Opinions on Chevy Volt Battery part time BMS ?

[Salts]

I'm using a Chevy Volt battery as storage for off grid solar energy and am having a few design issues.

Its a 2014 battery that came from a vehicle with 24k on it. I've broken the battery down and reconfigured it into 8 individual 12 series modules. I think its called a 3p12s configuration because each plate has 3 lithium pouches in parallel. Each module is within 4mV of the others.

I'm going to be hooking up a couple of SMA Sunny Island 6048 inverters which is probably more than I need in the way of power. My well pump is 240 volts so whatever I went with needed to be able to power the water pump.

It is unfortunate that I can not break down the Chevy battery into something like an 18p14s configuration as I'm sure that would be the best way of doing this. Such is life I guess. So my plan is to charge the battery to 4.1 and discharge to 3.5. I'm guessing that will let me use about 10kw out of a 16.5kw pack yes? On the up side, the battery should last forever.

Anyhow, I'm suck on the BMS. Technically, I think I'm supposed to have 8 BMS's.. one for each series string right? But can you imagine how expensive that would be? I was in another forum where a member had paralleled his BMS wires between strings. I've seen this twice now, one guy did it with 2 strings, another did it with all eight.. In other words, the cell numbers in each string go from 1 to 12. All the BMS wires from Cell 1 together, all the BMS wires from Cell 2 together, and so on.

Was this some crazy idea that's going to go boom? Or could this work safely if the BMS lines were fused?

One more question: Lets assume the whole parallel BMS wires is a bad idea. I'm using my battery bank in a solar backup application where charge and discharge rates can be held to less than 1C or even 0.5C. the reality is it would probably be more like 0.2C. I'm looking at purchasing a Rec BMS because it has the proper CAN communication protocols to work with the SMA Sunny Island. The Rec BMS will have both a pre-charge unit and main battery contactor. Each 12s string will also have circuit breakers installed as well as a main breaker on the line going to the inverter. Would it be reasonable to use a single 12s BMS on one string at a time and just swap it from string to string? The individual modules have quick-disconnect orange plugs on them that would make it pretty easy to swap the BMS lines.

Thank you,

[Ampster]

That is a problem using a Chevy Volt pack. You really need 8 separate BMSs. The issue with trying to parallel Volt batteries is finding a way to connect the individual 3P buddy cells together. The connections need to be large enough to carry the eddy currents that will develope as the various packs charge and discharge. The existing BMS wires are too small for that.

With regard to CAN protocols do you mean the explicit messages to set voltage at the constant voltage phase and the current at which to stop charging during the CV phase?

[Salts]

That is a problem using a Chevy Volt pack. You really need 8 separate BMSs. The issue with trying to parallel Volt batteries is finding a way to connect the individual 3P buddy cells together. The connections need to be large enough to carry the eddy currents that will develope as the various packs charge and discharge. The existing BMS wires are too small for that.

With regard to CAN protocols do you mean the explicit messages to set voltage at the constant voltage phase and the current at which to stop charging during the CV phase?

Hi Ampster, thanks for responding.

At the moment, I do not know what the Sunny Island is expecting from the Lithium BMS. I was surprised to find out that it would even bother to communicate with it since the Sunny Island has all its own measurement controls built into it. The Sunny Island has high voltage and low voltage settings for safety and charging, a current shunt sensor port, battery temperature port, etc.

I have a message out to SMA about the CAN BUS protocols and they will get back with me.

The Sunny Island operating manual leaves a lot to be desired. Their instructions are very clear in some of the easy things, and very cryptic in others. For example, under the "Interface For External Communications" heading, they say "You will find a detailed wiring diagram in the communication device manual, the software or on the Internet at www.SMA-America.com.".

Kind of strange that no where on their website or a google search is there any SMA Communication device manual" that I can find.. and I have no idea what software they are talking about. They seem to have a lot of tech videos and documentations for the European version of the Sunny Island.. The 6.0H, models and such, but very little for the US version which has a different front panel. What's that all about?

So anyhow, I have my Chevy Volt batteries, they're mounted in a 4ft x 3ft x 20 inch deep Hoffman steel enclosure (that's my fire box in case they go all 4th of July on me).. and I have some rather hefty 1/4 x 2 inch copper bars, DC circuit breakers, and lots of 2/0 battery cable.

Now I just need to figure out how to keep the batteries relatively safe without breaking the bank. Looking at Rec BMS, Zeva (master with slaves), Orion (very expensive), and a few other non-Chinese BMS's. I bought an entire unmolested battery so I still have the original BMS modules and associated cabling.

Thanks for your help,

[tom rickard]

I can help a bit. I use a Sunny Island and REC BMS. When connected, the REC completely controls the charging algorithm. It is required as the SI doesn't have the ability to monitor individual cells. You can program the REC via PC to the voltage set-points you need.

Are you able to have a high current connectors between all the parallel sets of cells? If so, then one BMS sensing/balancing wire per parallel set will work fine.

My BMS has one wire for each parallel pair in my LiFePO4 bank, and works as it should.

One thing i'm not sure of is your low voltage in that configuration. If you get voltage sag under load i think you will trigger the SI low voltage disconnect.

[emartin00]

Look up DavidPoz on YouTube. He has a Volt battery that he is using for off grid solar.

[Salts]

I can help a bit. I use a Sunny Island and REC BMS. When connected, the REC completely controls the charging algorithm. It is required as the SI doesn't have the ability to monitor individual cells. You can program the REC via PC to the voltage set-points you need.

Are you able to have a high current connectors between all the parallel sets of cells? If so, then one BMS sensing/balancing wire per parallel set will work fine.

My BMS has one wire for each parallel pair in my LiFePO4 bank, and works as it should.

One thing i'm not sure of is your low voltage in that configuration. If you get voltage sag under load i think you will trigger the SI low voltage disconnect.

Thank you Tom! Now we're getting somewhere! LOL
I was beginning to think I was the only person in the USA with a Sunny Island running off Lithium batteries. Regarding the paralleled BMS wires to the cells, how much current are we talking about when you say "high current"? I don't see anything preventing me from drilling the welded tabs and attaching a larger wire than what the Chevy Volt BMS connectors use.

The low voltage sag is going to be a problem if we put a big load on the system when the batteries are at the bottom of their SOC. Fortunately, I can can control the loads, and if I really need to put on a large load at low SOC, I can always hook in the generator.

[Salts]

Look up DavidPoz on YouTube. He has a Volt battery that he is using for off grid solar.

I've seen his videos. Its basically what I'm doing for the most part.

[SunEagle]

How different is the Volt battery compared to the Leaf?

[Ampster]

How different is the Volt battery compared to the Leaf?

The chemistry is similar. The Volt was water cooled and the Leaf air cooled. The Leaf consists of modules containing 2P2S in a Sardine can like shape. The Leaf has had some problems with battery degradation especially in hot environments because they are air cooled. They can easily be reassembled into packs from 8 volts to 384 volts.

The Volt modules vary from 24 volts to 48 volts. The Volt modules are hard to separate into smaller groups.

[Salts]

How different is the Volt battery compared to the Leaf?

You can't re-configure the cells. They come out of the pack in 12s and 6s modules and have welded terminals that can't be unwelded. You can combine two of the 6s modules into a 12s, but its damn near impossible to do much of anything else.

Its my understanding that the Chevy Volt battery has two things going for it over the leaf.. 1) its a higher quality battery.. but this doesn't matter much for solar storage.. 2) they are usually in better condition on the used market because the Chevy Volt has an engine and GM set the vehicle up so that the battery isn't stressed. There are multiple reports of Chevy Volts going 200,000+ miles without any real battery degradation. No first hand experience on that though.

The biggest drawback is that the Chevy battery cells can't be reconfigured.. its a major stumbling block.

[Salts]

That is a problem using a Chevy Volt pack. You really need 8 separate BMSs. The issue with trying to parallel Volt batteries is finding a way to connect the individual 3P buddy cells together. The connections need to be large enough to carry the eddy currents that will develope as the various packs charge and discharge. The existing BMS wires are too small for that.

With regard to CAN protocols do you mean the explicit messages to set voltage at the constant voltage phase and the current at which to stop charging during the CV phase?

How much eddy current are we talking about? I'm going to have 8 packs paralleled and a 6000 watt inverter so max continuous current is going to be around 136 amps. Divided by 8 packs, that about 17 amps per 12s string. We're not talking time travel in a DeLorean with Flux capacitors here..

That David Poz on youtube ran his volt pack for a year without a BMS and the pack stayed balanced. So here's what I'm considering, please chime in with opinions. What if I just purchased the Rec BMS and hooked it to a single string. Every two or three charge cycles, I would unplug the BMS leads and plug them into the next string. All the strings would still be paralleled so most of the protections would still be in place.. the only thing that wouldn't be protected is the balancing of any of the packs that aren't connected.. and they would eventually get their "time" on the BMS to make sure the series is balanced, it just wouldn't be a continuous thing.

I was also looking into the Zeva BMS system. Its far cheaper than the Rec BMS but I don't think its CAN BUS is set up to talk to the Sunny Island.

[Ampster]

How much eddy current are we talking about? I'm going to have 8 packs paralleled and a 6000 watt inverter so max continuous current is going to be around 136 amps. Divided by 8 packs, that about 17 amps per 12s string. We're not talking time travel in a DeLorean with Flux capacitors here..

I don't think they would be very large unless one cell loses a lot of capacity. Then you would only see it as the other cells attempt to equalize voltage. That would not occur at resting but only as they charge or discharge. I think 17 Amps is a good starting point.

That David Poz on youtube ran his volt pack for a year without a BMS and the pack stayed balanced. So here's what I'm considering, please chime in with opinions. What if I just purchased the Rec BMS and hooked it to a single string. Every two or three charge cycles, I would unplug the BMS leads and plug them into the next string. All the strings would still be paralleled so most of the protections would still be in place.. the only thing that wouldn't be protected is the balancing of any of the packs that aren't connected.. and they would eventually get their "time" on the BMS to make sure the series is balanced, it just wouldn't be a continuous thing.

That sounds like a plan. The important BMS functions of high voltage and low voltage disconnects would be in place no matter which string is connected to the BMS. I have a pack of mixed Nissan Leaf cells and I just use my Orion BMS for monitoring. I have turned off the balancing since. What I would look for when you are monitoring is the difference between the highest cell and the lowest. If that delta changes as you discharge the pack then that might be an early warning that one cell has slightly less capacity and is heading for the knee of the curve before the others. If there is any it will show up at 4.1 at the top and 3.5 at the bottom. Because I have some weak cells {60% of original capacity) in my Nissan pack I set my voltages at 4.05 and 3.80. When the lowest cell is at 3.82 I have a 120 millivolt difference to compared to the top cell at 3.94. At the top the difference is only 10 millivolts. I did not bin my modules before assembling my pack and it consists of modules from 3 different Nissan Leafs. I think you have the advantage of your cells all coming from the same pack and the Volt packs are more robust than the Nissan packs.
Regarding balancing, I don't think you will need to balance very often. When you do, you will only want it to happen when you are in the CV phase of charging because that is when the Amps begin to taper off and the balancing shunts will be more effective. Presumably you can set the balancing on or off or at least the voltage at which it should start.

I was also looking into the Zeva BMS system. Its far cheaper than the Rec BMS but I don't think its CAN BUS is set up to talk to the Sunny Island.

I think that is an important issue. When I got my Outback Skybox I was hoping to have CAN BUS connectivity but found out that unless there is similarity between the CAN BUS statements (not sure that is the right term) you can only monitor the CAN BUS traffic but not create actions. That was why I was skeptical in my earlier comment.

Hopefully the REC gives you some remote capability with logging. That is what I really like about my Orion Jr. I can look at the graph of my cell voltages at different phases of my charge and discharge cycle.

[SunEagle]

The chemistry is similar. The Volt was water cooled and the Leaf air cooled. The Leaf consists of modules containing 2P2S in a Sardine can like shape. The Leaf has had some problems with battery degradation especially in hot environments because they are air cooled. They can easily be reassembled into packs from 8 volts to 384 volts.

The Volt modules vary from 12 volts to 24 volts. The Volt modules are hard to separate into smaller groups.

Thanks. I have been getting a number of website advertisements that take me to a company selling the Leaf batteries in different configurations. Even though they are used batteries, most if not all are still expensive IMO.

[Salts]

I don't think they would be very large unless one cell loses a lot of capacity. Then you would only see it as the other cells attempt to equalize voltage. That would not occur at resting but only as they charge or discharge. I think 17 Amps is a good starting point.

That sounds like a plan. The important BMS functions of high voltage and low voltage disconnects would be in place no matter which string is connected to the BMS. I have a pack of mixed Nissan Leaf cells and I just use my Orion BMS for monitoring. I have turned off the balancing since. What I would look for when you are monitoring is the difference between the highest cell and the lowest. If that delta changes as you discharge the pack then that might be an early warning that one cell has slightly less capacity and is heading for the knee of the curve before the others. If there is any it will show up at 4.1 at the top and 3.5 at the bottom. Because I have some weak cells {60% of original capacity) in my Nissan pack I set my voltages at 4.05 and 3.80. When the lowest cell is at 3.82 I have a 120 millivolt difference to compared to the top cell at 3.94. At the top the difference is only 10 millivolts. I did not bin my modules before assembling my pack and it consists of modules from 3 different Nissan Leafs. I think you have the advantage of your cells all coming from the same pack and the Volt packs are more robust than the Nissan packs.
Regarding balancing, I don't think you will need to balance very often. When you do, you will only want it to happen when you are in the CV phase of charging because that is when the Amps begin to taper off and the balancing shunts will be more effective. Presumably you can set the balancing on or off or at least the voltage at which it should start.

I think that is an important issue. When I got my Outback Skybox I was hoping to have CAN BUS connectivity but found out that unless there is similarity between the CAN BUS statements (not sure that is the right term) you can only monitor the CAN BUS traffic but not create actions. That was why I was skeptical in my earlier comment.

Hopefully the REC gives you some remote capability with logging. That is what I really like about my Orion Jr. I can look at the graph of my cell voltages at different phases of my charge and discharge cycle.

Oh man, thank you for all that, you're a huge help. Now I have to digest it for a bit while I do more reading. My current thoughts are that I'm going to go with the Rec BMS and just swap the BMS leads from string to string every couple of cycles.

[tom rickard]

The problem is when you go to switch the sensing wires the REC will see low voltage and shut down the system. I'm not sure what it's sensing interval is - if you are quick it might be ok!

The current between parallel cells will be low, i