DIY 48v 280ah Build - Blowout/Fail

[NCislander]

I completed my DIY build(s) a few days ago. 32 Cells total, 16S, 280ah EVE Cells, JK BMS(s) B2A24S20P, Growatt 5000ES units, Solar Assistant.
All electronics appear working as designed.
My 1st 16S battery build running the past month, maybe past 30-40 days. Clean, no observed issues. JK Active balancing normal. My 2nd 16S battery build put together and online this past Friday.
All looked clean, running normal.
I walked out to the Shop (external garage) and Growatt fans noise running normally. All indications looked fine from a 6-7 ft distance. As I get closer I can smell this funky smell. Strange smell, liquid on the lower battery cells. Vented cells I believe.
Both battery banks appear as blownout/swollen/fail cells #9/#10.
Weird thing, all electronics (Growatts & JK BMS) running and reporting normal. JK units actively balancing. Growatts actively charging. No fire, but massive heat around Cells #9/#10. Those cells swollen, Cell #9 lower/Bank2 vented. No breakers tripped (Nader 125Amp) on either bank. No actual explosion.
I immediately shutdown system.
Still Assessing damage. Possibly wait until they cool down. Waiting to disconnect and evaluate.
Very lucky no fire.

Damn, damn, damn. Expensive fail.
Need to assess and decide what I did wrong. Obvisously something.

BANK-1 Upper & BANK-2 Lower (Before Event)

BANK-1 (1st Build) Before Event

BANK-1 After Event (Swelling Cell#9)

BANK-2 Lower (AFTER Event) Cell #9/#10/#11


Constructive comments appreciated, anything obvious noted?

 

[Nobodybusiness]

I completed my DIY build(s) a few days ago. 32 Cells total, 16S, 280ah EVE Cells, JK BMS(s) B2A24S20P, Growatt 5000ES units, Solar Assistant.
All electronics appear working as designed.
My 1st 16S battery build running the past month, maybe past 30-40 days. Clean, no observed issues. JK Active balancing normal. My 2nd 16S battery build put together and online this past Friday.
All looked clean, running normal.
I walked out to the Shop (external garage) and Growatt fans noise running normally. All indications looked fine from a 6-7 ft distance. As I get closer I can smell this funky smell. Strange smell, liquid on the lower battery cells. Vented cells I believe.
Both battery banks appear as blownout/swollen/fail cells #9/#10.
Weird thing, all electronics (Growatts & JK BMS) running and reporting normal. JK units actively balancing. Growatts actively charging. No fire, but massive heat around Cells #9/#10. Those cells swollen, Cell #9 lower/Bank2 vented. No breakers tripped (Nader 125Amp) on either bank. No actual explosion.
I immediately shutdown system.
Still Assessing damage. Possibly wait until they cool down. Waiting to disconnect and evaluate.
Very lucky no fire.

Damn, damn, damn. Expensive fail.
Need to assess and decide what I did wrong. Obvisously something.

BANK-1 Upper & BANK-2 Lower (Before Event)
View attachment 170284
BANK-1 (1st Build) Before Event
View attachment 170280
BANK-1 After Event (Swelling Cell#9)
View attachment 170281
BANK-2 Lower (AFTER Event) Cell #9/#10/#11
View attachment 170285

Constructive comments appreciated, anything obvious noted?

Second one I have seen like this is the last week.

No clue as to why other than over charge or something in the battery composition.

 

[ArthurEld]

It looks like over charging to me. If the BMS settings were set right they should have shut the charging down.

 

[featherlite]

Bummer!

It looks like: (i) the strap(s) have crushed the end cell(s), (ii) caused by lack of end plates to spread the load. The cardboard was not sufficient.

Also, I suspect that the hooks and the ratchet mechanisms will eventually pierce the thin blue plastic wrap and/or the thin aluminum skins of the cells.

Otherwise, it looks like a tidy build.

I would not use those straps as a fixture.

 

[KauaiMolokai]

With all respect to those who have far more experience than I, I am not a fan of "tight clamping".

My understanding is that cells expand when being charged. This is a physical manifestation of a chemical event. This expansion "could" be contained if the entire cell casing was constructed like a "pressure cooker", or similar device. A tight strap around the cells will only cause the inevitable swelling to occur in another path of least resistance. But it WILL occur. The "bus bars" are pretty much rigid, unless built of flexible material. They will maintain the distance between the terminals while the cell expands. This could put great force on the terminals.

I'll be happy to be shown to be wrong. But it's not easy to break the laws of chemistry and physics without unwanted consequences.

 

[Quattrohead]

Bummer!

It looks like: (i) the strap(s) have crushed the end cell(s), (ii) caused by lack of end plates to spread the load. The cardboard was not sufficient.

Also, I suspect that the hooks and the ratchet mechanisms will eventually pierce the thin blue plastic wrap and/or the thin aluminum skins of the cells.

Otherwise, it looks like a tidy build.

I would not use those straps as a fixture.

I agree with everything here. Also no divider material between cells.

 

[NCislander]

I’m suspecting something I did wrong connecting Cell#8 to Cell #9, back row onto front row/right side.

 

[Hedges]

The busbars had a shape maybe for compliance, but last one was stretched straight!

A couple cells have top plastic peeled - heat?
Last one on right looks dented on corners, including bottom. Was that present earlier, or did it happen after failure developed?
Possibly, strap and U-channel applied too much pressure in small area, causing internal damage.

And those were nice cells, welded terminals with double bolt holes.

 

[NCislander]

The busbars had a shape maybe for compliance, but last one was stretched straight!

A couple cells have top plastic peeled - heat?
Last one on right looks dented on corners, including bottom. Was that present earlier, or did it happen after failure developed?
Possibly, strap and U-channel applied too much pressure in small area, causing internal damage.

And those were nice cells, welded terminals with double bolt holes.

Battery busbars are new factory flexible units. Cells are new 280ah EVE units (Grade A)
All damage happened earlier today.
No Cabling overheated.
Red straps were minimal tension originally. Mostly to keep cells aligned for battery busbars.

 

[NCislander]

Hoping I can recover majority of cells. JK BMS units still running/reporting.
No balancing/charging//discharging.

 

[NCislander]

It looks like over charging to me. If the BMS settings were set right they should have shut the charging down.

Straps were not orignally cutting into the cells. Possibly once cell began expanding, but not early on. If overcharging, I need to get a handle on it. Both JK BMS's set the same. (AND BOTH UPPER/LOWER CELL#9 FAIL)

 

[NCislander]

The busbars had a shape maybe for compliance, but last one was stretched straight!

A couple cells have top plastic peeled - heat?

Definitely Heat. Lower Bank Cell#9 blew its breakable seal.

Last one on right looks dented on corners, including bottom. Was that present earlier, or did it happen after failure developed?

All damage after failure. Pristine New cells prior.

Possibly, strap and U-channel applied too much pressure in small area, causing internal damage.

And those were nice cells, welded terminals with double bolt holes.

 

[Hedges]

I’m suspecting something I did wrong connecting Cell#8 to Cell #9, back row onto front row/right side.

Mechanically, that doesn't have the double bolt. Which would matter mostly to prevent rotation and loosening of bolt, not a problem between cells with these busbars.
Polarity is consistent with the others

The odd thing is no voltage issue reported by BMS. Over/under charge could affect a cell, but all cells get identical current (barring contact elsewhere, like cell can.)

If voltages diverged we would suspect that. Or just voltage too high. Is there a history of highest voltage?

Otherwise, that leaves mechanical damage (or defect). Normal charging causes swelling, so thinking about the stress concentration at U-channel.

How about terminals, did they ever get twisted in the can? People have reported that happening, but I suspect there are sheets of metal going from terminal to jelly roll, and twisting a bit in the can causes no severe damage. Maybe compromises seal.

Hoping I can recover majority of cells. JK BMS units still running/reporting.
No balancing/charging//discharging.

Hopefully so, but sleep with one eye open.

Maybe reject all end cells; the others got somewhat better force distribution.

Remember different series sets may be at different SoC, may need balancing.

 

[NCislander]

Mechanically, that doesn't have the double bolt. Which would matter mostly to prevent rotation and loosening of bolt, not a problem between cells with these busbars.
Polarity is consistent with the others

The odd thing is no voltage issue reported by BMS. Over/under charge could affect a cell, but all cells get identical current (barring contact elsewhere, like cell can.)

If voltages diverged we would suspect that. Or just voltage too high. Is there a history of highest voltage?

Otherwise, that leaves mechanical damage (or defect). Normal charging causes swelling, so thinking about the stress concentration at U-channel.

How about terminals, did they ever get twisted in the can? People have reported that happening, but I suspect there are sheets of metal going from terminal to jelly roll, and twisting a bit in the can causes no severe damage. Maybe compromises seal.



Hopefully so, but sleep with one eye open.

Maybe reject all end cells; the others got somewhat better force distribution.

Remember different series sets may be at different SoC, may need balancing.

Battery voltage was set to peak at 56.6v, just below 57volts. My Solar Assistant shows this.

 

[Hedges]

The question is cell voltages.
We've seen lots of bloated cells because someone charged in series without a BMS. But you've got a BMS.
Is your BMS capable of disconnecting cells from charger?
What highest cell voltages did it record?

"JK Active balancing normal."

While I've never worked with any of this stuff,
1) People say active balancers may make things worse, and balance can't be determined below knee of the curve.
2) Active balance I think charges cells based on other cell voltages. If it had a bad connection it might read wrong voltage and charge wrong.

Your BMS is not capable of disconnecting cells from active balancer. Can it shut off the balancer?

 

[NCislander]

The question is cell voltages.
We've seen lots of bloated cells because someone charged in series without a BMS. But you've got a BMS.
Is your BMS capable of disconnecting cells from charger?
What highest cell voltages did it record?

"JK Active balancing normal."

While I've never worked with any of this stuff,
1) People say active balancers may make things worse, and balance can't be determined below knee of the curve.
2) Active balance I think charges cells based on other cell voltages. If it had a bad connection it might read wrong voltage and charge wrong.

Your BMS is not capable of disconnecting cells from active balancer. Can it shut off the balancer?

57vdc is the highest bank voltage I can find. BMS is set for over voltage cutout 3.64v/cell. Active balancing enabled. It appeared to be doing its job.

 

[RV10flyer]

I use 1 inch oak boards on each end of my 15 ea diy batteries, 1/2” of ceramic fiber batting between board and cells, then two 1” straps with one of them near the top and one near the bottom. I do not tension them very tight when they’re at 30% state of charge. If you do they will be way too tight when fully charged. I think you had an internal cell short from over tightening on a small surface area, which will not trip your main breaker. I agree with those posts above. I’d check IR (at same voltage as when you received them if you tested that and wrote it down) and capacity test remaining undamaged cells. Then start over with 5+ new cells. 18 mo…no issues. I have a bunch of those great gradeA cells from 18650Battery.

 

[RV10flyer]

57vdc is the highest bank voltage I can find. BMS is set for over voltage cutout 3.64v/cell. Active balancing enabled. It appeared to be doing its job.

I absorb at 56.0, float at 54.0V. OVP at 3.60V.

 

[Brucey]

Thinking compressed at low charge, swelled up with more charge and put too much force on the terminals? I like the braided connectors, will allow for change im distance between the terminals. That ratchet strap definitely give me concerns with the single unistrut holding it on one end.

 

[NCislander]

I use 1 inch oak boards on each end of my 15 ea diy batteries, 1/2” of ceramic fiber batting between board and cells, then two 1” straps with one of them near the top and one near the bottom. I do not tension them very tight when they’re at 30% state of charge. If you do they will be way too tight when fully charged. I think you had an internal cell short from over tightening on a small surface area, which will not trip your main breaker. I agree with those posts above. I’d check IR (at same voltage as when you received them if you tested that and wrote it down) and capacity test remaining undamaged cells. Then start over with 5+ new cells. 18 mo…no issues. I have a bunch of those great gradeA cells from 18650Battery.

All above makes sense. I appreciate the advice & feedback.

 

[NCislander]

Thinking compressed at low charge, swelled up with more charge and put too much force on the terminals? I like the braided connectors, will allow for change im distance between the terminals. That ratchet strap definitely give me concerns with the single unistrut holding it on one end.

Yes, I’ll not be using the rachet strap again.

 

[ArthurEld]

The rachet strap isn't the best but I don't think that had anything to do with this problem. It seems odd that it happened to both batteries.
What was OVP set to on both BMSs?

 

[Brucey]

Wonder if the hot liquids/discharge from the upper battery dripped down to the lower battery and set that one off?

 

[NCislander]

The rachet strap isn't the best but I don't think that had anything to do with this problem. It seems odd that it happened to both batteries.
What was OVP set to on both BMSs?

OVP- 3.64

I'm at "work" away from system currently. I plan to disassemble batteries this evening. Also I will review and document my JK BMS settings. All I can think of BMS settings wise, I must have made same BAD decision and changed them both to effect both banks.

 

[NCislander]

Wonder if the hot liquids/discharge from the upper battery dripped down to the lower battery and set that one off?

No liquid was blown out of upper battery bank. Only the lower.
Upper Cell #9 is swollen, but not near as bad as lower bank, same Cell.