ROAR of the Pelican (CR123 Explosion during use, firsthand account)

[hank]

The main lesson about both primary and rechargeable lithium batteries:

-- if these get hot, they vent flammable material plus oxygen
-- if one ignites it will ignite others nearby
-- normal fire extinguishers will not put the fire out
-- the pressure produced is enough to burst a sealed metal cargo container

e.g., from Google's cache:
http://209.85.173.104/search?q=cach...lithium+battery"++"cargo+container"++aircraft

 

[Gary123]

To be on the safe side, consider single-celled flashlights.

I thought about this. I have numerous double cell lights and have just purchased a 2 x 18650 light. If this one blew ...

Is it a good idea to have a battery tester to test the charge on multiple (rechargeable) batteries that go into the same flashlight? Would a battery tester tell you everything you need to know? Or could a tester indicate that two batteries are the same but they might still explode because they perform differently under load?

Can anyone recommend a good battery tester? I'm concerned about this thread and am inclined now to always test the batteries in flashlights that have multiple rechargeable lithium batteries.

 

[Brooke]

Hi Gary:

A Battery tester will NOT tell you. For example suppose you were going to use one "D" cell and one "AA" cell in the same flashlight. The battery tester tells you both are at 100%.

There is no tester on the market that can really test the primary "123" cell. I have a ZTS and have studied their patent. If the tester works like the patent then it tells you something but not capacity.
http://www.prc68.com/I/BatTst.shtml#MiniMBT

If you're using rechargeable "123" cells then a tester like the Triton2 that measures actual capacity and at about the same current as the load will tell you a lot. It also will tell you the capacity of a primary cell, but that's not much good since when it's done you need to throw away the cell.
http://www.prc68.com/I/Triton2.shtml

I sure would like to see a data sheet for the batteries that exploded. What is their max discharge rating? What's their capacity at that rating?

Have Fun,


Brooke Clarke

 

[Spence]

Lunar-
Thanks for your incredible, eye-opening posts and account of your trajic event. Hope you have a speedy recovery, it could have been so much worse, thank God for small miracles, in your case.

Battery Station, Kevin-
I personally want to thank you for your rapid and professional response, You are a class-act, all the way. I've run a ton of your 123's through my lights (including 2 Pelican's) with no problems. I don't expect this to change my M.O. because I always put my volt meter on my cells, coming and going. My question: How much difference in voltage do you consider safe in both primaries and protected rechargeables?
Thanks again.
Spence

 

[Gary123]

And to reiterate my question above -

If a battery tester can't tell you everything you want to know to avoid explosions, could it still help? Is some sort of testing better than none? Would it help to test batteries again after they have been used for a while? Would one need different testers for primaries vs rechargeables?

Should I forget about that Wolfeyes 3 x Q5 running on 2 x 18650's?

 

[Brooke]

Hi Gary:

Yes.

My first thought after reading the account of what happened was that there was a weak cell, a cell with a capacity considerably lower than the others and that when it ran out the remaining cells drove it into reverse polarity operation. This can happen with any multiple battery system.

But now I have a new theory since this was a new flashlight with new primary 123 cells. Extremely unlikely that they would have different capacities. There was a cell that was not sealed properly. It was either already leaking (I've heard about this) or was a defect.

If this was the case then testing a primary cell with a load current the same as it would see in the actual application would cause the cell to vent.

It's interesting that when the No. 6 Dry Cell (6" tall 1.5 V battery) was popular the common way of testing them was to measure the short circuit current (now called Flash Amps). For more on that see: http://www.prc68.com/I/No6.shtml#FA

I wouldn't recomend measuring Flash Amps on a 123 cell using a pocket meter, but doing it in way where a cell explosion would not be harmfull might also be a way to screen for leakers.

 

[e2x2e]

Wow that is scary. Yet another reason I'm not a big fan of CR123 cells. With the incredible advances, we can do more and better things with regular old AAs.

 

[Gary123]

Brooke:

So will it do me any good to buy an inexpensive battery tester for when two batteries go into the same light?

What would you recommend? Your responses above are a little confusing to me.

 

[SilverFox]

Hello Gary,

Since this incident occurred during early summer of 2006, I would suggest you go back in the archives and check some of the other threads related to this. There was a whole series of testing done to understand what happened, as well as a "recipe" developed to duplicate this explosion. About the same time frame, the ZTS tester was put under the microscope and you will find those discussions also illuminating.

The 18650 is a Li-Ion cell. We have found that you can obtain a reasonable idea of the state of charge by simply measuring the open circuit resting voltage of the cell. You want your cells matched when fully charged. They will tend to fall out of balance during the discharge, but that is normal. You want to limit the low end of the discharge voltage so you don't over discharge them, but understand that in a multi cell application, the cells will probably not be in balance at the end of the discharge.

Fully charged, balanced cells should have voltages within around 0.05 volts of each other. I prefer mine to be within 0.01 volts. Using these precautions should give you trouble free use of your Li-Ion powered light.

Tom

 

[SilverFox]

Hello Brooke,

Welcome to CPF.

I am afraid your venting theory doesn't work out.

There appears to be some quality issues with some cells made in China. Inconsistencies have been observed from cell to cell. I am not sure if these were from the same batch of cells, or differing batches. At any rate, it seems that the least expensive CR123 cells have a less stringent QC program that often leads to unexpected behavior such as rapidly venting with flame.

One of the unexpected features of these cells is that they seem to go to sleep with age. We believe a passivation layer forms within the cell, limiting its ability to handle current. These cells tend to get overly hot during discharge. While we don't know what causes this, it is often mentioned in conjunction with comments on the purity of raw materials used.

It just so happens that the ZTS results are based on high quality cells and it is very good at finding "marginal" cells. The general procedure is to test the new cells and if they don't register 100% on the ZTS, set them aside for single cell use. Don't use them in a multi cell application.

In addition, if you use the same cells that were used to determine the percentage points on the ZTS tester, you will get reasonably accurate test results. Their data base was set up by first discharging cells completely to determine their capacity, then by using a series of partial discharges to set their percentage points.

I will be the first to admit that the ZTS is not perfect, but I believe it gives a person valuable information about their cells.

Tom

 

[bwaites]

Thus spake Zarathustra!!

 

[PlayboyJoeShmoe]

Every now and then I test my 123 lights for "flash amps" I look for 2 (or 3) with very similar FA.

I will always put the slightly higher one closest to the bulb/LED.

On a multimeter with a 10A scale put the red lead into the 10A plug in, touch both ends of the battery you are testing just long enough to get to where the number stops climbing fast. Generally speaking I'll toss any 123 cell with 1A or less.

I'm having a bit of NimH trouble just now and have seen 10+ FA on Kodak LSD and 11+ on Eneloops. I don't think there is any question that a modern LSD NimH cell is up to the task asked of it!

Anyhow I like to try and catch problems before they get bad!

 

[Gary123]

Every now and then I test my 123 lights for "flash amps" I look for 2 (or 3) with very similar FA.

I will always put the slightly higher one closest to the bulb/LED.

On a multimeter with a 10A scale put the red lead into the 10A plug in, touch both ends of the battery you are testing just long enough to get to where the number stops climbing fast. Generally speaking I'll toss any 123 cell with 1A or less.

This is good information. Thanks - I like putting the higher one closer to the head - but is this only when the head is the positive side.?


and Tom: Since you haven't closed this post, I'm hoping you'll forgive an inquisitive mind: If two batteries are within tolerances, is it a problem if the capacities don't match? (Same battery - different manufacturers; same battery - same manufacturer - different batches, 2200 mAh with a 2400 mAh)?

Thanks

 

[SilverFox]

Hello Gary,

If you only run the batteries part way down... you might be able to get away with it, but it is risky.

You are never supposed to mix capacities or brands of cells when using cells in series. You can get away with it in parallel applications, but not in series.

Tom

 

[LuxLuthor]

Remembering this thread which was recently linked from battery section topic.

 

[Black Rose]

Just finished reading it. Man, what an eye opener.

I'm glad I only have a couple single cell CR123 lights.

 

[leon2245]

Really wish I would have read this before buying so many 2x 123 lights. Especially considering the surefire batteries I bought on ebay came in a ziplock baggie. From now on only S.F.'s direct from S.F.

I know it's an old thread and not likely to happen now, but if I had it to do over again, it'd be more AA's and 1x123's. The advantages aren't worth the risk, to me.

 

[toolpig1]

I never read this thread when it was "fresh", but I have to say that lunarmodule is one of the best writers on this forum. Thanks for a seriously eye-opening account.

 

[ZMZ67]

Really wish I would have read this before buying so many 2x 123 lights. Especially considering the surefire batteries I bought on ebay came in a ziplock baggie. From now on only S.F.'s direct from S.F.

I know it's an old thread and not likely to happen now, but if I had it to do over again, it'd be more AA's and 1x123's. The advantages aren't worth the risk, to me.

A few precautions and a little common sense have quelled my fears regarding multiple cells.Keep in mind these incidents are rare and usually have contributing factors like mixed cells.I do think it is advisable to get a ZTS tester to check for dead/low cells and avoid the cheap $1 cells from China.It is also important not to mix different brands or old cells with new.
Even name brand CR123s are as cheap or even less than lithium AAs when purchased online.CR123s provide more power and allow lights to be more compact than AAs.

 

[Monocrom]

A few precautions and a little common sense have quelled my fears regarding multiple cells.Keep in mind these incidents are rare and usually have contributing factors like mixed cells.I do think it is advisable to get a ZTS tester to check for dead/low cells and avoid the cheap $1 cells from China.It is also important not to mix different brands or old cells with new.
Even name brand CR123s are as cheap or even less than lithium AAs when purchased online.CR123s provide more power and allow lights to be more compact than AAs.

A big +1 on all of that.

A few simple rules of thumb to keep in mind when dealing with CR123 cells, and a ZTS battery tester, go a long way.