Thermal runaway and electrical shorting have been found, but cause remains elusive; a whistle blower surfaces.
bf sv nation dreamliner probe update FLAGSTAFF, Ariz.--The probe into the battery fires on two Boeing 787
Dreamliners deepened late this week, and the story expanded in
unexpected direction as a whistle blower linked to an electronics supplier surfaced.
Federal investigators said Thursday (Jan. 24) that they still have
no indication what started
a fire in a lithium ion battery Jan. 7 in Boston, aboard
a parked and just-landed 787. Investigators with the National
Transportation Safety Board did announce three findings, however:
Fire was present
There was evidence of thermal runaway
There were signals of an electrical short
National Transportation Safety Board Chairman Deborah Hersman said, "We have not ruled anything out as a potential
factor in the battery fire; there are still many questions to be
She added, "One of these events alone is serious; two of them in
close proximity, especially in an airplane model with only about
100,000 flight hours, underscores the importance of getting to the
root cause of these incidents."
NTSB Materials Engineer Matt Fox examines the casing from the battery involved in the JAL Boeing 787 fire incident in Boston.
Investigators also released this timeline of the events on Jan. 7:
10:06 am EST - Aircraft arrived at gate in Boston from Narita, Japan
10:32 am - Cleaning and maintenance crew noticed smoke in cabin
10:35 am - Mechanic noted flames coming from APU battery in aft
10:37 am - Airport Rescue & Fire Fighting notified
10:40 am - Fire and rescue personnel arrive on scene
12:19 pm - Fire and rescue personnel report event was "controlled"
One aspect of the investigation that has been ruled out is the
impact of the battery's auxiliary power unit (APU), which was manufactured by Securaplane
Technologies Inc. in Tucson, Ariz. Investigators were at the
electronics company Wednesday testing the APU charging unit and
found no anomalies. The NTSB said the same investigative team
traveled to Phoenix to conduct an examination of the APU controller
at UTC Aerospace Systems.
But at the same time, a twist in the tale emerged, as a former
Securaplane engineer Michael Leon said he was fired from the company
six years ago for raising safety concerns about rushing
chargers to market out of spec. Leon filed suit several
years, but a judge ruled he was fired for misconduct. Leon has
appealed the case.
Securaplane's corporate parent, U.K.-based Meggitt, has said there
is no connection between his dismissal and any battery issues.
In Japan, an NTSB team examined the JAL B-787 APU battery
monitoring unit at Kanto Aircraft Instrument Company, Ltd., in
Fujisawa, Kanagawa, Japan. The team cleaned and examined both
battery monitoring unit circuit boards, which were housed in the
APU battery case, the NTSB said. The circuit boards were damaged,
which limited the information that could be obtained from tests,
the NTSB added.
It is my belief that LiIon batteries exibit positive thermal runaway. For a little less capacity LiPo batteries do not exhibit this thermal runaway problem. They probably were not around when the design and specifications were penned for the 787.
Is it possible that the failure is in an individual cell within the LCO product? Just what if, the charging process is montoring the average of all the cells within the makeup adjusting the charge and rate of charge to this set average read. If we have a cell or group of cells that go into premature death, the overheating would be brutal. Basically cooking them to death. What kind of monitoring system is being used?? One perhaps that would allow this condition.
I am not a battery expert but I was wondering about the effects of the vibration and normal shocks due to the flight and takeoff/landings on the battery. I do know that dropping a SLA (sealed lead acid battery) can cause a lot of damage and internal heating if plates short together, so I wonder about the robustness of the plane batteries...
But this battery had nothing to do with passenger entertainment systems.
I just don't think it's likely to be a load profile issue, because excessive load currents are always protected in well known ways. I also find it hard to believe that the charging circuit could be inappropriate for LiIon batteries, because that would have been a really simple problem to find and would have occurred during the multiple thousands of hours of testing.
What did they find to be the problem in those laptop batteries that burned up? Was it not the battery itself that overheated and failed internally, not caused by external faults?
I hadn't heard any inkling of battery problems during the flight test program, though other teething problems were documented. Apparently one battery blew up in a ground-based test fixture back in 2006 though. It's odd (but far from unheard of) that it would get all the way through flight test without issue and then have two incidents so close together after certification.
There was a time in my career where I had to do a lot of testing of products pre-launch. I followed test scripts, but I also spent time just doing strange and random things to try and break the stuff. I always found more problems with the strange and random part of my test program than with the formal part. Invariably, the ways that I broke the products would end up showing up in the field under conditions not anticipated during design.
I could certainly imaging that a plane full of passengers fiddling with the entertainment system added into the other power consumers could create a different power profile than was anticipated and tested for.
The NTSB site linked to in the article has some interesting info - it shows a report that has a photo of a battery cell with a damaged electrode - internal s/c.
The battery chemistry used is Lithium Cobalt Oxide. I would have expected the safter Lithium Iron Phosphate to be used (or other safer-chemistry) instead in this safety critical application.
I tried to see from the photos what (if any) power conversion electronics were in the battery box, but couldn't see any high-power, only low power circuit boards. Power electronics can ignite.
In other words, no new information, as far as I can tell.
My thinking all along has been that it's not a load issue, because presumably the load currents are protected by fusible links, breakers, or other standard measures. So the problem has to be either in the charging circuit, somehow violating LiIon charging protocol (e.g. continuing to trickle-charge when it must not), or some internal short in the battery pack?
The fact that the load device was ruled out as a cause is perhaps the least surprising possibility.