voltage configuration. Probably also has an additional box that talks to the modules and has contactors to disconnect.

Yes, I believe it uses 42 of these modules, for 320kWh total. Not sure what the voltage is when complete.

I did some searching of the model number and make of the batteries and couldn't find any references to them online other than one page. They are from XALT Energy, called XPAND 76P. Best I can tell they use in-house BMSs on each module and then an additional control box that they plug into.

 

[CyberBill]

Following up - I have acquired 10 of these modules, ~76kWh worth!

I only had a short look at one of the modules outside of the crate and wasn't able to grab a pic, but I'll get one this week. Had to also buy a hydraulic table cart so I can move them around.

The modules are all faulty in some way and were being scrapped, which is why I was able to get them. Some have markings that say the problem (broken studs, for example) but most just say 'BAD' and likely have a dead cell. The BMSs on each module seem like they are just a big array of resistors, so those are getting tossed no matter what.

A slight hiccup that I wasn't thinking about was that I'm used to doing 48V batteries with LFP cells, but these are NMC/G. LFP cells are 2.5V (empty) -> 3.7V (full) compared to NMC cells that are 3.0V - 4.2V. With LFP you string 16 together in series and end up 40V -> 59.2V. But with NMC if you put 16 together you would be way too high voltage. So I need to do 14 in series instead, ending up with 42V -> 58.8V, which is comfortably within the FlexBOSS21's 40-60V range. Of course then 14S is a very weird number when each module is 48 cells. Common denominator between 14 and 48 is 336 cells (7 modules). :D

At 14S I will have a total size of ~34 parallel strings and targeting 24kW (240V @ 100A), which works out to 15A (per string? per cell? same thing). Each string would be 48V, 45.8Ah, 2.2kWh, and 1/3C at max. A very comfortable draw rate, lots of overhead.

I'll have a better idea of whether I should target 1P, 2P, or 3P once I open up a module and see how it is constructed. 1P is nice because it would be under 50 pounds, easy to move by hand by an individual. 2P would be easy enough to drag around and lift with a hydraulic table or another person. 3P is a strain to move by an individual.

For mounting I am going to weld together some angle steel into a rack, so each module can slide in and out. It will be fixed to the wall in the corner of the shop with cement backerboard for fire safety. I have very tall ceilings, so I could easily stack them 12' up.

 

[CyberBill]

Had a few minutes today to crack open one of the crates and see what these batteries are made of!

This pack is 2P24S using pouch cells, the tabs come out at the top and there are copper bus bars between the cell groups. The cell tabs are ultrasonic welded to the copper bars, which makes them a pain to remove. As far as I know, the only way is to cut them off - either cutting the tab to get to the raw cells, or cutting at the bus bar if you're OK with keeping them as 2P. Each bus bar has a cell tap connected to a flexible ribbon cable that heads up to the two BMS boards on the front of the pack.

The pack is built using two identical sub-packs of 12 cell groups that normally has a bus bar/jumper in the middle of the pack to string them together to 24S. At the 'back' of the pack is an Eaton 400A fuse. Connecting the packs up to the front, where the main battery studs are located are some bus bars. They are, in my opinion, really nicely made. Nickel plated copper with (looks like) powder coating as insulation. As I said the front of the pack has two (seemingly identical) BMS boards. Each one contains three connectors - one that goes to the cell taps and two others were not occupied. I would guess that they are connected to the higher level controller in the bus.

Mechanically the pack appears to use cast aluminum end plates and a lot of injection molded plastic pieces. I didn't have time to figure out how the pack is kept together, though. It seems pretty similar to how other pouch cell packs are built - I've disassembled the Chevy Bolt EV pack and there are a lot of similarities.

Cooling wise, there is integrated liquid cooling that appears to maybe run down the left side or the bottom. But I didn't get a chance to disassemble it far enough to see how it works. I doubt that I would be using it for my home battery storage system in any case.

The makeup of the pack here really pushes me to try and take one module and split it into two 2p12s packs, and then take a spare module and cut pairs of cell groups out of it. Then put the two together, giving me a 2P14S pack. I should have enough good working sub-packs to duplicate it at least 10 times, though I might need to make a few 'Frankenstein' packs if I find bad cells. Then find a >30A 14S BMS for each one, and stack them up.

I did a quick test of the cells in this module and all of them had an identical voltage at 3.57v.

 

[wpns]

Welded cells, I was afraid of that. Plus the fire hazard of NMC, yikes! Keep up the progress reports!

 

[EricBarbour]

FWIW these modules would be of some use to EV hobbyists. The internal BMS is a question mark, but it probably talks to a central system controller over CANBUS and there are people with experience hacking unknown CANBUS devices.

You might consider reposting to diyelectriccar.com.

Also, if you really want to make them into 2p14s packs for a 48v system, I would suggest cutting the bus bars and getting a suitable Daly or other 48v BMS. These EV packs with internal BMSes are notorious for being poorly documented and/or obsolete.

 

[Solar Guppy]

Don't use them.

NMC and pouch cells is possibly the worst combination possible for home ess. And since your likely going to dismiss this at least do any modifications outdoors as if you break or puncture a nmc pouch cell it will bust into flames and nothing can stop it from burning and spreading.

 

[CyberBill]

I agree with you all that pouch NMC cells would not be my first choice. BUT... they were free. Also, sitting in the same building is three other EVs using pouch NMC cells (MachE, Lightning, and Chevy Bolt) - and this will be installed in a shop, not in my home, so there is minimal risk to personal injury.

For sure during the phase of cutting things up, I'll do it on the cart in the shop where I can roll it right out onto gravel and it can't spread.

I started dreaming up a rack mount that I think should work - using 1" square tube to construct a 4-post rack, with 1" angle steel making rails. The sides of the rack will be covered in cement board, and on the rails will be cement board, effectively giving each battery module a little cement board house so any fire can't spread to other modules. The rails will be slightly angled so that coolant (if I decide to use it) will run out the back and not drip down onto all the modules.