Lithium-Ion Batteries Linked to Airplane Fires

smellsofbikes writes "The National Safety Transportation Board thinks it's possible that lithium-ion batteries caused a fire that destroyed a United Parcel Service airplane on Feb 8, 2006. The FAA already bans non-rechargeable lithium batteries from air shipment because aircraft don't carry fire suppression equipment capable of extinguishing lithium fires. The interesting thing is: these batteries aren't being used or charged, they're just being shipped: spontaneous battery combustion. Is this something that happens in the back of computer stores, or just on airplanes?"

Comments

Environmental stress

[morcheeba]

It's not just spontaneous, it's environmental stress. A cargo hold is a cold, low pressure, high vibration environment . This may be the first time a newly-made battery is exposed to these factors, causing infant mortality flaws in manufacture to become aparent. Even after the infant mortality portion of the bathtub curve, reliability calculations typically rate one hour of cargo flight time as worth 10-20 hours on the ground. That flight from china may be equal to 10 days on the ground.

Re:Environmental stress

[schnikies79]

The main cargo area on UPS jets has standard cabin pressure. The lower cargo areas on select planes are not pressurized.

Nope, it happens in plenty of places

[Anonymous Coward]

Flashlight geeks have been dealing with this issue for a while.

http://www.candlepowerforums.com/vb/showthread.php ?t=78843

http://www.candlepowerforums.com/vb/showthread.php ?t=124776

There have been several documented "venting with flames" of primary CR123A batteries. Rechargeables seem to be a lot more stable, occasional Dell laptop conflagarations notwithstanding.

Re:Nope, it happens in plenty of places

[dpaton.net]

A majority of CR123s aren't designed for contsant discharge at a relatively high rate. They are marketed to the photo market, where there are pulses of high power and long periods of very low draw. They do function at higher draws, but with reduced lifespan. This is hidden deep in the spec sheets, where the pulsed current recovery and discharge profile math is. I'm not terribly surprised that people have problems with lithium primary cells (NOT Li-po, Li-Ion, or any of the rechargable Li chemistries) in use for high current loads like the high power miniature flashlights out there like the Pelican M6 (the example cited in the second CandlePower link). The Xenon bulb version will suck the power out of a pair of CR123s in 1 hour. Calling the batteries 1300mAH (an average, according to Google), that means they're being loaded to about 1.3A each. That's a ~1C discharge rate. Most cells I found data sheets on didn't show a 1.3A discharge curve, instead showing a 1A curve or 1200mA pulse discharge measurement, using a 3s on / 7s off (30%) duty cycle. 10% can mean a lot in these cases. Odds are a lot of those cells are being used on the edge of or well past their design envelope. Beating up batteries like that can cause trouble, especially for cells that are fragile. Of course, not all are. The Energizer E2 photo lithium CR123 shows a capacity of 1.5AH and a 1000mA discharge life of 1.2 hours. It's probably the one used by Pelican to reach the rating of their flashlight, even if it looks like they did push the cells a little past their design limits.

Lithium primary cells generally do not have construction compatible with fast discharge. Often it can be gotten away with if the discharge is under 0.6C or is of a pulsed nature. Continuous discharge will kill them tho, a flaming, explosive kill.

Batteries have ever-increasing power densities, and deserve respect from designers. Just tossing 123s in is a BAD idea IMO. I was an engineer on a project where someone did just slap one in without consideration. When we put the test unit through its paces, blammo. Pulling 2A out of a 1.5A battery for 7 seconds is OK in NiCads and NiMH cells and even rechargable LiPoly prismatics if you know what you're doing. This was a dime store photo battery, and it went off like a small cannon after a few seconds.

People don't think about the design envelope for batteries as much as they should any more. It's unfortunate.

My US$0.02 as an engineer.

UPS = Ooops

[the eric conspiracy]

Given how some of my UPS packages arrive looking like they were dragged to my house behind the truck, I would say that it is pretty likely that UPS is doing things to the batteries that my computer store doesn't.

Re:UPS = Ooops

[CheddarHead]

Back when I was in college there was a brief period where I payed my rent and got beer money by working at a UPS facility. I worked loading UPS semi trailers with packages. The packages would come off of a conveyor belt, and our job was to load the truck as fast as possible.

To make a long story short, we were not particularly gentle with the packages. In fact if you saw the way the trucks were loaded, you'd be surprised at what good conditions your packages are in. I still use UPS, but I always make sure that things are packed very, very well.

Re:squished?

[ResidntGeek]

Nope. Lithium is an alkali metal. Alkali metals ignite on contact with water. The more active ones (Cesium most of all) violently explode. I imagine a small puncture in a battery could let in enough atmospheric water vapor to ignite a battery.

Re:squished?

[treeves]

That would be a plausible explanation if the battery contained elemental lithium. They don't. They contain compounds of Li.

Warning...

[CaseM]

Oxygen linked to fires...time to take ACTION!!

Re:Too slow?

[cat6509]

It is the NTSB that researches things like this, not the FAA per se. They are very methodical and precise with their work. They are slow to publish thier findings, this doesn't mean they are slow to identify the cause, just very carefull that they have come to the correct conclusion. Check out the NTSB aircraft accident database, this contains detail over every aircraft accident reasearched by them for several decades ( 1962 ) http://www.ntsb.gov/ntsb/query.asp

Fragile

[Archangel Michael]

UPS Dictionary says .....

Fragile (fra-gil-lay) from early French n. To toss about with reckless abandon.

lipo fires

[heli_flyer]

Electric RC flyers have been dealing with this issue for a while.

Here is an informational thread about lipo batteries:

http://www.rcgroups.com/forums/showthread.php?t=20 9187

Uh... "Vent with Flame" anyone?

[tlhIngan]

One of the failure modes of a Li-Ion battery is what the industry calls "vent with flame", or what everyone else calls, a fire. (A very spectacular one, at that - not just ignition, but the fire actually shoots out like a jet).

Li-Ion batteries are extremely volatile and sensitive, which is why good batteries have a variety of protective circuits on them (or can have) - e.g., physical distortion (detects if the battery balloons), over temperature (charging/discharge), over current, unsafe low voltage (if the battery voltage falls too low, you can't charge it safely), and many more. That's also why their charge regimen is so complex (charge at constant current to ~90% capacity, then constant voltage charge to 100%. Then stop all charging until capacity is around 90% again, then restart CV charge - this is why the first 80% can happen relatively quickly, while the last 20% can often take as long as it took to get to 80% in the first place) since they need charge controllers and "smart chips" to monitor the state of the battery.

Usually these events happen when the battery is actually used, but there isn't anything to say that it can happen otherwise. Those protective circuits require power, and they get their power from the battery while outside the device. And since you cannot store Li-Ion batteries discharged very well, they are often charged at the factory, during assembly and final sale. A nice short somewhere along the line and battery will vent with flame.

There's a reason why most LiIon batteries have hard to get at terminals or come with protective covers. It's not for convenience, but more for during storage/shipping, so the terminals don't get shorted.

Oh yeah, those protective circuits are optional - not all batteries have every one (some may not need it or find a way to protect it in another way - battery distortion can be handled by having the battery having to fit in a slot - if it can't fit, well...). Third party ultra cheap batteries may have *no* protective circuits at all (hence those "Nokia Exploding Batteries").

Re:Uh... "Vent with Flame" anyone?

[mmclure]

Yes, but "below 50% charge" actually means "85% of the maximum voltage." For Lithium-Polimer batteries, the maximum voltage is 4.2V and 3.0V is "completely discharged" (in fact dropping lower than 3.0V runs the risk of permanently damaging the battery and causing "vent with flames" the next time you charge it.) 3.8V is considered the ideal storage voltage in R/C circles.

R/C circles probably have the largest percentage of failed Lithium Polymer batteries, because they are discharged at pretty high rates, and the protective circuits on most consumer lithium batteries are removed. Most failures are caused by the following:

  - Overcharging due to the use of a bad charger or using the wrong settings on a Lithium-Polymer-capable charger.
  - One cell of a multicell pack overcharging due to being at a higher initial voltage than the other cells in the pack. This is being mitigated with the use of "balancers" which connect to each cell and maintain equal voltages across the pack.
  - Physical damage due to crashes.
  - Shorting out the battery.
  - Allowing the voltage of a cell to drop below 3.0V.

Many of these have been ameliorated due to technological improvements in the last two years - for example the aircraft-side electronics now cut off the motor at a point where the battery voltage is still safe, chargers have been much improved, and battery packs now come with special connections to allow checking and adjusting voltages on a cell-by-cell basis.

R/C enthusiasts tend to treat Lithium-Polymer batteries with quite a bit of respect as far as storage and charging is concerned. Many will set up a cinderblock on a concrete floor, place the battery in one of the holes of the cinderblock (the holes in the block are set vertically) and cover the top of the hole with a plastic bag full of sand. In case of a "vent with flame", the heat will melt the plastic allowing the sand to cover the battery and put out the fire. Compared to a couple of years ago, however, the incidence of Lithium-Polymer fires has greatly decreased even as use of Lithium-Polymer batteries has increased.

This forum thread has a lot of information about dangers and safe handling guidelines for Lithium-Polymer batteries.

More on the destructive power of Lithium....

[Cherita Chen]

Check out these photos here of lithium polymer batteries (commonly used in r/c models) in action... SUPER FUN HAPPY BURN THE HOUSE DOWN TOYS!

Re:squished?

[Directrix1]

Let me just say that a Lithium-Ion battery can do some pretty nasty stuff. I had one out of my camera (a small Nikon digital) sitting on my bedside table next to my camera. One night I dropped it on the floor. I don't know what that did to it, but it started to bulge and become untouchably hot. I put it inside a pyrex container on the kitchen floor for the rest of the night in case it went poof. By morning it was fully discharged, but still had the bulge in it. I thought that thing was going to explode for sure, but luckily it didn't.