Lithium polymer battery pack problems

My mechanical-engineering buddy Dave Ruigh bought an $8,000 dollar Go-one carbon fiber recumbent tricycle. He uses it to ride to work to keep his heart healthy. The Go-1 is completely enclosed — it looks like a little jet or BD5 airplane. Since it is enclosed Dave can peddle to work even when it is raining. But Dave is never happy even with out-of-the-ordinary transportation, so he decided to make an electric vehicle out of it. He points out that what the electric does is make the ride safer — rather then sailing though intersections and lights in order to keep you speed up, you don’t mind coming to a stop and then you can use the electric to get back up to speed and then peddle.

Dave installed a hub motor. Then Dave bought two $800 dollar lithium polymer battery packs. The top figure is one of the packs with the outer covering off. Each pack has 30 batteries with 10 sets of 3 parallel cells. If you are following LiPo performance in the model airplane world you know how amazing these batteries are for weight to power. Dave’s bike was 20 pounds heavier but the LiPo pack could drive it at highway speeds the 4 miles to work and back with no problem. Dave only used it a few times and then the pack had a problem. One of the cells in one of the packs swelled up. We both knew how hard it is to keep batteries balanced, especially when they are paralleled as in this pack. Dave was a little worried about handing the pack, there are kilowatts of energy in them, but after he got the bad cell out, he shot it with a pellet gun and then threw it in a tub of salt-water with no real drama. LiPo can burn but there is still a limited amount of energy. Jim Williams gave me his take on electric cars last week. He said “take a laptop battery and try to move a car with it—you might the down the driveway and a block away. Now that that same volume or weight of gasoline and pour that into the carburetor — the car should go a mile or two.” Dave reports::

Here are some images of the pack after discharging and dismemberment. The Schulze charger was used to discharge the whole pack at about 0.4A. Although both the charger display and my direct measurement of the pack said it was at 33V, when disassembled, all the cells I could get to measured 3.7V. The positive tab of the bloated cell tore off flush when I attempted to extract it. It's likely the bad cell was taking a good portion of the current.

After the pack was disassembled and the offending cell removed, what to do with it? Well I would like to shoot some holes in the theories about LiPo safety, and that's exactly what I did. The puncture wounds were made with a Weihrauch HW45 4.5mm caliber spring piston pellet pistol. Apart from a tiny wisp of smoke that may have been dust from the plywood, no explosions or other violent results were noted.

I heard somewhere that a bucket of salt water is a good thing to immerse bad lipo cells in, and since vaguely remembered, anecdotal stories are as good as peer-reviewed science to me, I proceeded to mix up some saline and dump the cell in. A fair amount of gas was evolved initially, but it soon settled down to a slow boil. It now appears to have dissipated most of it's energy and only an occasional bubble can be seen.

I would think it would be nearly impossible to rebuild this pack into something useful, the way the wires are soldered nearly flush with the tabs would make in unworkable, and would also make selling the individual cells on eBay impossible.

Now it is certain that the batteries will be less dangerous once they are discharged. Too bad that Dave already used the bad cell for target practice since I would have been curious to see if it would take and hold charge. There are two more cells from that 3-cell parallel arrangement in the pack that I will get from Dave and abuse at our leisure. It will be fascinating to see if I can get them to swell and cause problems. And I will drive an ice pick thought a charged cell and see what happens. Batteries are a pain in the butt and if you can use a single cell instead you always should. And when you do have a battery try and avoid parallel cells. Also be sure to monitor the voltage and charge in each cell in a pack, it is the only sure way to protect and diagnose upcoming failures. This is why I am not too enthused about the Tesla battery pack that uses thousands of lithium ion cells all in series/parallel combinations. What a charge-discharge nightmare. And lithium may not be ready for EV use. Maybe that is why Toyota uses nickel metal hydride (273V, 228 1.2-V cells). Regenerative braking requires that you stuff charge back into the battery. Well, you can slam a ton of current into either of the nickel chemistries or lead-acid. Lithium batteries have far more restrictive charge regimes, so if you cannot push all the energy returned from braking into the battery by charging it fast, well then you have to dissipate it as heat in the brake pads. And here is your tax dollars at work: Thermal Evaluation of Toyota Prius Battery Pack (pdf).