Boeing 787 Li-Ion Meltdown

[senso]

So they use that new chemistry to lower weight and then add some steel plates, I smell it would be lighter with a more stable chemistry..

[jerry507]

The case was ALREADY some form of metal, and I'd bet it was probably steel. They're simply sealing it so they can control the oxygen in the box.

Sinnetts analysis is spot on. They can't figure out what went wrong, so look at everything again and try to fix any potential problem. Why is the forums general consensus they can't figure out what went wrong because they're incompetent? The thing burned, it isn't strange that there isn't enough evidence left.

They've made a lot of good changes that won't really add much weight. This is a good example of quality engineering.

[EEVblog]

http://www.gizmag.com/787-lithium-ion-safety/26676/

Back in the air in weeks?

And a lousy 68kg added weight?
That's down in the noise when you consider the variable weight of the hundreds of passengers and their luggage! 

[TerraHertz]

From http://seattletimes.com/html/businesstechnology/2020561865_787batterybriefingxml.html

In addition, Boeing is adjusting the battery charger to narrow the acceptable level of charge for the battery. It will both lower the highest charge allowed and raise the lower level allowed for discharge.

Ha. It would be nice if they mentioned what the resulting effective accessible charge to mass ratio is after the charge margins narrowing. So funny if the end result is worse than NiCads.

[SeanB]

Probably would be now very close to NiCd ability, but with a very steep cutoff at low voltage, unlike a NiCd which will deliver power until it dies. If you have an oops moment in the air and need battery power I hope there is a "break tab and operate but will mean battery replacement needed" guarded switch that overrides the low voltage cutout for use when the fertiliser hits the ventilator and the battery is the only thing keeping the wings from becoming ground augers. No good going in Valuejet style with a battery that is still good but which says no more power.

[staxquad]

Probably would be now very close to NiCd ability, but with a very steep cutoff at low voltage, unlike a NiCd which will deliver power until it dies. If you have an oops moment in the air and need battery power I hope there is a "break tab and operate but will mean battery replacement needed" guarded switch that overrides the low voltage cutout for use when the fertiliser hits the ventilator and the battery is the only thing keeping the wings from becoming ground augers. No good going in Valuejet style with a battery that is still good but which says no more power.

Current provided for airplane startup:  Li 150A, NiCad 16A
weight: Li 28.6kg, NiCad 48.5kg

It was mentioned on PPrune that they balanced the cells in groups of 3.  If that's the case, there's your problem.  One cell goes low, damages, then they try to recharge, the 2 other cells go too high and damage, and with repeated use the battery goes critical from increased internal resistance.

edit: Above seems to be erroneous information.  The battery consist of 8x LVP65 in series.  The cells are folded in 3 layers inside each LVP65, but that's one cell.  (There's no 3 cells in parallel as mentioned in PPrune.)  So balance, temperature, voltage min/max, charging monitoring should be done to each 8 cells in the battery.  Monitoring should shut down the battery before becoming problematic.

Also, the cells can damage from heat, and now they've insulated the cells even further to heat up even more than before.  They should not have stacked the cells the way they have, with the additional insulation, they will just heat up more.  They worry about one cell damaging the next, but let each cell damage more easily.   The cells should be configured side by side along the walls of a box with a center courtyard of emptiness (a square doughnut) so that each cell is not adjacent to another.  Right now the center ones are chocking.  There's no confidence in their fix.

When I charge my LiPo batteries, the charger monitors voltage and impedance on each cell showing if there is a damaged cell to be replaced immediately.

One of my batteries had a cell too low that the balance charger could not deal with.  So I discharged the other two to be at the same voltage as the low one and recharged.  Have used for 4 or 5 cycles and the battery remains balanced.  (there is a difference in impedance between the former low cell and the former two high cells that still remains, 3S 12V 8000mAh 30C; 9m?, 9m?, 5m? impedance; the same battery but 2S  12V 8000mAh 30C and healthy has 4m?, 4m? impedance).  If I did not have a balance charger letting me know the condition of the cells, but just a wall wart, charging or using that battery might have resulted in a fire.

Boeing should have hired a 15 year old RC hobbyist to design those batteries in the first place. 

They better be balancing each cell, monitoring each cell, not in groups of 3.

[LaurenceW]

In some of the photos of a "still good" battery pack that I've seen, there is clearly supervisory/monitoring wiring going to each cell, but it doesn't look like it would carry different charging currents.

SeanB, I have done a few informal experiments with 18650 lithium cells. The manufacturers tell me not to go below about 2.7V per cell, or damage is done. so I tried it! To be honest, there is precious little energy left in the batteries at this point, anyway.

[staxquad]

In some of the photos of a "still good" battery pack that I've seen, there is clearly supervisory/monitoring wiring going to each cell, but it doesn't look like it would carry different charging currents.

SeanB, I have done a few informal experiments with 18650 lithium cells. The manufacturers tell me not to go below about 2.7V per cell, or damage is done. so I tried it! To be honest, there is precious little energy left in the batteries at this point, anyway.

Did you monitor the cell's impedance before and after you conducted your experiment?

You observe precious little energy outside the battery.  Can you observe the chemical damage inside the battery?  That's why you're supposed to check internal impedance.

If you don't monitor, everything will always seem hunky dory, which is exactly Boeing's problem.   

[tom66]

So a NiCad weighs about twice the LiPo.
A piece of baggage easily weighs the difference.
I suppose, "every bit helps", but compromising (so very obviously) on safety worries me...

[ee851]

Does the steel battery box need to be vented?    Why can't the box simply be sealed and filled with either halon or carbon dioxide, a gas that cannot support fire?    Not to mention pressure sensors and temperature sensors.....Seems to be that keeping the contents of the box under one atmosphere of pressure all the time might speed up the testing process.

[NiHaoMike]

So revert to the well proven batteries and then make sure at least 3 passengers onboard every flight are as skinny as Tiffany Yep. (Though that might not work out so well in Japan where most of the passengers are very skinny to begin with...)

[staxquad]

So a NiCad weighs about twice the LiPo.
A piece of baggage easily weighs the difference.
I suppose, "every bit helps", but compromising (so very obviously) on safety worries me...

The Yuasa LiIon used by the 787 can recharge at a 1C rate (recharge at a rate of 65 amps for one hour), discharge at a 5C rate and has a 65Ah capacity.

Boeing designed the 787 around the current draw, weight, size and recharge time of the LVP-65 LiIon batteries.

If a match is required for a current draw of 150A (for airplane powerup), they would need 9x the batteries used by the 777, with a resulting weight of 436kg compared to 28kg.

Other batteries do not fit into the 787's current design.

[Hypernova]

IMO they should do the battery SiFi style - Have a row of ports on the side of the plane and eject any cell that catch fire!

[chickenHeadKnob]

Does the steel battery box need to be vented?    Why can't the box simply be sealed and filled with either halon or carbon dioxide, a gas that cannot support fire?    Not to mention pressure sensors and temperature sensors.....Seems to be that keeping the contents of the box under one atmosphere of pressure all the time might speed up the testing process.

As far as I know Li- cobalt and similar chemistries will continue thermal run-away even in an anoxic environment as the full complement for redox reaction is right in the cell. And a vent is "extremely" necessary, consider what will happen when all those evolved gases build up and finally breach your containment there by generating shrapnel, a good pipe bomb design.

[G7PSK]

CO2 is not an inert gas especially if you have an electric arc as it breaks down. you would need Argon or Helium or possibly Nitrogen.