Electrolyte leakage and corrosion of HV battery

[Mr. Mik]

Hi everyone,

can you tell me how to determine if a cell or stick is leaking electrolyte vs. being passively contaminated by leakage from another stick?

I have just opened up and inspected a Toyota Prius NHW10 half-pack. It consists of the same 20 sticks as a complete Insight pack.

I knew there was a single and total cell failure from readings taken with an S2000 scanner. I had reconditioned this battery about half a year ago, following years of neglect and very, very deep discharge. Some sticks had been down to a few millivolt only, none had been even close to normal voltages.

They all appeared to come back well with gentle treatment. I only opened and inspected the sticks in half-pack 2 and replaced one stick that had increased self-discharge in a test over several weeks. I don/'t know how thoroughly I inspected the sticks on that occasion, I might have missed some mild signs of leakage. The stick which I removed from the pack half a year ago looks as if it might have a small amount of electrolyte between the cells and the shrink-wrap, but no corrosion-bubbles.

The cell that died suddenly and completely after a few months of running in the car was in the half-pack 1, which I had never before dismantled. I found the dead cell just where I expected it (except that it was at the opposite end of the battery - Toyota appear to be going against the usual convention of numbering batteries in series beginning from the negative battery end.)

This stick with the dead cell had very obvious corrosion problems. There was a small amount of green stuff at the positive end and liquid between the steel and the shrink-wrap. It smells sweet and a little bit like chewing gum.
Corrosion bubbles were visible and palpable through the wrap. I removed the wrap and found severe corrosion on cells C2 and C6. The surface of cells C1, C4 and C5 looks pristine. To my surprise I found that C4 is the dead one with 0.003V! It looks fine but is dead.

This prompted me to remove the stick located below the dead one to check for electrolyte. Of course I ended up taking the entire half-pack apart....so much for the quick repair job I had in mind!

Results of the half-pack inspection:

1) The serial numbers are a bit jumbled, but consecutive. It does not appear like a stick was replaced previously.
2) Most sticks look like there is some electrolyte between wrap and steel. The color is darker than usually. Three look fine.
3) Many cells show corrosion bubbles which can be seen and felt through the shrink-wrap.
4) Only one stick appeared to have some electrolyte on the outside of the shrink-wrap.
5) The supportive frame parts do not look wet anywhere.

Do you think the wide-spread electrolyte contamination and corrosion in this half-pack is due to one (or a few) leaking cells, or due to individual leakages in all affected sticks?

Could the prolonged deep discharge have weakened the seals, causing multiple slow leaks once the battery was "restored" and then put to use?

Why do some cells corrode severely, while the one next to them remains pristine?

Cheers and thanks for any explanations!

[IamIan]

You can check for loss of chemical reactants ... by performing tests on the cells to determine Wh that they will discharge ... and the dV or voltage drop at a given current, SoC, and temperature.

Chemical reactants could be the Anode , Cathode, or the Electrolyte... electrolyte being by far the most common type to loose in batteries ... but it doesn't really matter which one is lost as you need all three in certain ratios for the cell to function properly.

A shortage of any one of the three can result in lower Wh of discharge capacity , and higher internal resistance in the cell.

Sense Wh of capacity and internal resistance both change with different SoC % and at different temperatures , etc.... if you test the cells compared to the others in a pack ... ideally getting ones to match as closely to each other as possible in as many performance characteristics as possible... of course all tests should be done in as close to the same methodology , and conditions as possible... I also recommend at least 3 identical tests of each cell in order to try and weed out statistical anomalies... for Insight battery packs ... I usually myself only went down the 6 cell stick level ... as I only found marginally small benefits from going to the individual cells , but it did increase the work load substantially, for that marginally small benefit.

The cells that test to be significantly different in terms of discharge Wh , or internal resistance, or self discharge rate ... those are the ones you want to remove / replace ... whatever the specific numbers you get for Wh , Internal Resistance, and Self Discharge Rate ... you want the pack to be matched as closely as possible.

Resist the temptation to rely on terminal voltage ... Discharge Wh , and Ohms are much better indicators ... for example ... I've had NiMH cells down to 0.01 V terminal voltage ... but they still took a charge and worked fine ... some with over ~90% of original capacity remaining.

[E27006]

There were reports of Prius cells failing with corrosion dating back a number of years ago.

Toyota initially not recognising the problem then began replacing packs.


The Insight battery pack draws warm air from the cabin venting into the hatchback area, how does the Prius ventilate the pack?

[Mike Dabrowski 2000]

The toyota issues with corrosion were mainly on the copper bands that connect the prismatic cells in series.
I believe they are using a nickle plated copper band in the newer systems.

Any venting of electrolyte on the prismatic cells should be handled by the rubber vent tubes.

On the cylindrical cells, the vent is right in the positive terminal, so the junction between cells will be where the juice would collect.

The electrolyte is potassium hydroxide, which is a strong base. If you get it on your skin, it feels like soap and is slippery. Use a mild acid like vinegar to change the base into a salt, and then it will wash off with water.

In my experience, the main reason for venting is that the cells got too hot and the internal pressure exceeded the vents max pressure.
Overcharging is the likely cause, or fan failure.