Published Saturday August 2, 2014: Updated August 4, 2014
Dump Mistakes in the Geo EV: ID10T Error
Operator Caused Battery Failure:
Commercial electric cars are designed with many fail safes in them. My Geo EV, which did not have the enormous engineering budget of say Nissan, has a few quirks that still plague its design. One of these is an imperfect battery management system.
Recently, I did one of the dumbest thing you can do in a home-made electric car. I killed one of the 24 LiFePo4 160 Ah batteries.
A flooded lead-acid battery on the other hand, is very forgiving when it comes to overcharging and over-discharging it. Yes doing either will shorten its life somewhat but at least it will forgive you and it will live for another day.
On the other hand, a LiFePo4 (Lithium-Ion Iron Phosphate) battery is not forgiving and will die if you give it the chance. It is vitally essential that whenever a LiFEPo4 cell voltage drops below 2.7 volts, you must charge it. If you decide to keep pushing on, you MUST bypass the low cell or else you will absolutely kill it on the spot.
It is crucially important to have a good battery management system (BMS) when you use LiFePo4 batteries. The BMS will prevent any cell from becoming over-charged or over-discharged.
This is a non-issue for all commercially made electric cars because they all have a superior BMS, designed to be foolproof, (so far as we know), against battery damage. Even if something does go wrong at first, most EVs carry a 100,000 mile warranty on the battery anyway. The risk is on the manufacturer and not the owner.
In theory, the Ligoo BMS on my home-made electric Geo keeps all the cells perfectly balanced. In reality, it doesn’t and one of the cells, B9, over time, becomes undercharged.
I suspect this is because the battery pack is split into thirds with relatively long wire runs between each battery pack sub-assembly.
Cells 1-8 are under the back seat where the gas tank use to reside.
Cells 9-18 are in a group just under the hood and
Cells 19-24 are arranged behind the front bumper.
You would think that an eight foot length of 00 gauge wire would not have any measurable voltage drop. But at nearly 400 amps, the voltage drop is upwards of 0.5 volts.
With the longest wire run between the first two battery thirds, B9 ends up taking the brunt of the voltage drop and having to make up the most difference. There is a 4’ wire run of 00 AWG wire to B19. It too will occasionally get unbalanced but not to the extent of B9.
With each battery charge cycle, B9 falls more and more out of balance with the other cells, until after several months it may only be at 2/3rd capacity of the other cells.
The lowest I have seen B9 get was over a 6-month period where it was 60 Ah low. That is literally 30 miles or 37% of my 80 mile range, gone missing.
This problem isn’t even noticeable on my normal 40 mile commute. I drive to work and back each day (charging up only at home) and the whole trip only uses 45-55% of the battery’s capacity.
It’s the occasional trip where I run an extra 10-40 miles beyond the normal 40 mile commute, that an unbalanced cell will rear its ugly head.
I make it a point every 1-2 months to manually re-balance B9 (and B19) to the other cells. This involves throwing a special 3.6V battery charger (made from a modified ATX computer power supply) onto the undercharged cell. While I should have spent the time and money to correct this BMS issue from day one, I never have. As a sciencey, experimenting kind of guy, manually charging every couple of months is not that big of an inconvenience to me.
UNTIL ONE DAY!!!
A few months ago, I ran several errands after work, driving 67 miles instead of my normal 40. It was around mile 49 that I noticed B9 started alarming a low battery voltage. It was then that I remembered, that I was way overdue for balancing the cell. Crap!
While I knew perfectly well what to do in that situation, in a moment of mental dimness, I ignored the emergency requirement to pull over and bypass the low cell before continuing on. I have never had to do that before, so why should I start now?
Foolishly, I pushed onward and drove the last 18 miles on a totally dead cell. Surprisingly the performance of the car didn’t suffer one bit. This only masked the seriousness of the situation, adding confusion to the issue and reinforcing the conviction of my folly.
Arriving home, I pulled into the garage and popped the hood, I was mortified at the slight smell emanating from battery pack. A very hot, very expensive LiFePo4 cell OUT-GASSING! Fortunately LiFePo4 cells are not very exciting in failure mode. Nothing caught fire and nothing blew up. Measuring the “at-rest’ voltage of B9, the meter read 0.03 volts. Dang!
HE’S DEAD JIM!
I tried to resurrect the dead cell several times but to no avail. Luckily all the other cells were un-harmed.
A very DEAD LiFePo4 cell several hours after being taken off the charger.
I thought about just living with the car the way it was, with a missing cell. But the on-board battery charger would now try to charge up to the same total voltage as before. But with only 23 cells instead of 24, this could potentially overcharge other cells and cause further damage. It would be better to just bite the bullet and buy a new cell. I ordered one from Electric Motor Sport out of California.
While a little higher priced for LiFePo4 batteries, They are the only store in the US that will sell a single 160Ah cell with fast shipping. They are also the same company that sold me most of the EV parts for the Geo, almost 2 years ago.
This stupid mistake cost me nearly $280 for a new, 160 Ah cell, expedited to my door.
Four months later, the new cell is still working great. Now I am more careful about keeping the cells balanced.
I keep the dead cell around as a reminder. “John, Don’t Do Dumb Things!” Live and Learn.