Safety of Shorted Lithium-Manganese Coin Cell Batteries (CR2032)

[granth]

I'm building some light up Christmas ornaments as gifts for friends and family, they're basically three self-flashing LEDs in parallel powered directly from a Li/MnO2 CR2032 coin cell. I ordered the PCBs and just got them today. This seemed like a pretty safe design to me, until I realized I failed to account for people slotting the batteries in backwards. I included no reverse current protection, and in my testing the circuit gets as low as 1.6 Ohms when wired in reverse. Murphy's law dictates that at some point someone will install the batteries backwards and hang it on a Christmas tree.

Is thermal runaway, venting, or explosion really something I need to be concerned about when heavily loading a small battery? I can't imagine that there's enough energy in that cell to be a problem but I'm not sure. The specific cell that I'm using is detailed here https://www.jauch.com/downloadfile/5bf5283b3496e3fc615261f65915827ad/cr2032_jauch.pdf Max current discharge of said batteries is 15mA, which means at peak voltage (3.3v) the battery will only discharge 50mW of power before tapering off and dying. I really don't think this could heat the battery past 70C before draining it completely. That being said, these are lithium batteries and that's not an element to mess around with.

TL;DR
Should I add reverse current protection when using a CR2032 to prevent thermal runaway due to incorrect battery installation?

[I wanted a rude username]

Welcome to the forum, granth!

Connect your ammeter across one and see what happens. 

In my test, using a generic AliExpress CR2032, the current starts at 75 mA, drops to 50 mA in a couple seconds, and then settles down to 25 mA. The cell barely gets warm to the touch, even after a minute. Personally I would be more worried about component damage, but you don't have a microcontroller on board to fry ...

That said it's generally a good idea to include reverse current protection in end-user devices, particularly since you can do it with a MOSFET, which avoids the large voltage drop you'd normally get with a diode.

[granth]

Thank you! Long time watcher, first time poster.

Don't know why I didn't do the ammeter test. My name brand batteries start off pretty high (~200mA peak) and taper down to a pretty consistent 90mA. At that current, the backwards circuit is about a 15 ohm load. The batteries get warm to the touch, but I can definitely still hold them. They're braced against a metal CR2032 clip and the ground plane of my PCB, which should dissipate even more heat. I'm leaning more on the side of me being paranoid for not including reverse current protection. It seems like a catastrophic failure is unlikely.

Do those reverse current mosfet circuits work reliably at ~3v? I've tried to build them before at this low of a voltage but could never get them working properly (my background is CompE, not EE)

[I wanted a rude username]

Do those reverse current mosfet circuits work reliably at ~3v? I've tried to build them before at this low of a voltage but could never get them working properly (my background is CompE, not EE)

I haven't researched it yet, just know it's possible even at 3 V. You might need a low-resistance signal level MOSFET ... something like the DMG2305UX.