diy solar

diy solar

Audi EV parallel battery sketch V2.0, hand drawn (oh boy) safety recommendations and opinions please!

cyberfed

New Member
Joined
Mar 23, 2024
Messages
44
Location
Florida
Hello bright minds,

I am still pondering the full build of the Audi 13S battery packs I got that are brand new. Yes 13S isn't ideal I know but I'm not going to tear open the factory metal box they are in with the intent of trying to reconfigure things, I'll just end up breaking things. So 13S it is. Yes these are the evil blow up bombs NMC chemistry. The box (besides the factory one) being designed is being designed to survive massive energy release and over 3000F of heat protection in multiple layers, with exhaust systems. That's going to be a heck of a build. For now I want to get the fundamentals of the parallel setup right before doing anything.

See the drawing and my terrible penmanship. The plan is one JK bms (already tested on all three batteries) per battery pack. A Class T fuse on the main POS of each battery pack going up to the busbar rated for 1000V. From there there's only going to be one load connection out, I added another class T on the POS after the busbar, followed by a disconnect switch. Then I'm no sure if I need anything more before the inverter. Load = 50-70watts max (yes virtually nothing I know).

Questions: You can see some notes at the bottom I scribbled.

1. Since JK does active balancing and each pack has its own BMS, in a parallel setup...what is the best way to handle this for proper battery stability? Only one BMS doing active monitoring? All of them? None of them?

2. Best way to charge, should I charge each pack independently and and then reconnect them after validating the voltages are almost identical so there isn't a surge of energy transfer? Or can I charge in parallel essentially replacing the inverter for a SCC and all the safety bits that go with that.

3. Other safety measures. What am I missing? What else would you recommend I add? Should I add a disconnect switch for each POS before the busbar as well?

Each 13S pack has the following properties:
  • Capacity: 37Ah
  • Energy: 1.78kWh
So I know I can google this and find out but I figured I will ask here too. What class T fuse size should I use since each pack is 37Ah? Add them all up (round to 40 so 120a and add a safety margin so like 130-140a fuses? What's the "right" way to size them in a parallel setup.

All this info will come in handy because I have 4 Renogy brand new batteries I bought, 24V/25ah each and they are going in parallel. They have their own internal BMS, which is nice and all and it says they balance themselves but I still plan to follow this similar type footprint. Is it allowed or safe to add a BMS to batteries like that have internal BMSs? Just be to be able to monitor each battery? I have a 500V shunt from them but I didn't know if there is a better way for me to be able to see how each battery is doing. I dont just want to throw some fuses on them and wire em up that scares me not being to "see" how each battery is doing. How is that normally handled??

Keep in mind this are "portable" packs, they do not stay connected to solar panels or SCCs, they are in boxes that have wheels so they can be placed in safe locations to power certain parts of the house. Mostly I've been building 16S LFP packs in steel enclosures right now, but these I'm kinda on my own in terms of design. I've read the battery documentation from Renogy it actually says the batteries can be paralleled up to 8 (I only have 4) and NO series connections allowed. Curious of what the community would design in that situation.

I appreciate any and all feed back on design and safety measures needed. These Audi batteries are the last of NMC batteries and the case I have to build is going to have to be ridiculous to protect myself in the worst case scenario. It's going to involve a lot of thick steel vent holes with proper handling of the venting, I'm going to add sensors from microcontrollers I program to give me warnings ahead of time for temp and VOCs, the BMS's will be set very conservatively for temp shutdown and other things. Thermal insulating fireglass will be used, CellblockEX pellets will be automatically setup in a manner to douse each pack in the event they heat up too much.

Everything will be thick steel, firebricks and the like. Fans will be used to remove exhaust as well as keep BMS's and other electrical parts cool and they will have a seperation barrier betwen the batteries and the busbars/bms's ect..and with multiple layers of protection. Fire extinguishers designed for Lithium fires have been purchased along with "worst case" fire supression canisters you throw at the fire and it goes off almost like a tear gas can the police use to smother, cool, and coat the batteries to extinguish them.

Honestly I'm not looking forward to the box build its going to be a huge PITA, expensive, and heavy (though it'll be on wheels).
Batteries will be charged conservitively and same with dishcharge. They are capable (video is out there to prove it) they can do about 150A continous (minor sag) when they are fully charged dropping slowly down to about 100A at shutoff (3V) and remaining just slightly warm). Now I would NEVER pull that kinda amperage. Like said my loads are stupid small under 100W. Routine checks on the IR (with quality expensive tools) will be performed to ensure batteries are staying within spec and healthy. As always I appreciate in advance your support, opinions and advice. This about safety and best to set that up for a worst case scenario. I don't know if I mentioned but the battery packs will be seperated by thick steel, so all three are not together so to speak. The hope there is, if one decides to become a bomb it doesn't take out the others with it. If all goes correctly it may start to vent, the battery will be doused automatically with CellBlockEX (designed to smother and put out NMC fires) and hopefully that stops it, if not the thick steel exhaust and other measures I discussed will serve as containment. I hope. I know someone is going to say just put them far away from your house as possible. Not possible. This is FL where its hot as hell pushing high 90's today and I have no where to put them in my yard I rent. If I did put em outside now I have to worry about adding weather (heavy rain) protection and it spirals. I understand the dangers, I've seen it, but also if done conservitive and safely they can be quite reliable and safe. I've used NMC for other things many times they are great. But yes they are little bombs, in this case 3 medium bombs if combined, one giant bomb. Hence looking for the best way to protect my self from a worst case (internal short/thermal runaway) situation.

Cheers,
Cyberfed
IMG_0128.jpg
 
Don't be wuss... I use 14S NMC. :p

You size the fuse for 1.25X the wire rating. The wire should be sized to handle the peak current.

1. Link in line #6 of my signature. Engage in best practices when paralleling, and it's a non-issue.

2. See #1

3. A per-battery disconnect isn't a bad idea. Even if not for safety problems, if you need to take one battery offline, it's crazy easy.
 
lol thanks @sunshine_eggo. I think I'll stay a wuss and not crack open the factory sealed 13S pack :). I know that I have a low power Victron Inverter that will allow me to run it below 48V. This isn't going to be powering anything major. Yeah I decided on the disconnect switches for that very reason in case I want to pull one out of service ect.. plus more safety "just in case".

I do oversize my wires on all my builds and use high quality welding cable that is high temperature rated. I do follow the 1.25 rule for fusing as well. I use a hydraulic press to get nice solid cold welds to my ring terminals.

Two questions I still need advise on well 3. I think I may have asked above but that's a long post and even I don't want to re-read it. :)

#1 3BMS - active balancing in parallel. What do I do here? Leave all 3 BMS doing active balancing on each of their respective packs? Not sure how paralleling comes in since that tends to balance the packs (over time) naturally. Should I just have 1 BMS doing balancing and turn the other 2 off for balancing or??? Help!

#2 I've heard varying "best practices" for charging in parallel. I've heard its better to charge each pack individually, get them all to like within .01V of each other then reconnect as parallel. Is that the way to go? Or can I charge the whole set of 3 at once? pros/cons/ major do's/dont's?

Well now I can't remember the 3rd question so...yeah :)

Oh I saw your VRM neat! If I set mine up you would laugh its so pathetic. :)

Thanks in advance,
Cheers.
 
Last edited:
#1 & partial #2: The batteries in parallel are forced to the same total voltage, so they're working to the same absolute boundary conditions. The BMS will all work similarly on their own batteries. The key is to regularly achieve full charge to the BMS can top balance their respective batteries. This condition also limits deviations... If 1 battery voltage falls faster, it supports less than its fair share, so the other parallel batteries take on more.

Short answer: leave them on in parallel balancing individually. If you can limit it to a voltage range closer to the top end, then that's preferred.

#2: Best practices is more about input than output, i.e., you take the steps necessary to ensure the batteries experience the same resistance. This is typically done by using bus bars, same length cables, properly torqued connections, etc. The output tends to be what you get as a result of the quality of your input efforts. Effort done before the final configuration are going to be either undone or supported by the quality of the parallel scheme. Your scheme of same length cables to bus bars should minimize any issues.

NMC have a strong voltage to SoC correlation, so when you've achieved a given cell voltage with very little/zero current flow, you can be confident that you're at a given SoC.

The only reason I would charge them individually is to get them to within 0.2V of each other before connecting in parallel.

There is no such thing as a pathetic VRM. It's mere existence makes the world a better place.
 
Sorry for the delayed response. Thank you @sunshine_eggo! Yes I will be using the appropriate bus-bars and keeping the cable lengths equal. I'm actually in the process of paralleling another 3 different packs these are 8S 24V LifePo4's. I am top balancing each of them without a BMS and keeping a close watch on them once they hit 3.65V across each cell they will rest overnight. If the voltage delta is miniscule, then physically paralleling them will occur. Class T fuses, busbars, oversized cable, and disconnect switches will all be in play as well. I've got a few 24V inverters I wanna use so that's why for these I'm not going with 48V.

Thanks again!
 

diy solar

diy solar
Back
Top