I have experience with  LiFePO4 that discharged to below 1V and reviving them.
Some rebounded and were fine, with very low self-discharge even after many 
months of sitting at 3.2V.
However, a few did not take a charge but instead turned into resistors and with 
continued (low current) charging, became ever lower in resistance. So, if they 
would be part of a buddy pair then that would easily lead to thermal runaway 
and a fire.
Note that this is not unique to Li batteries. Many moons ago a long time EV’er 
in Petaluma lost his barn with EVs inside due to a fire that appeared to have 
started in a Lead-Acid buddy paired pack. Remember that Lead cells also have a 
tendency to fail, I had a colleague that for more than a year drove an old 
pickup truck that had a 10V battery. Most of the time that still started his 
truck just fine, so he kept using it instead of replacing, I even drove that 
truck to San Diego and back to Silicon Valley with a Leaf in tow, with that 
battery, 2 years ago and last I heard, now 2 years later, he is still using 
that 10V battery…
One way to avoid fire from a short in a cell is to have two individual strings, 
not buddy pairs.
If one cell shorts, it just pulls that string a little lower but not 
catastrophically much and any good battery install must have a fuse in each 
parallel path, so one fuse per string.
Note that a Tesla pack has 60-90ish parallel cells, each with an individual 
fuse (doubling as its connecting wire to the bus bar).
I have heard suggestions to interconnect the cells in parallel strings with 
thin wire to allow them to balance, I do not know the wisdom of such hybrid 
structure. If you are using a BMS it might be helpful to avoid two complete 
BMS’es on the two parallel strings, though that is what I chose to use: two 
Leaf packs, each with its individual BMS and the complete strings are 
paralleled, with a fuse in each string. My battery box has an off-the-shelf 
3-phase ON/OFF switch and I have wired 2 of the contacts to break each string 
individually.
Hope this clarifies.
Cor.

Sent from Mail for Windows 10

From: David Nelson via EV
Sent: Monday, August 6, 2018 4:48 PM
To: Electric Vehicle Discussion List
Cc: David Nelson
Subject: Re: [EVDL] Lithium Ion buddy pairing? (Waselectriccarpartscompany.com?)

On Tue, Jul 31, 2018 at 9:25 PM, Tom Hudson via EV <[email protected]> wrote:
>
> Those two were ZERO VOLTS, and show ZERO OHMS resistance.
>
> Is this a common failure mode of lithium-ions (or the CALB modules)?  I'm
> thinking that I'd better set up some kind of fusing between the buddy pairs
> because if one of the pair fails in this way, it's going to present a dead
> short to its buddy -- and nothing good is going to come of that.

I don't know how common of a failure mode that is but I suspect it
could happen. If they were placed upside down (terminals facing down)
there is evidence that that has corroded the internal contacts to the
terminal bolts but that caused an open circuit situation, not a dead
short.

Internally the cells are made up of several layers which are
"buddy-paired" at the terminals. I put 20 buddy pairs of TS-LFP100AHA
cells in my Gizmo back in January 2010 and they are still going
strong. While the probability of any given cell failing goes up with
the number of cells I claim that a 200Ah cell made up of the same size
materials as the 100Ah cells would also have a higher probability of
failure. Definitely give each cell a cycle test and if you have the
time a capacity test. I would recommend that you then pair them up
where the capacity of the pairs ends up being nearly identical. I did
a rudimentary version of this for my Gizmo and it has worked out just
fine.

Some claim that you have to have the cells at exactly the same SOC
before connecting the buddy pairs. Theoretically this doesn't make
sense and I have also tested it and practice shows it really doesn't
matter. As long as their voltages are similar the initial current
isn't that great. Somewhere in the EVDL archives I posted the numbers
from one of my tests of this. IIRC it maxed at ~100A for a second or
so and quickly dropped to below 20A or so. Charging the cell pair to
3.4V or discharging to 2.8V or so brings the two cells in line with
each other.

The 100% SOC resting voltage of LFP cells from CALB and TS is 3.38V.
Make sure you aren't hammering the cells with 3.6+V on every charge
trying to equalize them. A good set of cells doesn't drift that fast.
I have been running my pack top balanced without attached BMS boards
for several years with no issues. I recently sold the Gizmo and I
don't trust the new owner to be able to be the "human BMS" so I am
going to be reinstalling the BMS boards I have. Furthermore, it is an
unknown how much the cells will drift as they age. They have been fine
so far but I'm not driving it now so can't do my annual voltage check.
I've been charging to 3.465V/cell with my Zivan chargers and the
resting voltage with no load after 12hours has been 3.36V so I know
I'm not overcharging them.

HTH,

-- 
David D. Nelson
_______________________________________________
UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
http://lists.evdl.org/listinfo.cgi/ev-evdl.org
Please discuss EV drag racing at NEDRA (http://groups.yahoo.com/group/NEDRA)


-------------- next part --------------
An HTML attachment was scrubbed...
URL: 
<http://lists.evdl.org/private.cgi/ev-evdl.org/attachments/20180806/eeda6c92/attachment.html>
_______________________________________________
UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
http://lists.evdl.org/listinfo.cgi/ev-evdl.org
Please discuss EV drag racing at NEDRA (http://groups.yahoo.com/group/NEDRA)

Reply via email to