Thursday, January 24, 2013

Boeing 787 Battery Fire: Short Circuit and Thermal Runaway in One of 8 Battery Cells

From Flight Global (1/24/2013; emphasis is mine):

NTSB finds signs of short circuit, thermal runaway in JAL 787 battery failure

The Boeing 787 battery involved in the Japan Airlines incident on 7 January reveal signs of a short-circuit in one cell and thermal runaway that led to a fire, says the US National Transportation Safety Board.

The NTSB's first press conference 16 days after the JAL battery failure and fire in Boston also revealed that the investigation could still be far from discovering a root cause.

"It is really hard to tell how long the investigation will take," says Deborah Hersman, NTSB chairwoman.

"We have to understand why this battery resulted in a fire when there is so many protections designed into this system," she adds.

The short circuit was found in cell five of the eight-cell, 32V battery that starts the auxiliary power unit and is located in the aft electronic equipment bay in the 787.

But the NTSB has not yet been able to narrow down a set of possible root causes that it is ready to reveal to the public.

"We are seeing symptoms," she says. "We know there is something wrong here. The short circuit, the fire, these are all symptoms that something is wrong."

Hersman says a troubling complication of both the JAL 787 incident and the All Nippon Airways 787 battery event was the timing of the battery failures within 100 flight hours of service by both aircraft.

"We do not expect to see events like what we saw on the 787 in the battery system," Hersman says. "Two battery events in two weeks in the early flights of this aircraft are not what we expect."

The incidents prompted the FAA to order United Airlines to ground all six US-registered 787s on 16 January, an action that triggered a global grounding of the aircraft less than 15 months after it entered service with ANA in October 2011.

So far, the Japan Transport Safety Board and the NTSB have agreed that there is no evidence yet that either battery was over-charged when it failed, but they are continuing to examine the data. Hersman says, for example, that investigators are still considering if it was possible for one of the eight battery cells was over-charged, but not the overall battery.

Another critical part of the investigation is considering the certification process for the 787 batteries.

"These events should not happen as far as design of the airplane," Hersman says. "There are multiple systems to prevent a battery event like this. Those systems did not work. We need to understand why."


TechDud said...

Cosmic-ray damage, "whiskering" phenomenon (see NASA-Goddard site concerning "tin-whiskers" and other metals), or both?

TechDud said...

Better yet, how about thermal runaway?
Looks a little cramped in there.

Anonymous said...

Or what about manufacturing errors?? Where are these batteries produced, quality control checks? Good place to start. Over charging one cell in a battery..and not the others? HUM....

Anonymous said...

Thought I read these batteries are made in Japan. True?

Anonymous said...

The control circuitry should charge/discharge and monitor each cell individually. If it doesn't, this sort of thing will happen - especially if each cell of the battery is not at the same voltage on installation/supply (which they ought to be).

If the control circuitry was not specified to do this ( you can get away with not balance charging LiFe cells sometimes) then it's a design flaw!

This could also be a battery flaw although that should be picked up in commissioning the batteries. The batteries should also have some protection from thermal runaway etc in how they are stowed!

Anonymous said...

Why did they cram the circuit boards right next to the battery cells in the same compartment?

I take it that once these type cells begin overheating there is no stopping it so taking a failing cell out of the loop is meaningless.

Wiring still looks suspect to me.

In the meantime...

Anonymous said...

The Boston 787 fire resisted suppression and was allowed to burn out. There was smoke in the cabin, indicationg permeation of the entire aircraft. Since the effluent from a lithium-ion battery burn is corrosive, the whole aircraft needs evaluation, IMO - particulally the fly-by-wire electronics.

Ageless Yankee said...

Tin whiskers are exceedingly thin conductors, roughly one molecule wide. A short across one or more cells would result in the tin molecules being vaporized and the short being cleared. The battery wouldn't even notice.

Cosmic rays produce bit flips in memory. They don't short battery cells.

Both NTSB's seem to be engaged in misdirection. The statement "there is no evidence yet that either battery was over-charged when it failed" and the speculation that only one of 8 series cells could have been overcharged is chaff. They keep releasing information about what didn't happen to the battery, while remaining silent on the absolutely critical issue of charge and discharge profiles.

The questions to ask now are:

1. What are the battery manufacturer's maximum ratings for charge and discharge?

2. Does the onboard battery charger for the battery remain within those maximum ratings as well as any other limitations under all circumstances?

3. Can the airplane draw more than rated current from the battery in any circumstances?

4. What service, if any, does the APU battery get when the airplane goes in for maintenance? If the battery charge is "topped off", is it done with a battery charger specifically designed for this cell type and does it remain within mfr's maximum ratings/limitations throughout the charging cycle?

5. How many APU starts do the failed batteries have on them relative to others in service that haven't failed?

6. Does the current draw of the APU starter/ generator exceed any of the battery mfr's maximum ratings and/or limitations during an APU start operation?

Old battery aphorism: "Most batteries don't die, they are murdered".

TechDud said...

Nice post!

However; " A short across one or more cells would result in the tin molecules being vaporized and the short being cleared. The battery wouldn't even notice."

Look this up and you will find that tin whiskers can contribute to Metal Vapor Arcing currents as high as 200A.

... also, sorry for mentioning "thermal runaway"; that was already in the article. My bad.

TechDud said...

... also
"Furthermore, whisker formation is observed on Sn thin-film electrodes"
quoted from:

better yet:
"Lattice Energy LLC Growth of conductive metallic “tin whiskers” and random electrical shortingevents implicated in Toyota’s uncontrollable vehicle acceleration safety problems Whisker shorting issue analogous to metallic dendrite shorting events that can causeLENRs which can then trigger super-hot, fast-burning metal oxidation fires in Li-ion batteries "
quoted from:


TechDud said...

there are many people at Washington State University, researching the positive use of whiskering phenomenon to increase Lithium Ion battery capacity, reliability, etc.

"The problem with tin (as opposed to graphite) is that when you make electronics (or battery guts) with it in a cheap and efficient manner, the tin grows what's called "whiskers," which are spiky things that can cause short-circuits and other damage and after 60 years of trying nobody's been able to eradicate them in a mass-production environment. The WSU researchers figured that if you can't beat 'em, join 'em, and they've managed to just get the tin whiskers to grow in an orderly manner at the nanoscale, providing lots of surface area to store charge without causing lots of problems."
quoted from:


TechDud said...

... don't totally discount Cosmic Rays, as Solar output has declined since the last solar maximum, increasing the amount of Cosmic Rays reaching Earth. This would be more pronounced at altitude.

TechDud said...

... "While it is often thought to be associated with internal shorts and electrical arcing within a somehowdefective cell, some battery manufacturers will admit privately that this peculiar failure mode is not well-characterized and very poorly understood --- most of them are presently at a loss for ideas about exactlyhow to definitively mitigate such a problem"
"This additional new source of concern about the safety of advanced Lithium-based batteries has arisenbecause, in the course of our company’s ongoing R&D efforts, Lattice has applied the Widom-Larsentheory of Low Energy Nuclear Reactions (LENRs) on a practical level to try to help better understand thepossible role of nanoscale metal dendrites and nanoparticles in certain types of failure modes that mayoccur in smaller Lithium-based batteries as well as in extremely large, multi-thousand-cell battery packsutilized in all-electric vehicles and some military applications.In May 2010, academic researchers at Oxford University published a new and we think important paperthat many believe implicates the involvement of Lithium metal dendrites in a significant number of Li-ionbattery failures (please see R. Bhattacharyya et al., "In situ NMR observation of the formation of metallicLithium microstructures in lithium batteries," Nature Materials 9 pp. 504 - 510). What is of great concernfrom a safety standpoint is that nanoscale internal metal dendrites that are prone to shorting-out can growspontaneously over time as a given battery ages and goes through many charge-discharge cycles.A battery pack may well be perfectly safe during the first months of ordinary use; however, dendrites andother types of nanoparticulate structures grow inside over time, increasing the probability of dangerousinternal electrical shorts as the battery ‘ages’. The problem is that nobody in the world has any realworking experience with large multi-cell Lithium-based battery backs that have endured hard usage andvibration for periods of many years. Also, nanoscale internal metallic dendrites can potentially form andgrow in almost any type of Lithium-based battery chemistry.Approaching battery safety from perhaps a different technical perspective than many scientists, we havebecome increasingly concerned that some present/future Lithium-based battery chemistries couldpotentially be susceptible to rare, but potentially very damaging occurrences of LENRs in isolatednanometer to micron-scale regions within some failing battery cells. Cell field failures arising fromnanoscale internal shorts/arcs are thus very worrisome with regard to potentially triggering LENRs thatcan in turn readily initiate macroscopic, catastrophic thermal runaways."

quoted from

LENR = Low Energy Nuclear Reaction ????

TechDud said...

how this could directly, or indirectly relate to Cosmic Rays, i do not yet know.
Perhaps another reader here will take the time to comment.

TechDud said...

...perhaps if "lightening strikes the same place more than once or twice", that could be a similar failure mode possibly involving plasma physics and exothermic consequences. I do not yet know about the Widom-Larsentheory of Low Energy Nuclear Reactions (LENRs). Once again, i hope somebody will enlighten the other readers.

Thank you again, for faithfully dispelling ignorance, arevamirpal::laprimavera! A deep humble bow to you sir.

TechDud said...
quoted from johnnymotel: "battery cooling:
I saw a picture of a new battery system and my immediate reaction was the cells seems awful close. Then I came across an article that talked all about active cooling of this type of battery. From the photo I saw, I didn't see any active cooling."

quoted from Dave Hilling: "RE: battery cooling:
I cant speak to this particular battery but in many aircraft its just a small cooling duct that enters the back of the part from the air conditioning system. They are not always very big. That being said though AC doesnt generally run on the ground and it may just be a small ducted cooling fan which doesnt help a ton if the area is already 35C+"

quoted from ilmari: "Re: battery cooling
I've been wondering about thermal management in that bay, is it pressurized, if not, how does thinner air affect the thermal management?"

quoted from Real Ale is Best: "Re: battery cooling
Another thing to think about is the AC cooling is driven by the engines. When the battery is most needed, guess which component is most likely not running?"

TechDud said...

NASA pulled a F1 rocket engine out of mothballs at the Smithsonian to test fire it's Gas Generators, BTW.

fru said...

>> In the meantime...

This is really wrong info.
Yes, tin whiskers are a problem for electronics.
Yes, lead has to be removed definitely, there's no going back to lead.
Yes, US manufacturers are forced to produce ROHS, as are all WW electronics manufacturers who want to sell to EU.
Yes, manufacturers have to switch entirely to avoid contamination of ROHS products with lead.

But the rest is BS :

Wrong. Today, we use silver as a substitute for lead. The problem is not the solder, but the coating of surfaces.
>> There is a conformal-coating under development, that promises to "contain"... blablabla
Wrong. conformal coatings from 20 years ago also contain them, this has been validated and is in widespread use (especially in automotive)
>> There will NEVER be enough time to prove that tin whiskers will not grow...
Take the time and relax. A product dev is 2 years, a few hours and a microscope with tangent lighting are all that's needed. Yes, every HW dev doesn't do it today. I also don't do it every time.
>>THERE IS NO CURE to prevent tin whiskers from growing, unless you get rid of the pure tin
Wrong. Electroplated pure tin "sometimes" grows TW. Dipped surfaces quasi never grow TW. There are validation procedures for that.
>> The pure-tin-plated electronics are inside not only consumer products but industrial ... nuclear- and conventional-power-plant control systems, automobiles
For automotive, the industry was for many years exempted of ROHS, but made the switch anyway.
For nuclear, military, aero, space etc... there's no excuse. These people have to use reliability certified components, which never use electroplated tin, but rather ENIG (electroless nickel+immersion gold) If they do not respect their own standards, due to the cost lowering COTS politics, that's their fault. Don't expect from a non certified product to perform the same as a proper engineered one.
That problem is the same regardless of ROHS, anyway these industries are exempted of ROHS compliance.

To summarize, this US BS about mixing two different problems of ROHS and tin whiskers is really costing credibility for the US industry.

fru said...

>> Look this up and you will find that tin whiskers can contribute to Metal Vapor Arcing currents as high as 200A.

Yes, tin whiskers can cause prolonged arcing. But only in vaacum. Not in a plane.

Anonymous said...

Interesting fru:

I guess coating circuit boards with silicone spray does nothing except for some moisture proofing and maybe preventing stray crosstalk.

Whiskers seem to be caused by electromagnetic attraction and once whiskers are established, let alone grow in length, can go airborne spreading elsewhere to cause havoc around and in electronics. So I read.

TechDud said...

Wow, this is turning into a good discussion!

>> There Is No Substitute For Lead In Solder
Ask NASA, they will tell you "3% minimum".
That's why they have ROHS6. Silver's nice indeed, yet it can whisker &/or form dendrites too.

"Yes, tin whiskers can cause prolonged arcing. But only in vaacum. Not in a plane."
I note that it need not take prolonged arcing to start a fire in lithium batteries, and this phenomenon is not confined to tin.

To remove most whiskers, i use the method called "cleaning". :)

TechDud said...

That is to say, cleaning is an effective method to remove most whiskers and other debris, prolonging electronics MTBF. Not practical for the consumer as there are "no user-serviceable parts". Not practical for batteries, either.

They could be thought of like miniature snakes.
We'd have to ask Samuel Jackson on that one, yet i don't believe he would be pleased to hear about it!

Ageless Yankee said...

Thanks to all of you who pointed out various tin whisker behavior that could damage a high current circuit. That said, I'm disinclined to view vapor arcing as a likely cause given that each cell is only 4 volts.

Additionally, I'm disinclined to buy the entire tin whisker hypothesis because the batteries are relatively new. My experience is that tin whiskers typically take a couple of years to manifest. Anybody seen it happen in say, months?

Finally, the news is out this morning that the NTSB has cleared the battery, so we're back to charge/discharge profile hypothesis. The 787 has two distinct categories of software applications on board. The first is "loadable software" which is designed for easy upload to the airplane. The other type is called "firmware" and generally requires special equipment to load it into "Read-Only-Memory" of the box (a.k.a. LRU/LRM) (These days, "ROM" can be written to electronically). The 787's functionality, behavior, and performance is largely determined by which versions of SW have been loaded. Not all versions of a SW applications "play nice".

An interesting question to ask at this juncture would be, "Were there any loadable software or firmware updates during the service period that were related to the APU starting and charging circuitry?".

TechDud said...

"Finally, the news is out this morning that the NTSB has cleared the battery"

Will you provide a link to this new information, please?

" The other type is called "firmware" and generally requires special equipment to load it into "Read-Only-Memory" of the box (a.k.a. LRU/LRM) (These days, "ROM" can be written to electronically)."

There's a possible path to failure in a multiplicity of ways.

Can you pinpoint any other possible "murder weapons"?

Anonymous said...

Cosmic radiation is at it's worse at higher altitudes and over the poles. A radiation "hit" can involve millions of electron volts and thereby easily break down the insultaive properties separating the two "poles".

The 787 is pressurized to fly at higer altitutes than any previous commercial passenger aircraft and JAL and ANA would both be flying the polar route.

It would/might only take, initially, a microscopically small short to begin a localized thermal runaway and that, in turn, might take hours to grow, spread, to a hazardous level.

rsgoldsale said...

just think it is a example, i think everyone should have their own opinion.

Post a Comment