2p16s: What to do with faster (and slower) 'runners'

[addieboy]

Hello guys. My setup has been under observation for the last 10 days and would like to ask your inputs on the best practice on how to deal with the faster (and slower) cells on what I could only conclude as those with lower or higher internal resistance than the rest.

1. The 32 cells are Eve 280k, all came from Shenzhen Basen with about similar mfg dates. Since I am using a JK BMS with around 2A active balancer, I did not bother to top balance the cells so after completing the wiring, they are immediately charged by solar (no AC) using the MPP 5048 MGX in an off-grid setup. They are snug fitted on a 3/4 plywood (not compressed) and will be closed soon. This is a 48v setup, bulk charge at 57.6v, float at 54v, cutoff at 48v. Cells start to balance at 3.4v and 0.01v delta. For the past 10 days, I did not see the voltage at the batteries past the 54.4v mark, current always tapers until it reaches around less than 1A, even if the sun is still out.

2. I noticed that there are certain group of cells that stand out from the rest. Upon charging, cell group 1 seem to be a slow starter, always a laggard in reaching the voltage vis-a-vis the rest. Cell group 3 charges really fast. While group 1 is at 3.36v, group 3 is now at 3.43v.


3. At the higher state of charge, delta is less than 0.01v, indicating that the balancing worked well.


4. However, when they are discharging, cell group 3 seems to have no pacing at all, discharging voltage like there's no tomorrow. Here, they are now at 3.10v, switching places with group 1 who are still at 3.17v while the rest are around 3.14-3.15v.


5. At the discharged state, cell group 3 is now almost empty at 2.95v. The rest are still above 3.17-3.19v.



So, any suggestions or opinions on how to deal with this? My goal is just a 90%-15% SOC usage but it will be preferable to maximize the energy as much as possible since we are not only looking for a battery but a hobby as well. Some options:

1. Try to switch partners with 1 and 3 to 'balance' out the internal resistances.
2. Charge to full and top balance cell group 1 only.
3. Clean the terminals.
4. Dont worry about it as they are normal.

 

[sunshine_eggo]

You need to top balance.

 

[Kemills1]

Hello guys. My setup has been under observation for the last 10 days and would like to ask your inputs on the best practice on how to deal with the faster (and slower) cells on what I could only conclude as those with lower or higher internal resistance than the rest.

1. The 32 cells are Eve 280k, all came from Shenzhen Basen with about similar mfg dates. Since I am using a JK BMS with around 2A active balancer, I did not bother to top balance the cells so after completing the wiring, they are immediately charged by solar (no AC) using the MPP 5048 MGX in an off-grid setup. They are snug fitted on a 3/4 plywood (not compressed) and will be closed soon. This is a 48v setup, bulk charge at 57.6v, float at 54v, cutoff at 48v. Cells start to balance at 3.4v and 0.01v delta. For the past 10 days, I did not see the voltage at the batteries past the 54.4v mark, current always tapers until it reaches around less than 1A, even if the sun is still out.
View attachment 88536
2. I noticed that there are certain group of cells that stand out from the rest. Upon charging, cell group 1 seem to be a slow starter, always a laggard in reaching the voltage vis-a-vis the rest. Cell group 3 charges really fast. While group 1 is at 3.36v, group 3 is now at 3.43v.
View attachment 88532

3. At the higher state of charge, delta is less than 0.01v, indicating that the balancing worked well.
View attachment 88533

4. However, when they are discharging, cell group 3 seems to have no pacing at all, discharging voltage like there's no tomorrow. Here, they are now at 3.10v, switching places with group 1 who are still at 3.17v while the rest are around 3.14-3.15v.
View attachment 88534

5. At the discharged state, cell group 3 is now almost empty at 2.95v. The rest are still above 3.17-3.19v.
View attachment 88535


So, any suggestions or opinions on how to deal with this? My goal is just a 90%-15% SOC usage but it will be preferable to maximize the energy as much as possible since we are not only looking for a battery but a hobby as well. Some options:

1. Try to switch partners with 1 and 3 to 'balance' out the internal resistances.
2. Charge to full and top balance cell group 1 only.
3. Clean the terminals.
4. Dont worry about it as they are normal.

If the group 3 pair of cells appears discharged with a remaining capacity of 211Ah, it seems to me one of those cells is pretty much useless or these is no connection to it.

 

[Patrick_Cashman]

You need to top balance.

What he said.

 

[Sky-HHI]

Yes, top balancing is needed, also in your diagram on the wall and on the battery between group 8 and group 9 you only have one bus bar , which means all the current between the groups are running through that single bus bar. I think you need a second bus bar between 8 and 9 to equalize the current between these groups.

 

[sunshine_eggo]

Yes, top balancing is needed, also in your diagram on the wall and on the battery between group 8 and group 9 you only have one bus bar , which means all the current between the groups are running through that single bus bar. I think you need a second bus bar between 8 and 9 to equalize the current between these groups.

Nope. Please review:



I have circled each cell group interconnect. There is only one bus bar between between each cell group. The 8-9 connection is consistent with all other series interconnects.

 

[Sky-HHI]

Sorry sunshine_eggo, look in the resource section, under Cell Configurations for 12v, 24 v and 48v batteries. Filter Guy has examples for 2P 16S 48v batteries and look at example E , also you can look at 2P 8S 24v battery example D. Both have double busbars at the far ends. Is Filter Guy wrong?

 

[400bird]

It looks to me like cell pair 3 is both the highest voltage (at high SOC) and lowest (at low SOC)

That says to me that pair is weak. Maybe both cells, maybe one of the two, maybe one isn't connected (seems unlikely considering the pictures, but still worth mentioning)

I'd say you've got the options about right:
Definitely inspection the connections, just in case.
If it were me, I'd take it apart enough to capacity test some cells and replace the worst performing.
Alternatively you could just swap some cells around. If you have a pair that is low voltage (at high SOC) and high voltage (at low SOC) that's your strongest performer, swap one of those. Which now that I look back at your post, you're already aware that is probably pair #1.

 

[toms]

Sorry sunshine_eggo, look in the resource section, under Cell Configurations for 12v, 24 v and 48v batteries. Filter Guy has examples for 2P 16S 48v batteries and look at example E , also you can look at 2P 8S 24v battery example D. Both have double busbars at the far ends. Is Filter Guy wrong?

Yes, in this instance Filter Guy is wrong. Double busbar isn’t required. It only takes a quick look to see that a single busbar connects every other pair to the next pair in the chain - the end pair has the same current and requires the same busbar.

 

[sunshine_eggo]

Sorry sunshine_eggo, look in the resource section, under Cell Configurations for 12v, 24 v and 48v batteries. Filter Guy has examples for 2P 16S 48v batteries and look at example E , also you can look at 2P 8S 24v battery example D. Both have double busbars at the far ends. Is Filter Guy wrong?

@FilterGuy is neither right nor wrong. There is more than one way to skin a cat.

Perhaps you should look at example B....





Double everywhere EXCEPT the crossover between the two 8S groups.

 

[time2roll]

This is a 48v setup, bulk charge at 57.6v, float at 54v, cutoff at 48v. Cells start to balance at 3.4v and 0.01v delta. For the past 10 days, I did not see the voltage at the batteries past the 54.4v mark, current always tapers until it reaches around less than 1A, even if the sun is still out.

If the energy is sufficient I would tend to run them and see if the active balancer will bring the two problems in line over a few more weeks or months. Otherwise if there is a pair that hit top voltage and bottom first I would be inclined to replace them if they are significantly reducing the pack capacity.

In the mean time 57.6 seems high while having balance issues. Has there be an over volt situation? If bulk is set to 57.6 (seems high) I would be concerned to look closer to why the battery never gets above 54.4 volts. Is there enough sun? Are amps tapering nicely at 54.4 indicating not using boost voltage?

 

[addieboy]

It looks to me like cell pair 3 is both the highest voltage (at high SOC) and lowest (at low SOC)

That says to me that pair is weak. Maybe both cells, maybe one of the two, maybe one isn't connected (seems unlikely considering the pictures, but still worth mentioning)

I'd say you've got the options about right:
Definitely inspection the connections, just in case.
If it were me, I'd take it apart enough to capacity test some cells and replace the worst performing.
Alternatively you could just swap some cells around. If you have a pair that is low voltage (at high SOC) and high voltage (at low SOC) that's your strongest performer, swap one of those. Which now that I look back at your post, you're already aware that is probably pair #1.

Thanks. Yes, this makes sense though, maybe pair #3 has definitely some issues, either 1 of them is not at par with the rest. Will try swapping out with pair #1 and see if it balances out

 

[addieboy]

If the energy is sufficient I would tend to run them and see if the active balancer will bring the two problems in line over a few more weeks or months. Otherwise if there is a pair that hit top voltage and bottom first I would be inclined to replace them if they are significantly reducing the pack capacity.

In the mean time 57.6 seems high while having balance issues. Has there be an over volt situation? If bulk is set to 57.6 (seems high) I would be concerned to look closer to why the battery never gets above 54.4 volts. Is there enough sun? Are amps tapering nicely at 54.4 indicating not using boost voltage?

I did try to turn off discharging for two days to try to charge the batteries to full (my panels are only 5000w but another 5000w will be added in a few weeks) and the pack voltage did reach 54.68 before the overprotection kicked in (3.6v).



Now, at full charge, (and contrary to my first post) pair #1 has the highest voltage and pair #3 is the lowest (which makes my first post invalid).

And to add to more confusion, the MPP is registering a voltage of 57.6 (at the C- side).



I have never been so confused. lol

 

[Sky-HHI]

@FilterGuy is neither right nor wrong. There is more than one way to skin a cat.

Perhaps you should look at example B....

View attachment 88953



Double everywhere EXCEPT the crossover between the two 8S groups.

Why are you showing example B when example E that I referred to is how the batteries are arranged in the picture of the first post. No reason to rely as I am done with this issue.

 

[time2roll]

I did try to turn off discharging for two days to try to charge the batteries to full (my panels are only 5000w but another 5000w will be added in a few weeks) and the pack voltage did reach 54.68 before the overprotection kicked in (3.6v).

View attachment 88956

Consider reducing the charge voltage enough to stay under the 54.68 and avoid the over volt. Give it time to balance. Could be a few months depending on how much time is actually spent balancing each day.

 

[sunshine_eggo]

Why are you showing example B when example E that I referred to is how the batteries are arranged in the picture of the first post. No reason to rely as I am done with this issue.

Your reply is not relevant. The two configurations are electrically identical.

 

[codfish]

You can push the float up to 3.4v / cell and cutoff to 3.65v / cell while you are balancing the pack. And as previously suggested bring the charge voltage down to just below the most recent pack cutoff voltage, and slowly increase it. That way you can get the maximum balancing time. Just back them back down once it is balanced at the top of charge.

 

[halibut58]

Just a quick note if this has not been resolved. There could be a couple of different issues at play here. I found typically if you get symmetric errors at similar charge/discharge rates on a single cell (group), there may be some bus bar resistance issues. What I see here is +60mV at 60A charge, -60mV at 60A discharge (approx).

I would check your bus bars where the yellow circle is. You can compare it to the next one upstream during high current condition.
It could be high internal resistance in your cells, but since you have good quality cells (eve) and you have 2 in parallel, most likely it is the busbar.
One way to check is to move the sense 3 wire to the other end of busbar (blue). If the problem moves to cell 2, it is the busbar. If it stays with cell 3 it is the cell (group).
Another way to check is to measure the voltage at the studs at the points shown. During high current condition you may measure at least 60mV, most likely more. During comparison to cell4, it will most likely be 60mV more than the same reading from cell4 - then it is the busbar.
There could be other effects, but the above I consider low hanging fruit so would check first.

Once you find this problem, it will be easier to analyze balance.
As far as balance is concerned, if you actually have a battery where a given cell group hits high voltage disconnect and the same cell group low voltage disconnect, your battery is actually balanced enough to give you full capacity (The max capacity is given by your weakest cell group connected in series, and that cell will hit both if balanced). You can leave it there, or try to actually align at the top (top balance), but your total capacity will not change.

Also please note that I attempted to figure out which busbar is the culprit. What is given is my best guess and it is very easy to make a mistake - from your picture, since V3 is the problem, I would say it is busbar going cell 2 to 3.

 

[Samcat]

Remove all bus bars and use flat no conductive bar seating on top of all cells terminals. Use gap Guage and see if you have any gaps. If you are not using flexible bus bars this is step #1. Report back what you have observed.

 

[Kemills1]

Just a quick note if this has not been resolved. There could be a couple of different issues at play here. I found typically if you get symmetric errors at similar charge/discharge rates on a single cell (group), there may be some bus bar resistance issues. What I see here is +60mV at 60A charge, -60mV at 60A discharge (approx).
View attachment 90642
I would check your bus bars where the yellow circle is. You can compare it to the next one upstream during high current condition.
It could be high internal resistance in your cells, but since you have good quality cells (eve) and you have 2 in parallel, most likely it is the busbar.
One way to check is to move the sense 3 wire to the other end of busbar (blue). If the problem moves to cell 2, it is the busbar. If it stays with cell 3 it is the cell (group).
Another way to check is to measure the voltage at the studs at the points shown. During high current condition you may measure at least 60mV, most likely more. During comparison to cell4, it will most likely be 60mV more than the same reading from cell4 - then it is the busbar.
There could be other effects, but the above I consider low hanging fruit so would check first.

Once you find this problem, it will be easier to analyze balance.
As far as balance is concerned, if you actually have a battery where a given cell group hits high voltage disconnect and the same cell group low voltage disconnect, your battery is actually balanced enough to give you full capacity (The max capacity is given by your weakest cell group connected in series, and that cell will hit both if balanced). You can leave it there, or try to actually align at the top (top balance), but your total capacity will not change.

Also please note that I attempted to figure out which busbar is the culprit. What is given is my best guess and it is very easy to make a mistake - from your picture, since V3 is the problem, I would say it is busbar going cell 2 to 3.

Not sure I would agree here. In photo 5 the cells are discharged and at rest. If there were a poor bus bar connection between the cell pairs 2 and 3, the cell voltage would have recovered with no load. Since the voltage has not recovered, I doubt it’s the bus bar.