Dive lights always have a catalyst pellet in them. This helps reduce the danger if the vented gas is hydrogen.

Hello Arney,

I also go by SilverFox on CandlePowerForums. I am heavily involved in battery testing. We have had several reports of the CR123A Lithium primary cells "rapidly venting with flame." In an effort to understand why, we set out developing a test plan and conducting a series of tests. NewBie (on CPF) did the actual testing and we were able to come up with a recipe to cause this "venting with flame" to occur.

The formula we discovered goes like this:

You need an incandescent lamp that draws about 1 C (or around 1.2 - 1.5 amps).
You need one cell at 100% capacity, and the other at around 70 - 80% capacity.
You need an ambient temperature of around 80 F.
You need an enclosure of some type. (More on this later).
You need to run the light completely down, then you need to leave the circuit connected.

Rapid venting with flame occurs 5 - 20 minutes after the lamp goes out. By the way, the temperatures exceed 1000 F for the short time the cell is venting.

This works almost every time with the inexpensive CR123A cells. It never works with the premium cells such as SureFire, Streamlight, Energizer, Duracell, and Sanyo. We haven't tested all of the brands, but we are seeing a trend here.

The problem is that the stronger cell tries to charge the weaker cell. With some temperature, the reaction is able to continue and after "cooking" for a few minutes... bang.

Cells in an open environment will not cook off. You need an enclosure to maintain the heat within the cells. We started out with a flashlight body, then progressed to simply wrapping the cells in some aluminum foil to simulate the flashlight body. This worked quite well.

It would be my "guess" that the pilot, upon discovering that his light had gone out, had pressed the switch several times and may have lost track if it was turned on or not. I believe he ended up with the light in the on position and that allowed one cell to reverse charge the other.

We have discovered that the non-premium cells often are not stable. I have been unable to determine why, but think it has something to do with the quality of raw materials, and perhaps, something to do with the sealing method of the cells. I have cells that tested (on a ZTS tester from www.ztsinc.com ) at 100% initially in April 2006, that are now testing 10% as of November 2006. While I have only observed this with 4 different brands of cells (Radio Shack is not one of them) I don't believe it to be unique. It is quite possible to end up with cells of different states of charge when you purchase them. If all of the other conditions are met, you can end up with cells that "rapidly vent with flames."

This does not seem to be the case with LED lights. We have not been able to get an LED light to vent or flame under the same conditions. I believe the LED just breaks the circuit when the voltage drops below a certain point.

To protect yourself against this, I recommend using premium cells in multi cell applications. If you want to try the "cheaper" cells, it is a good idea to load test them (ZTS tester or simulate a load with a resister and measure the voltage under load and compare it to a premium cell under the same conditions), then REMOVE them from the light when the light dims or goes out. If the circuit is broken, the reaction stops.

To summarize, we need non premium cells, a load that is high enough to heat up the chemistry within the cells, ambient temperatures high enough to sustain the chemical reaction, and a connection that is not disconnected, and of course 5 - 20 minutes for the reaction to complete.

To protect against this, when the light goes dim, remove the cells from the light, and use premium cells in multi cell applications.

Here is NewBie's test thread on CPF.

Tom

Was that with any LED light or regulated/semi-regulated? Many LED lights use a regulation circuit which probably does turn off the light at some point, but will the naked LED bulb do the same?

That question asked, the only non-regulated lights I have 2D & 3D MagLights in which I changed the bulb to an Epieon LED.

Like any diode, the LED itself has a "forward voltage" rating below which current through it drops off rapidly. So theoretically an unregulated "dumb" LED light would be less likely to cause this failure than an incandescent one.

Silverfox:
Are the BatteryStation 123's safe or not? They are among the cheapest batteries that I have seen (with the CPF discount), but they are only about $0.25 cheaper than Surefire batteries, which you categorize as "premium".

Has anybody measured the current and temperature of the premium cells as they are discharged? Perhaps they have a thermostat that disconnects the battery as it overheats.

Excellent! Great to have input from some of the CPF experts on the topic, like you and Newbie. Let me just say that I am always impressed by the professionalism and care that go into the reviews, write-up's, and tests on CPF. This ETS post is no exception. I hadn't visited CPF in a long, long time but after this topic came up on ETS, I started visiting CPF more regularly since I'm a flashaholic-in-denial. But darn it, my flashlight collection has already multiplied in the short time that I've been on CPF again. *sigh* But I digress.

The 100% vs 80% differential between the two cells isn't much, plus your observation of how variable the capacity of unused non-premium primary cells can be in a short time is quite alarming to me. So, even grabbing two brand new or unused CR123a cells from the same non-premium batch is not necessarily safe. Yes, surprising and worrisome indeed.

Anyway, thank you, Silverfox, for your input to ETS. Any reservations I might have had about these guidelines due to lack of data are now resolved.

Hello Ras,

There has been no venting, regardless of how hard we try, with LED lights.

All of the venting has been in conjunction with incandescent lights.

Tom

Hello Harrkev,

Is any battery "safe?" <img src="/images/graemlins/smile.gif" alt="" />

Kevin at BatteryStation is testing all of his cells before they ship. He is only shipping cells that test 100% on the ZTS tester. I believe (and know first hand on a limited basis) that the odds for getting good cells from BatteryStation is extremely good.

Keep in mind that there have been problems with all the brands of batteries. They are extremely rare with the premium brands, but they do occur. The ZTS tester came into being because photographers were having problems with their batteries and wanted a way to insure that their equipment would work when they wanted it to.

The ZTS tester is not perfect. As a matter of fact, it seem to lack precision. However, it does a good job of matching cells to keep you out of trouble.

Tom

Hello Arney,

NewBie tested a bunch of cells that were new and mismatched. I provided him with cells that tested on the ZTS tester from 0% to 80%, and he tested several samples of each. In each test, he coupled a lower tested cell with one that tested at 100%. There were no cases of venting during these tests. To get them to vent, he had to discharge the cell roughly 25 - 30% and couple it with one that was at 100%

I am not sure why this is so. We are still working on this.

I do suggest that in incandescent lights that drain the batteries in about an hour, that you start with cells that test at 100%. There is no reason to add any additional risk to your flashlight use.

Tom