NTSB 3rd investigative update on Boeing 787 battery fire in Boston; battery not overcharged
20 January 2013
The National Transportation Safety Board (NTSB) released a third update on its investigation into the fire aboard a Japan Airlines Boeing 787 at Logan International Airport in Boston on 7 January. (Earlier post.) NTSB said that examination of the flight recorder data from the JAL B-787 airplane indicated that the APU battery did not exceed its designed voltage of 32 volts.
The lithium-ion battery that powered the auxiliary power unit has been examined in the NTSB Materials Laboratory in Washington. The battery was x-rayed and CT scans were generated of the assembled battery. The investigative team has disassembled the APU battery into its eight individual cells for detailed examination and documentation.
Three of the cells were selected for more detailed radiographic examination to view the interior of the cells prior to their disassembly. These cells are in the process now of being disassembled and the cell’s internal components are being examined and documented.
Investigators have also examined several other components removed from the airplane, including wire bundles and battery management circuit boards. The team has developed test plans for the various components removed from the aircraft, including the battery management unit (for the APU battery), the APU controller, the battery charger and the start power unit.
On Tuesday, the group will convene in Arizona to test and examine the battery charger and download nonvolatile memory from the APU controller. Several other components have been sent for download or examination to Boeing’s facility in Seattle and manufacturer’s facilities in Japan.
In accordance with international investigative treaties, the Japan Transport Safety Board and French Bureau d’Enquêtes et d’Analyses pour la sécurité de l’aviation civile have appointed accredited representatives to this investigation. Similarly, the NTSB has assigned an accredited representative to assist with the JTSB’s investigation of the 15 Jan battery incident involving an All Nippon Airways B-787. Both investigations remain ongoing.
Just because the "...APU battery did not exceed its designed voltage of 32 volts" does not mean it was not over-charged. All lithium batteries are supposed to follow a charging schedule, with a higher voltage at the start (perhaps 32 volts in this case), then slowing to a trickle charge after perhaps they are 70-80% recharged. If Boeing/GS Yuasa had a bad charging schedule, or a bad voltage reading, it would be easy to cause a thermal runaway.
--Keith
http://www.keithdpatch.com
http://blog.fuelcellnation.com
Posted by: Keith D. Patch | 20 January 2013 at 05:02 PM
@Keith - typical lithium battery charge protocol is constant current (say 100A) until pack voltage reaches maximum voltage at which point the charging algorithm switches to constant voltage (say 32V).
Let's say that they are using a typical Li-Ion cell with a 100% charge voltage of 4.2V - anything higher than that results in overcharge. That would mean that they are using 8 cells in series and limiting each cell to 4.0V maximum which should be very safe in theory as long as they are accurately monitoring the voltage of each cell in the pack.
I think it's more likely that there was a bad connection somewhere in the pack which resulted in overheating...
Posted by: Dave R | 20 January 2013 at 10:10 PM
If so, the question would become "how was this bad connection not detected in assembly and testing"?
If the connection failed due to e.g. vibration or thermal cycling in use, this means the battery design is inadequate. We can hope that it's an assembly flaw like the A123 packs which had the connection welded in the wrong place and had to be recalled.
Posted by: Engineer-Poet | 21 January 2013 at 07:59 AM
It is a good thing that this plane was grounded before an in-flight fire caused a fatal crash?
The cause or causes will be identified. Future batteries and their installation will correct whatever is wrong with the current version?
Posted by: HarveyD | 21 January 2013 at 12:03 PM
As I mentioned in a previous posts the problem might be not in the battery itself and not battery management system but rather in power system integration and distributed DC bus arrangements. So, under regular testing nothing wrong will show up but being integrated in the plane power system a battery can overheat depending how the loads are distributed and connected. That is why it is not that straight forward and nobody is blaming (although it would be so convenient)the battery company.
Posted by: CarCrazy | 22 January 2013 at 03:06 PM
@Dave R - My point is that maybe the battery charging schedule was too agressive, in order to make a force-fit match to Boeing's specified battery duty cycle design requirements. In your example, reducing the initial constant current of 100A to 75A would obviously be safer (less stressful on the battery), but perhaps resulted in an excessive recharging time period, and meant that GS Yuasa did not meet the design criteria. Thus, after cycling the battery multiple times with an overly-agressive recharging algorithm, it becomes prone to thermal runaway failure.
Or perhaps the problem was that limiting each 4.2V-rated cell to 4.0V maximum was too agressive, but allowed GS Yuasa to meet the Boeing design energy density requirements. Again, further derating to perhaps 3.9V as the operating maximum cell voltage would be safer, and perhaps avoid this type of battery problem, but would not meet the Boeing design criteria.
So perhaps the fault lies with an overly-aggressive design by Boeing?
--Keith
http://www.keithdpatch.com
http://blog.fuelcellnation.com
Posted by: Keith D. Patch | 24 January 2013 at 07:28 AM