(ET) Major pack failure in BB600 NiCD pack last night.

[Chris Zach <cz alembic crystel com>]

Well, after 20 years of running my E20 on the NiCD batteries without an 
issue everything.... went... bad. I'd consider this a catastrophic 
failure and figured I'd write it down for posterity.

Symptoms:
Symptoms started on July 4th, after a night of fireworks I was driving 
the tractor back through the brush when the deck got hung up on a root. 
No biggie, I just needed to lift the deck. Lift went partially up, then 
hung up.

Lift was dead. Thought it was the fuse, abandoned tractor. Next day came 
out, got the deck over the root, drove it back to house. And the lift 
was *dead*. Over the next month I did troubleshooting, lift was fine, 
switch was fine, etc. Finally I settled on the wire back to the battery 
pack which was connected to the middle of the last battery block to 
provide 18 volts of lift voltage. Rather than get under and trace out 
the wire I hooked the lift's return to the middle of the front pack and 
abandoned the wire in place.

Mistake 1: The wire was not the problem, something was wrong in the pack.

Another issue is that if I left the tractor the pack would go dead. It 
would charge up, and run and drive, but would die after a day. Figured I 
had a parasitic load on the lift and put it down to something I would 
work on after my trip to Iceland last week. So I parked the tractor in 
the yard, turned off the main disconnect and left it.

Came back yesterday and needed the tractor to haul out junk. It was dead 
again, but I put it on the charger, charged it, then drove it out over a 
bouncy path. Voltages were all over the place which was odd and after a 
bit of driving I drove it back to the yard, put it on a 6 hour charge at 
6pm with the timer, then went out for the evening.

Got call from wife at 6:30 pm saying my son noticed the charger was 
"very loud" and smoke was coming from the tractor. They unplugged it and 
I came home by 8 to disconnect the four packs.

Note: My E20 has a total of 120 BB600 cells. They are arranged in 4 
packs of thirty wired in parallel at the string ends. Two sets of thirty 
are in the back of the tractor, one is in the front, and the last one is 
in a battery/tool box off the rear. There is also a 36 volt line from 
the batteries to the rear box (AMP disconnect) and an 18 volt tap for 
the rear lift.

When I got back last night I saw that the rear batteries were hot (bad) 
front ones were warm (bad) and the ones in the external box were 
warm-ish (not good). I disconnected one set of interconnects from each 
battery to ensure nothing could discharge into another pack, cooled it 
down with a hose, then left it for this morning.

This morning I checked the cells and they were in bad shape. I could see 
that a ground fault hit the rear pack in the center where the battery 
box had sagged down enough to make contact. Not good, I have a long 
standing ground fault in the drive motor that places the compound field 
in the armature as a ground fault location. It's not fatal (it just 
energizes the field, and pulls limited current) but it is there and I 
can't figure out how to fix it (pulled motor, it's in the large series 
field, not the shunt field). Anyway that probably drained 1/2 of the 
rear pack, and the rest of the batteries then discharged into that 
unbalanced fault.

However that was not the bad part. The bad part was the charger was 
merrily charging away. Now, the E20 charger goes to a max of 43 volts 
(tapered down current after 40 or so volts) which is less than the 1.5 
volt peak charge of a BB600 NiCD. As a result the batteries are always a 
bit low but having that headroom keeps the batteries from every going 
into thermal runaway (once you go past 1.6 volts on a BB600 it will heat 
up, boil the water, then short as the polyethylene cell separator melts)

In this case, the short meant that 1/2 of the back pack was now 
essentially an 18 volt pack being fed by the charger with the only 
current limiting being the field (which can allow a number of amps 
through). Thus the cells got hot, boiled, and shorted. This then caused 
the other strings to feed their current through the last pack, damaging 
more cells in the process.

As of this morning I have a lot of cells with 0.0 volts. Putting a 3a 
current limited charger shows no volts and 3a current after 15 minutes 
which means the cells are shorted and bad. Great. What I can see is this.

String 1: Rear of rear battery box.
Left side 18v: All cells good. These were probably protected/shadowed by 
the short and thus didn't get overcharged. Great.
Right side 18v: All cells but one dead. These were the main source of 
the steam/smoke I guess. The one cell that is not dead I need to check 
in more detail but it's probably damaged.

String 2: Front of rear battery box:
Left side 18v: 4 cells dead, all closest to the rear pack. One cell dead 
in each string. Odd.
Right side 18v: All cells dead again. Great.

String 3: Front of tractor.
First two cells in the string are shorted. I wonder if the charger was 
just chewing through the cells at this point, when one cell shorts the 
remaining cells get more voltage. As more cells short the remaining 
cells boil and are destroyed. Looks like it was caught before this happened.

String 4: Battery box.
All cells in the battery box were ok. One was a bit low (.8v) but is 
being charged up to match the others.

Thoughts in another post.

Chris