Primary lithium-thionyl chloride battery in a sealed enclosure - Hazards?

I've had multiple designs with lithium-thionyl chloride cells get through UL (and various other compliance) tests. Mine were smaller though; 2/3 AA vs your D. Also, the enclosures weren't sealed.

The impression I got from the tests was that basically they didn't care as long as a few conditions were met:

1. There must be a reliable way of making sure the charging current is never greater than the datasheet max value (assume 0.0uA if one isn't specified).
2. The maximum discharge current is limited. I think that might be more related to temperature rise rather than current per se.
3. The temperature range was limited.
4. There is no chance of mechanical penetration.

Note that those tests may be performed with a specified number of faults (typically one) active.

Example 1. You need to meet the reverse current spec with your reverse current protection diode shorted out.

Example 2. If you have a DC/DC converter on the board (it might be unrelated, but still thermally coupled), give it a reasonable current limit, which will control temperature rise when a short circuit happens.

Also, if using e.g. current limiters, using ones that already have UL recognition may make your own approval process easier.

Your test lab may have different ideas. Talk to them.

Regards, Allan

Thanks for the detailed response Allan. There's some good feedback there for me to work on.

In my experience the LSH20 is extremely stable, I've only seen a few failures while running a few billion hours with these cells. These required rather extreme circumstances to happen, though.

What kind of environment (temperature range, any moisture ?) and load profile do you expect ?

Dry environment inside a small IP67 ABS enclosure. Temp range could be anything from -10C to +60C

Most of the time its asleep drawing microamps, when it wakes it may draw

I worked with similar cells some years ago and came across one issue. If the cell has been "sleeping" for a long time and is then suddenly called on to deliver a lot of current the internal resistance may initially be higher than usual.

John

Sounds safe, our problems have been mostly with even hotter cells getting some mechanical shocks with high power load.

Sound like a long lifetime application. Take care of the cell passivation, the voltage can drop quite a bit if you have long inactivity period if the load comes online fast.