787 Batteries and Chargers - Part 1

[RR_NDB]

Torque at intercell connections

Hi,

Rocchi:

Good question!

Low or high torque are not good. Low IMO is worse. Could overheat the connection.

The model i imagine is another: Heat, short to case and LOC.

Your point could be one factor to damage (cell # 3).

Welcome!

[Lyman]

I think at least in ANA, the stainless case contained the event. The "top" is deformed, having been subjected to sufficient pressure to significantly bulge the structure with the fasteners preventing separation....

[FullWings]

If you compare the BTU heat storage of these batteries, with the BTU storage of the fuel in the wings, and the airborne fire coursing through the bleed air systems of all other aircraft, this is really minor.

I'm not sure this is a totally valid comparison. The fuel is inert on its own and the bleed system can be shut off as soon as abnormal temperatures are sensed. If the wings carried a mixture of LOX and kerosene, then yes...

And the system worked. All of it.

Errrrm, not really? Uncontrolled fire in one aircraft on the ground, meltdown and discharge of battery internals in another, leading to an emergency landing and evacuation? What would a non-working system look like?

There is more energy in half a pound of butter than in an average grenade but which would you rather hold in your hand? Now pull the pin.

It just appears to me that the combination of cobalt chemistry and a lot of high-capacity cells jammed together isn't the right way to go for anything that flies. Or drives. Or floats...

[saptzae]

@grebllaw123d

Rating is 5 * C (A). C is 65Ah. 5 * 65 = 325A

The inverter powered starter motor soft-starts and draws well below the the battery limit.

@hetfield

Yes, lead acid can do it, Li-Ion not.

No, lead acid can't either and also consists of sub-cells.

[Lyman]

Failure mechanism, example.....

Quote......
"The use of copper as the current collector for the negative electrode has particular reliability and safety implications. At very low cell voltages (usually approximately 1 V for the cell), the potential at the copper current collector increases to the point where copper will begin to oxidize and dissolve as copper ions into the electrolyte. On subsequent recharge, the dissolved copper ions plate as copper metal onto negative electrode surfaces, reducing their permeability and making the cell susceptible to lithium plating and capacity loss. Usually, once a severe over- discharge event has occurred, cell degradation accelerates: once the negative electrode has become damaged by copper plating it will no longer be able to uptake lithium under “normal” charge rates. In such an instance, “normal” charge cycles cause lithium plating, which result in a greater loss of permeability of the surfaces. Ultimately, over-discharge of cells can lead to cell thermal runaway." ....Unquote

http://www.nfpa.org/assets/files/pdf...rieshazard.pdf


So it would seem that "internal protections" would be limited practically to a strict control of voltage within a frame of charge/discharge. protection from mechanical damage, and environental stressors are straightforward, nothing below -18 or above 60 (C).

Since the listed resrictions are "mitigating" rather than "prohibitive" (eliminating) as to failure, they can certainly be construed as "permissive of" smoke, runaway and even fire "events". A "mitigation" is by definition "may" not "may not"....

[bill good]

B787 battery

On the question of containing a problem within a cell. How many prismatic shaped pressurised aircraft are flying? The question on the backup internal protection is where the fault will be found. I cannot find the same sort of protection built into the prismatic cells as can be found within the cylinder type which has at least 2 methods to prevent this sort of trouble. What I would like to find out is if there was power being feed back to the batteries in both cases while this fault condition existed or if the process of disassembly was all within the battery case. i.e the one on the ground- was ground power or APU supplying DC to the batteries & I assume the A/C flying had the main battery powered from its charging source. By design, if the battery is fully charged it should be disconnected to protect the aircraft (imo). Li-Ion has no loss of capacity over any flight time of any aircraft. All the Aux. battery does is to start the APU when Ext, AC power is not available after which the AC alternators provide the systems power for normal ops. (maybe needed to "clean up" the DC but that could be addressed) There are several issues not just one cell failure that must be addressed before this aircraft is returned to service on just how safe the interface is to other components within the DC system. I hope all are identified.

Hopefully the fix is close.

[TURIN]

Bill, the battery does not start the engines.
The APU battery can be used to start the APU if no other AC available.
Main engine start is from high energy AC sources only.

[RR_NDB]

Short as consequence of heat inside box

Hi,

FullWings @ # 523

"It just appears to me that the
combination of cobalt chemistry and a lot of high-capacity cells jammed together isn't the right way to go for anything that flies. Or drives. Or floats..."

And for anything in ground too. So the question is: Why the cell(s) started to overheat?

The "equivalent circuit" derived from a visual analysis of the ANA battery shows clearly the temperature inside the case was so high that caused a dramatic short circuit that "fused" an intercell strap and a thick ground wire. I tend to think as the increasingly hot cells inside the case like you put created the short circuit of cell # 3 to ground. (~ 12 V applied in resistance of miliOhms. Surge current of hundred Amps. (probably thousands Amps).

I don´t put as probable a short circuit as the trigger of the battery destruction. FDR iirc data is compatible with the model i presented.

[saptzae]

@RR_NDB

The ground strap shows little thermal damage and its cross section is too small to have carried any significant current in destroying #3.

The ground strap was likely mechanical damaged and partially cut either before the event or by intervention.

Severe damage seen is tertiary. I would not spend time on it.

Important is Primary failure, cause of deterioration leading to shorting cell and Secondary failure, if/why the battery was not protected from charging.

Edit: If ground strap was cut pre-event and could have touched BMU PCB, it possibly could have induced the battery failure.

Edit: Arcing at the pole of #3 due to loose nut, I think it would have welded shut, not burned the pole off.

[bill good]

B787 battery

Ths Turin, I was focused on the local problem at the battery installation level!!
I hope it reads correct now.

Anyone able to give min voltage alowable before dispatch for this A/C?

[RR_NDB]

APU battery usage in BOS

Hi,

TURIN,

JAL APU at BOS was started by AC or from battery?

A model i am imagining is:

1) Batteries discharged (for many possible reasons) more than expected
2) Thermal aspects (cell heating, improper thermal cell monitoring, inadequate battery case, inadequate cell spacing) created conditions for thermal runaway (a positive feedback that in the end destroy the device)

So, my question is because ANA 787 main battery very probably was being recharged (hi curr.)when event started.

BOS JAL APU battery probably was also being recharged? (high curr.)

My feeling, was NOT.

[FlightPathOBN]

I'd like to see some of the other 99 batteries they replaced, and some images of what the battery box looked like when the fault messages showed up...

[RR_NDB]

Hi,

Saptzae @ # 529

First i need to clarify some points to prepare my comment.

1) Destruction of # 3 was by what? Internally generated heat (due subcell inbalance, dendrites, voltage), ohmic losses due thousand Amps flowing to case (and ultimately to ground), or what?

2) On ground THICK wire you has two details: One seeming mechanical and another VERY PROBABLY caused by excessive current. Please refer to the equivalent circuit where i put the (highly probable due evidences) short circuit path.

3) Severe damage shows a PATTERN. It´s up to us to establish a model rhat if is robust enough could explain most. I am just taken into account everything. Could we discard sabotage? Answer is NO. I prefer look to all details and balance them in the best proportions.

4) Shorting cell? Which one? Due over voltage, improper charging? Temperature of individual cell(s) not being measured?

5) to be continued...

[FlightPathOBN]

RR,

I would suspect the way the cells are wired as well...not with bus bars, or ground straps, but simple braided wire...

and I certainly would not consider this 'thick' wiring by any definition...

[saptzae]

@RR_NDB

First i need to clarify some points to prepare my comment.

1) Destruction of # 3 was by what? Internally generated heat (due subcell inbalance, dendrites, voltage), ohmic losses due thousand Amps flowing to case (and ultimately to ground), or what?

I continue to rule out thermal runaway as cause of Primary failure. My scenario is same as at BOS, the same pattern. Primary failure was in #3 (#5 at BOS) due to deterioration of unknown origin. Secondary failure overcharge and thermal runaway of several other cells. Tertiary failure is total destruction of #3, how the pole can burn away, leaving the nut unscathed, beats me. The cause of "deterioration" leading to Primary failure remains as elusive as ever. By now, NTSB could have said something, perhaps them looking for micro-shorts by testing last week gives a hint toward the short of a cell as the cause of Primary failure.

2) On ground THICK wire you has two details: One seeming mechanical and another VERY PROBABLY caused by excessive current. Please refer to the equivalent circuit where i put the (highly probable due evidences) short circuit path.

Yes, it could be that the frayed, not severed wire burned up the way you suggest. It could also be that the severed wire caused shorts on a BMU PCB. Now, that would be really awkward and destroy the pattern of both batteries and feed sabotage conspiracies.

3) Severe damage shows a PATTERN. It´s up to us to establish a model rhat if is robust enough could explain most. I am just taken into account everything. Could we discard sabotage? Answer is NO. I prefer look to all details and balance them in the best proportions.

I think it is tertiary damage pattern, like structural breakup, it can go any way.

4) Shorting cell? Which one? Due over voltage, improper charging? Temperature of individual cell(s) not being measured?

#3, (#5 at BOS), deterioration of cells by mishandling after assembly** / by maintenance or mismanagement by BMS during flight operations.

** I don't know how battery could be safely assembled when cells carry a charge. Any mistake like a short would lead to a firework or damage. Perhaps, battery is assembled while cells not carry a charge, and are charged after assembly.

I am not confident about cause of deterioration leading to Primary failure.

Mishandling like for example resetting BMU cutout of deep discharged battery, thereby charging damaged cells would be a convenient and fixable explanation. Mismanagement by BMU would be harder to fix.

If I would want to figure evtl mismanagement out, I would connect transient processing by way of a 8 channel differential 12bit transient recorder at 1megasamples/sec to the battery, one channel per cell and perform real-time transient analysis to find unique voltage transients on cells.

I feel confident toward the Secondary failure pattern, it is a very tough job for BMS to detect cell short in a timely manner. That will have to be fixed. It may well take something like the above mentioned transient processing to be responsive enough. Another way could be an infrared camera in the case. BTW, Temperature sensors would be not much use as the response is way too slow.

Tertiary failures - breakup - I don't really think about.

It all may well boil down to lack of understanding in a highly decentralized operation. Examples are that nothing seems to have been done about deep-discharged and failed batteries over many months.

Edit: BMU -> PCBs in battery box. BMS -> BMU + charger + all other related functionality.

@FlightPathOBN

I would suspect the way the cells are wired as well...not with bus bars, or ground straps, but simple braided wire...

and I certainly would not consider this 'thick' wiring by any definition...

The cells are connected using at least 350A rated bus bar and the wires are only for monitoring and balancing at currents below 5A.

Edit: I also would like to see CT scans of in-service as well as of failed batteries.

[FullWings]

RR,

And for anything in ground too. So the question is: Why the cell(s) started to overheat?

Agreed. Reading through this thread there are many cogent theories as to why the overheat began and I have to say most of them seem at least possible.

The main question for me is: why did batteries whose technology was known to be vulnerable to catastrophic thermal runaway (more than virtually any other kind of cell) get put into a configuration that sent the whole lot up in flames if there was a *single* failure? Who in their right minds would install that in an aircraft that could be 3hrs from the nearest airfield?

If it had been designed to contain an individual cell going AWOL, then you'd get something like a status message "ELEC BAT MAIN SYS" which would show reduced redundancy in flight and maybe no dispatch at the other end until the problem had been sorted. You don't expect (or want) a large percentage of the chemical energy in the whole battery being released over a short period of time inside the aircraft...

Looking at it from a statistical POV, if all it takes is a microscopic rupture in a nanometre-scale film in one cell to trigger a destructive event, when there are 800 787s flying 18hrs a day carrying (insert number) of cells, how likely does this become?

[cockney steve]

Regarding the number of batterie that have been changed-out.-

First, one has to go along with the officoal viewpoint at Certification....."safe, well-proven, long service-life......."

So, What we're looking at, is an overgrown cellphone or Laptop battery.

they cost a fraction of (16,0000 USD?) the price of the Thales "box of energy" They're charged by non-technical and sometimes low-intelligence owners ,using cheap and nasty power-supplies.

IT ALL WORKS MILLIONS OF TIMES BETTER THAN THE "SCREAMLINER" SETUP

The number of documented failures have been a minute percentage of the Thales failures.

All these "Domestic" batteries are properly fed/controlled by their interface circuitry.

When your phone needs charging, it switches itself off....you connect a supply and it automatically "sees" it and commences recharging....does this for hundreds of cycles.

So, WHY is the "screamliner's" system so pi55-poor that it won't stop an over-discharge but instead converts it to a disposable?

The lack of sub-cell monitoring and the whole lack of fitness for purpose makes one think of short term profiteering.

For whatever reason, Boeing plumped for the most sensitive and unstable of the Lithium technologies.

You would have thought that the person(s) making that decision, would have used every resource available to ensure that it worked properly.

they didn't ......it doesn't. someone has a lot to answer for. that many "dead" batteries in that short an operating-life points to a major problem......so, who was brushing it under the carpet?


If the volume of components ordered is twice the number that the customer can use (new-builds) the balance is going somewhere.....in this instance, a massive pile of replacements well before their anticipated service-life.

I feel there's a lot we don't know, as yet.

[TURIN]

When your phone needs charging, it switches itself off....you connect a supply and it automatically "sees" it and commences recharging....does this for hundreds of cycles.

So, WHY is the "screamliner's" system so pi55-poor that it won't stop an over-discharge but instead converts it to a disposable?

I think this was covered in either this or one of the other (five?) threads.

In an emergency you don't want your battery to shut down just to save itself from the scrap heap.

Imagine this scenario.

Full inflight electrical supply failure (doesn't matter why-lets call it fuel starvation so no APU either. The Airtransat A330 gliding into the Azores springs to mind.)


The main bat kicks in while the RAT drops and spins up.

The a/c lands safely due to excellent pilot skills but on the roll out as the RAT winds down due to lack of airspeed, the brakes are being powered by the main battery. The battery monitoring system at this point senses a low charge state and shuts the battery down. Pilots get a black screen, microsoft windows icon, the words "hibernating" flash on the PFD and a little yellow box pops up and says "please connect external power or change to a new battery".


Meanwhile, the poor chaps at the front end are smashing their feet through the brake pedals as the end of the runway looms at about 140kts.

OK I'm being facetious, and you could argue that a simple logic based on air/ground, airspeed, wheel speed etc should be incorporated to stop that.

Maybe it was suggested and dismissed as unnecessary expense/weight/design complication. Who knows?

[TURIN]

Hi,

TURIN,

JAL APU at BOS was started by AC or from battery?

No info on that sorry.

In normal ops the APU would have been started on taxi-in off the normal AC supply. However, in my experience it is not uncommon for the ground electrical supply to drop off line (after engine and APU have been shutdown) requiring the APU to be started off the battery.

Happens a lot where I work as the airport infrastructure is diabolical in regard to suitable multiple 90KvA supplies.

[Lyman]

Separator
Lithium-ion cell separators most commonly are porous polyethylene, polypropylene, or composite polyethylene / polypropylene films.29 These films are typically on the order of 20 μm thick, although thinner (approximately 10 um) and thicker films can be found (approximately
40 um). The function of the separator is to prevent direct contact between the anode and cathode. The pores in the separator allow transfer of lithium ions by diffusion during charge and discharge. These films soften and close their pores at elevated temperatures (usually in the range of 130 to 150°C / 270 to 300°F), and stop charge or discharge processes by impeding the transport of ions between the anode and cathode. Thus, these types of separators are commonly referred to as “shutdown” separators. If a minor internal short occurs within a cell (e.g., from small contaminants penetrating the separator), local separator shutdown will effectively disable a small point within the cell by melting slightly and closing the separator pores (Figure 13). The shutdown function will also permanently disable the entire cell in the case of an abnormal internal temperature rise to approximately 130°C (266°F) (e.g., due to high current draws caused by an external short circuit of the cell) (Figure 14). However, should internal temperatures rise

This is where the problem is... If Boeing admits it is this simple, they will not lose some face, their face will fall completely off....