Opinions on Chevy Volt Battery part time BMS ?

[Salts]

I think this conversation is becoming a distraction to the needs of the OP. In his situation the voltage is 48 volts and the maximum amperage of two 5k inverters at full load is 200 Amps.
I do agree that high voltage and low voltage disconnect logic is needed in a stationary pack. Damage to lithium batteries can happen at either end of the charge/discharge curve and that is why that is necessary.

Rec-BMS offers a 48 volt contactor that I plan to order with their BMS. I have two 175amp DC circuit breakers I plan to use with 2/0 battery cable. I purchased a TEMco 11 ton hydraulic crimper and will probably also add solder to all the large battery cable lugs in order to reduce any and all possible points of adding resistance to the connections. Total battery cable length should be less than 10 feet to each Sunny Island.

The Sunny Island also has its own DC Breaker built in. I'm wondering if I should add a fuse or if the DC breakers will be sufficient.

[Ampster]

When you get to higher current contactors they often have arc suppression circuitry, this only works for one polarity. In an EV, the contactor primarily is used for a low voltage disconnect. In a stand alone power system, you are as likely to get a full charge current failure as a low voltage failure.
(i've seen more systems fail due to overcharge than low voltage)

In any case, plenty of people are using the Gigavac as a disconnect for both HV and LV - that doesn't make it right.

I think this conversation is becoming a distraction to the needs of the OP. In his situation the voltage is 48 volts and the maximum amperage of two 5k inverters at full load is 200 Amps.
I do agree that high voltage and low voltage disconnect logic is needed in a stationary pack. Damage to lithium batteries can happen at either end of the charge/discharge curve and that is why that is necessary.

[Salts]

To do that you will have to switch at least 13 wires, which means a lot of wires and a lot of relays. When you switch the BMS will likely open the safety contactor because the sensed cell voltage will go to zero momentarily. So you put in a filter or some caps or something to "hold up" either the battery sense leads or the output to the contactor. Now you've slowed down the response of the system, and it will not work as designed. Will it still protect the pack? Maybe, maybe not. Plus you are now dealing with 39 wires - and those are raw cell voltages, and if any of them gets misconnected or shorted, a fire is inevitable.

Hi jflorey2,
Thank you for your input. The REC-BMS I am considering using has this long green plug that all the BMS cell tap wires connect to. After they are all connected to the green plug, the plug is then inserted into the BMS. My idea is to order 7 extra plugs so all that I have to do is turn the system off, unplug a single connector, and plug the next one into the BMS. Restart the system, and I'm done. In other words, I'm not physically switching 13 wires going into screw terminals and having to connect and disconnect each individual wire. Its just a single multi-pin plug that each 12s string will have pre-wired.

Yes, the main battery contactor will open and shut down the system, I will shut down the system manually in the proper order so as not to put any stress on the BMS or the Sunny Island.

This is an off-grid system for my home. We do not run a surgical operating room where having power all the time is a life or death issue. Its just an off-grid emergency system to keep the freezers cold, the sump pump running, and a desktop computer or two going.

I have a lot of flexibility to work with.

[jflorey2]

Exactly what is going to cause a fire with switched sensing only wires that is not going to happen with 3 individual BMS's?

To do that you will have to switch at least 13 wires, which means a lot of wires and a lot of relays. When you switch the BMS will likely open the safety contactor because the sensed cell voltage will go to zero momentarily. So you put in a filter or some caps or something to "hold up" either the battery sense leads or the output to the contactor. Now you've slowed down the response of the system, and it will not work as designed. Will it still protect the pack? Maybe, maybe not. Plus you are now dealing with 39 wires - and those are raw cell voltages, and if any of them gets misconnected or shorted, a fire is inevitable.

[jflorey2]

The problem is that I have 8 individual 12s strings (Chevy Volt) and each of them would require a BMS which would be very very expensive.

You can get a 12S li-ion BMS from Alibaba for $10.00 each. (I would post a link but that doesn't work well here - search for Battery PCB BMS for 12S 44.4V Li-ion/Li-Polymer Battery Packs)

I assume $10.00 isn't very very expensive to you.

Then there's the Sunny Island. I think it is preferable to have a BMS communicate with it so that it benefits from the best charging protocols. Problem is that I know nothing about how CAN BUS works.

CAN BUS is just the physical layer. It's basically a serial interface. The hard part is what protocol rides on top of the CAN bus. MODBUS is a popular one.

So what's better: 1) A cheaper BMS like the Zeva (Master + 8 Slaves) but no Sunny Island CAN BUS communication, or 2) A higher quality Rec-BMS on a single string but with full Sunny Island communication through the CAN BUS ?

I'd do one good BMS so the Sunny Island gets good info and the rest cheapo BMSes. You can use the onboard power switch for the BMS, or use a contactor (or remotely triggered breaker) to cut power when a cell goes out of bounds (safer.)

I started to think that I should start with the Rec-BMS. Install the battery safety contactor and have the CAN BUS control the charging. Then later, I could get a Zeva BMS (Master + 7 slaves) for the rest of the pack. I could wire the Zeva BMS so that it also controls the battery safety contactor along with the Rec-BMS through a relay. If both BMS's don't agree then the contactor stays open until they do.

That sounds good. Again, consider a remotely operated breaker rather than a contactor; safer and less power required.

[Salts]

It sounds like you have done a lot to reduce the risks associated with Lithium batteries. I disagree with the comment about absolutely needing a BMS to prevent a fire. They can give you an early warning that a cell has become weak but they can't necessarily prevent a thermal runaway. There is even a lively debate about whether a BMS creates more risks than benefits. I am not of that belief. I think a BMS is a very important management information system. You can be a BMS with a multimeter. As I have said earlier, it is most important to observe the cell differences as the charge or discharge cycles approach the knees of the charge/discharge curves. Sacrificing some capacity by conservatively setting your high and low voltage set points can extend the life of your pack and reduce the risk of thermal runaway.

The idea of connecting all the cells to be able to have one BMS is more about convenience. In the case of the Volt batteries it is not very convenient. You are correct in your comment #4 that paralleling the cells might mask the problem that only one cell is going bad.

Thank you for your help, you are a wealth of knowledge.

This goes back to using just one BMS and swapping it from string to string at regular intervals. Turn BMS Off (battery contactor opens), remove BMS Cell Tap Plug, swap in a different plug, turn BMS on)
Also, I don't plan to push my batteries to their limits. I have no problems with adding more battery capacity in order to stay away from those charge-discharge knees. I plan to charge to 4.1 and will discharge to 3.5. I'm having a hard time figuring out exactly how much energy this will harvest from my 2014 Chevy Volt pack, but my best guess is that it will give me about 10kw-11kw from a 16.5kw pack. The good news is that what I give up in capacity, I get back in life cycle longevity of the pack.

[Ampster]

It sounds like you have done a lot to reduce the risks associated with Lithium batteries. I disagree with the comment about absolutely needing a BMS to prevent a fire. They can give you an early warning that a cell has become weak but they can't necessarily prevent a thermal runaway. There is even a lively debate about whether a BMS creates more risks than benefits. I am not of that belief. I think a BMS is a very important management information system. You can be a BMS with a multimeter. As I have said earlier, it is most important to observe the cell differences as the charge or discharge cycles approach the knees of the charge/discharge curves. Sacrificing some capacity by conservatively setting your high and low voltage set points can extend the life of your pack and reduce the risk of thermal runaway.

The idea of connecting all the cells to be able to have one BMS is more about convenience. In the case of the Volt batteries it is not very convenient. You are correct in your comment #4 that paralleling the cells might mask the problem that only one cell is going bad.

[Salts]

I'm still not sure what you have against a linking wire (eg 10a) between each parallel pair? I can't see a problem with this approach.

Nothing against it specifically, but I do have a few concerns that I haven't worked out yet. What is (eg 10a) ?
1) What ga wire is needed?
2) I would have to fuse each wire, that's 96 fuses and 96 fuse holders. (Technically, I think it I could get away with 84 fuses) That's still a lot of possible bad connection points, points of possible failure, and points of more resistance. The Chevy Volt has a BMS wiring connector built into the cell modules, all I have to do is use the original connector, but as was stated, the wires aren't big enough for paralleling cells and dealing with the eddy currents.
3) I'm not sure if I understand the various failure modes that can happen with paralleling 8 cells with the BMS that way. I am hesitant to do things that I don't fully understand.
4) If a single cell in one string starts to have problems, the BMS won't know it until it starts to affect the other cells. I'm not entirely sure if this is even a problem or not because I think that eventually it would start to drag down the other cells and the BMS would at least see that, maybe not as early, but it should show up eventually.


I did consider using 4 BMS's and paralleling two strings to each BMS. That would cut the number of fuses down by half to just 48. Still a lot of fuses though.

This battery bank is going to be in my basement. It will all be mounted inside a large industrial Hoffman steel box that is 4ft tall, 3ft wide and 20 inches deep with a hinged door. I think its made of 14ga steel. Obviously, I need to balance safety and cost so I'm not for or against anything yet.

You guys are a boat load of help.. Can't thank you enough.

[tom rickard]

When you get to higher current contactors they often have arc suppression circuitry, this only works for one polarity. In an EV, the contactor primarily is used for a low voltage disconnect. In a stand alone power system, you are as likely to get a full charge current failure as a low voltage failure.
(i've seen more systems fail due to overcharge than low voltage)

In any case, plenty of people are using the Gigavac as a disconnect for both HV and LV - that doesn't make it right.

[Ampster]

To be clear, i mean the main contactor terminals have to be connected so DC current being disconnected under load is flowing in only one direction.

(ie the main contact terminals have a + and - )

No worries, it doesn't matter with Kilovac. I hung out on the diyelectricar.com forum for a few years and never heard that one before. But I never heard of Gigavac either. FWIW, I also thought DC flowed in only one direction. Never seen otherwise.

[Ampster]

I'm still not sure what you have against a linking wire (eg 10a) between each parallel pair? I can't see a problem with this approach.

Yes that would eliminate multiple BMSs. The way the tabs are configured it might be tricky to drill holes. I will have to see what Dave Poz did.

[tom rickard]

I'm still not sure what you have against a linking wire (eg 10a) between each parallel pair? I can't see a problem with this approach.

[Salts]

Geez, don't do that. If you need balancing/monitoring on all of them, get a BMS for all of them. Don't skimp on protection; the downside is a fire.

The problem is that I have 8 individual 12s strings (Chevy Volt) and each of them would require a BMS which would be very very expensive.. Then there's the Sunny Island. I think it is preferable to have a BMS communicate with it so that it benefits from the best charging protocols. Problem is that I know nothing about how CAN BUS works. My network skills are basic at best and my programming skills are limited to some Visual Basic and database stuff. Not exactly hacker extraordinaire. The Rec-BMS comes pre-configured for controlling the Sunny Island.

So what's better: 1) A cheaper BMS like the Zeva (Master + 8 Slaves) but no Sunny Island CAN BUS communication, or 2) A higher quality Rec-BMS on a single string but with full Sunny Island communication through the CAN BUS ?

I started to think that I should start with the Rec-BMS. Install the battery safety contactor and have the CAN BUS control the charging. Then later, I could get a Zeva BMS (Master + 7 slaves) for the rest of the pack. I could wire the Zeva BMS so that it also controls the battery safety contactor along with the Rec-BMS through a relay. If both BMS's don't agree then the contactor stays open until they do.
The Zeva BMS is just under half the price of the Rec-BMS, and if I add more Chevy Volt packs later, I can just add more of the cheap Zeva slaves.

Opinions?

[tom rickard]

I am still trying to understand what you mean when you say bidirectional. Are you talking about current flow like AC and DC?

To be clear, i mean the main contactor terminals have to be connected so DC current being disconnected under load is flowing in only one direction.

(ie the main contact terminals have a + and - )

You need to check with your manufacturer, i know this is true for the Gigavac GX14, which some people were selling as suitable for off-grid power systems.

[tom rickard]

Geez, don't do that. If you need balancing/monitoring on all of them, get a BMS for all of them. Don't skimp on protection; the downside is a fire.

Exactly what is going to cause a fire with switched sensing only wires that is not going to happen with 3 individual BMS's?

All well and good to be an alarmist, but all 3 packs will still have pack level over and undervoltage protection.

It would be helpful if you had knowledge of how the batteries worked, and what state you could see occurring that would cause a fire.

I can see issues with 3 BMS having communication conflicts on the CANBUS, so i wouldn't go that way at all.