The National Transportation Safety Board (NTSB) is holding a press briefing at 230 pm (ET) today. We’ll have a running synopsis of it right here.
Deborah Hersman, Chair of NTSB
- Expectation in aviation is never experience a fire on an aircraft. In two weeks time saw two battery events and groundings.
- Significance of these events can’t be overstated.
- Been working since Jan. 7 to understand what happened and why
- Lithium ion battery experienced thermal runaway, short circuits and a fire.
- Still trying to find out why.
- 10:49 to 12:15 fire fighters fought the JAL blaze.
- Fire confined to an area within 20 inches on the battery.
- The batteries are unique to the Boeing 787.
- The main battery is the final power source should all other power fail.
- Minor faults found, not unusual following damage. Trying to determine the significance.
- Still trying to gather data.
- Boeing, FAA conducting more activities. NTSB is participating as well.
- Evaluating failure modes, manufacturing records for potential issues or trends, supplier audits.
- Tremendous amount of work going on around the world and around the clock.
- Still trying to analyze failure modes.
- [NTSB will post its PPT presentation on its website later today.]
- [Going through slides right now of components of battery construction]
- Independent examination in DC area for NTSB has been conducted.
- Examining electrodes individually and a number of other methods looking for contaminates or defects.
- We’ve done CT scan of the entire battery.
- We’re looking for signs of thermal runaway (uncontrolled runaway), short circuits, manufacturing defects that could be foreign materials or material defects.
- We have found some short circuits.
- There is bulging present, signs of thermal runaway and short circuiting.
- Still have lab testing on damaged battery and will do testing on an exemplar battery.
- We’ll look at airworthiness, design components, to evaluate the safety of the systems that are integral to the battery.
- FAA and Boeing are parties to investigation, Japan JTSB, GS Yuasa, Japan Airlines, France’s BEA, Thales
- Naval Surface Warfare also assisting; have used lithium ion batteries for decades
- JTSB in Japan is leading investigation on ANA 787 incident.
- Japan ATC did see smoke around ANA aircraft.
- We have to understand why there was a fire when there were so many protections built into the system.
- Not yet determining the cause of the event; just sharing characteristics right now.
- Would not answer a question if the plane would be safe to fly based on information known today. Deferred to FAA’s authority on this.
- It’s very hard to tell how long investigation will take. Working hard to understand failure mode and the solution.
- Grounding is unprecedented and we are very concerned. We don’t expect to see fire on an aircraft.
- We are seeing symptoms right now but have to determine whether these are cause or effect. We need to understand why and how and what came first. Haven’t identified this.
- We are working with JTSB to determine if there is a common cause between JAL and ANA. NTSB is a bit ahead because JAL happened a week earlier, but as yet not enough data to know if there is a common thread.
- Does not answer question if Boeing thought grounding was an over-reaction, again deferring to FAA.
- This JAL airplane had been in the air less than 100 hours and would not expect to see a fire.
- We do not have any data to suggest the battery over-charged beyond design limit.
- Still unsure if there are internal defects to the battery. This is why we are tearing down the batteries.
- It is still open question if this is only the battery or if other components are involved.
- We don’t know if it is possible for one cell to overcharge but the battery overall is not overcharged.
- The ANA event has not been called a fire event, it’s a smoke event.
- We are looking at certification standards, whether they were adhered to and whether they were appropriate. What we have seen in these two events do not comport with any design to protect against the battery events. Those systems did not work as intended. We need to understand why.
- We are still evaluating all failure scenarios.
- [NTSB expert]: damaged cell, short circuits could result in thermal runaway.
- [NTSB chairman]: Testing exemplar batteries and how long it will take: this is interesting technology and are looking for other expertise to assist us. It can take a week to do one of the tests. This is not something we expect to be solved overnight. If we have a breakthrough, and find something that will point to a cause, we’ll get that out. But prepared to be methodical.
- Our investigation will certainly look at the certification process and if we find there are vulnerability we will make recommendations. In past focused on aircraft type but could be global if safety of flight issue.
- We’re looking to see if batteries were from the same batch. Looking at flight data recorders from ANA and JAL events.
- We don’t know if short came first or thermal runaway came first. Have not identified sequence of events. Trying to establish why the events occurred and the sequence.
Our take: the 787 will be grounded for an indefinite future.
So, the NTSB does not have a conclusion, yet, but they are still working on it.
Not a good sign. Obviously very complex.
A thorough investigation will take its time.
Think about how long the AF447 investigation took and the lack of initial obviousness in the result.
I think we can only hope this will not take quite that long. The fact that all articles involved are “ony” burnt rather than lost under kilometers of ocean should help.
From the rumour mill leading up the press conference, you would think they’d already found a manufacturing fault in the batteries.
But it seems all possibilities are still open.
Your report gives us an idea of the depth and scope of the investigation. I am delighted by the openness demonstrated by the NTSB, and so early in the process. Good reporting job from you, and thumbs up for the NTSB.
There seems to be no data (yet?) on the charging status of single cells.
So firefighters fought the small, closet like JAL blaze for 1.5 hours! It means they could do nothing except to wait out the oxidizer in the battery to burn out and prevent it from spreading outside.
It seems that these battery problems start occurring only after the batteries have been in use for some time. Somekind of degradations due to vibrations, temperature and pressure changes? Sort of like too old dynamite “sweating” or freezing the dynamite, both ways making it very volatile.
But Boeing has been flying and testing the 787 for quite some time.
They did experience an in flight, electronics related fire – but that was a part being tested that failed then (IIRC).
should we expect some operator influence on this? e.g. they left the engines running on the ground too long, or didn’t shut them down in time (because they never had to bother about that before)
The question is if they actually did contribute to the extinguishing of the fire or if they even prevented the fire from expanding? If their efforts did not change the outcome of the fire, then perhaps Boeing’s system at least worked correctly as far as keeping the fire contained is concerned.
If the charging device is built as per the Securaplane patent, it monitors and records single cell charging status to optimize charging – however, does anyone know whether the charging device for those on the 787 is built into the battery? This is how they advertize their batteries on their homepage. In this case, the data could be burnt and not retrievable…
That is difficult to say.
My understanding of the patent is that they determine a very distinctive point
in the charging characteristics indicating a known charging state (~80..90%.??)
Then they feed the remainder to 100% in a “blind trust” charging episode
into the battery.
This should work reasonably well in an environment of distinct charge and discharge cycles.
i.e. I’d instantly use it for my electric RC flyer 😉
With usage intermeshed into the charging I do have reservations
as there is no room for error in a system that goes poof at 101% charge.
Securaplane by their words provides the Chargers and the APU Start Unit.
They also show LI-Ion main ship batteries ( which on first blush look
similar to the 787 battery from the outside ( ok, black instead of blue ;-).
But the Battery seems to be sourced direct from Yuasa.
The charger is said to have 20..30 sensory wires to the battery ( plus the actual battery connections.)
The available images for the battery show 2 stacked PCBs just behind
the enclosure front. 8 cells in 2 rows in the remaining space to the rear.
It is not obvious what part of the electronics sits in the battery enclosure and
what is located at the charger. With the number of wires my guess would be the battery is “dumb”.
The battery box is a simple cut piece of sheet metal ( Alu going by how the lid deformed ?),
formed and riveted, a simple lid on top held by 8 screws.
No visible provision for holding or directing egress of material from the box was made.
I believe it was inexcusable and irresponsible for Boeing AND the battery
manufacturer, NOT to take advantage of the 3 year delay on the 787 program
itself, due to outsourcing, to make sure that the battery was “inflight o.k.”
It may now take that much time, heaven forbid, to correct the matter, with
incalcuable consequences to Boeing’s financial position and its reputation!
If the information is correct they took enough of an interest to change battery chemistry.
.. to get from 6 month useable lifetime to projected 4 years. Still significantly short of natural life expectancy imho. If Yuasa has the same cells flying on commercial sats I would expect 20++ years for use inside the parametric envelope.
Anyway I don’t think Boeing had much idle personel at hand to open yet another box of pain.
When I saw the first pictures, my reaction was that the containment vessel appeared quite flimsy for the task it had been designed to handle. But one of the principle attraction for this type of battery is its light weight; and therefore a more robust construction would probably defeat its purpose.
I don’t see much issue with designing a hermethic enclosure with controlled pressure relief and an electrolyte “barf bag” with a similar amount of material.
When you look around automotive power blocks tend to use cylindrical cells, often have active temp control and pressure controlled enclosures ( metal, composites ).
Why does the 787 battery underperform in expectable servicelife ?
Is the battery used in buffering the 28V busses ?
Maybe they can use one of the 3 written off prototypes to let a battery burn down under simulated ETOPS 330 conditions, with ventilation, all systems live. Boeing management , Bethune can be in the cabin to see if there’s an over-reaction somewhere.
In fact, why not remote-control the plane by a US Army drone operator, and do it in flight. Bethune can still be in the cabin. Bet that would make him feel manly. 🙂
“•The batteries are unique to the Boeing 787.” Yeah, about 720 times as heavy as the ones used on the A380. I wonder how they compare to those they plan to use on the A350.
I can guess that Airbus is paying very close attention to this process. If the batteries for the A350 are anywhere near as large as those on the 787, I can guess that the FAA & EASA may take a second and closer look at their plans.
Things might even go so far as to review the A380 system, even if it is smaller and has shown no problems this far.
Of course they are paying attention. Any special conditions attached to the Lithium Ion batteries are going to apply to them as well. Up to and including a ban. Any special protections or systems changes such as addition of a dedicated fire suppression system, or secure isolation, as Keesje suggested will also apply. Airbus is whistling past the graveyard at the moment with their press releases claiming they aren’t concerned, but be sure they are watching with baited breath.
Sorry it was Uwe suggesting stronger containments and “barf bags”.
ikkeman wrote “should we expect some operator influence on this?” Maybe Boeing should get some ANA pilots in the similator and try to see just how quickly an experienced airline crew can run through various routines. They might find the electrical system is being called upon to start up various systems a lot faster than they allowed for.
The 787 battery is more powerful and all that power is concentrated in a single unit. The A350 batteries are less powerful and that power is divided among four smaller battery units.
The physical size of a Li-ion battery is an important safety aspect. A smaller unit will dissipate heat more quickly, and that helps to prevent a thermal runaway.
Idly turning wheels:
In both cases the plane was essentially in steady state when the batteries popped their tops.
A: Sitting idle with the APU running B: Flying in cruise with both engines running after APU having been started to enable engine start ( APU off after that?).
In both cases the batterie fault must have been preceeded by an APU start using both batteries ( regular case to start APU ). Discharging the batteries will trigger a charging cycle asap(ower available).
What do we know about the time elapsed from use of battery to fault exposure?
Similar interval? Perchance the time required to (over)charge the battery to 100%
after an APU start cycle ?
Your hypothesis is interesting. We will find out when the NTSB gives us a detailed report of the unfolding of events in sequence, second by second.
This is an interesting question. At this stage of the investigation we don’t really know just yet what would have happened if the fire fighters had not intervened. Perhaps the airplane would have survive but not the passengers, possibly because they would have died from smoke inhalation. In that case the airplane would have continued to fly normally until it ran out of fuel.
In the Boston incident the airplane was on the ground and the aircraft ventilation system was not in operation. But in the ANA incident the passengers could smell the fumes from the battery fire shortly after take-off. I can only imagine what would have happened if they had had to inhale that smoke for 330 minutes.
To remove the potentially confusing statement from the first paragraph, please add “If it had been in the air at the time” before the word “perhaps”.
What strikes me about this situation is that so far there has been little if any progress by literally thousands of the most expert Li-ion battery engineers and scientists working around the clock. Given the uncertainty and lack of prgress in determining the cause and a cure for the Li -ion battery, why has Boeing not expedited a temporary replacement modification with Ni-cad batteries?
Yes the Ni-cad is heavier, and larger, and has quite different charge-discharge characteristics, but stiil — it is only a battery and the 787 electrical system can be safely adapted to it.
The weight penalty and (I’m guessing) relatively minor temporary retrofit costs would seem far preferable to the effect of the indefinite grounding.
What do you expect?
This is a job for 3 maybe 6 investigators. Additional personel will only produce interference.
What certification timeline would you expect for a replacement?
Boeing has nearly chewed through the words “complex” and “integrated”.
My impression is that the low impedance of Lithium batteries is an intrinsic prooerty of the design. ( and probably the early death of the batteries.)
NiCd then is no viable replacement.
apropos: in my book the synonym for the combination of “complex” and “integrated”
is “unwieldy hairball”. ( and the menetekel of Microsoft )
I wonder if anyone is looking at the batteries in the other 48 planes that so far appear to be problem free. Maybe there is something to learn there.
There is a distinct possibility that some of them, if not all, will undergo surgery under the NTSB knife. Unless the job is “sub-contracted” to the OEM. In any event, that would help them to uncover any potential manufacturing defects.
From the report: “•The ANA event has not been called a fire event, it’s a smoke event”
Since a Lithium-ion battery generates highly corrosive smoke when it burns, will the entire aircraft involved in each of the 787 incidents be evaluated? Particularly, the fly-by-wire electronics?
Do you mean to say that a low impedance battery would be a must on the 787, but not on say the A380, which uses Ni-Cd for its main batteries?
Maybe a low impedance battery is required to compensate for the high impudence of the whole 787 design. 😉
A “dead” landing will use the battery for braking power. I seem to have read that this will drain a significant amount of capacity..,. and in a very short timeframe at that 60..90 seconds.
Afaik you will have difficulty with NiCd ( of similar capacity ) performance .
Impudence: The reason for the massively reduced servicelife is still open. IMHO that will show impudence as that usually stems from hugging the parametric envelope.
Compare to small engine thrust derates that provide for significant extension of service life.