RR NDB:

The cells are charged serially? I thought early on i read that the cells are charged, voltage and temp monitored, individually. i.e. in parallel.

Have a bunch of closely stacked cells being fast charged and discharged in series seems like it would be a really bad design. If for no other reason than hot or cold soaking overnight, then cranking the APU, then fast charging it again, would cause huge temperature variations from the center of the battery to the outside cells, unless there was active cooling, which there is not.

Just asking. Are the cells charged and monitored serially or in parallel?

I do think Lithium is the best battery for the job. It just needs to be done right.

glenbrook, Hi

I could have been more clear; the post was meant to be somewhat tongue in cheek. It was (is?) Boeing's position that the rules were honored, that the grounding was a timid reaction to pressure from the political side.

They have a point, the fire was contained withn 20 inches of the case, no damage to nearby equipment, venting worked, etc.

But it is too late to unground... Boeing is keeping their screaming profanities behind closed hatches.

cheers

Radken
 

Machinbird

Ahh, yes. I stand corrected. Thought about checking my facts but failed to follow up.

Some say these Li's are a mature product now. They should be, and I guess Boeing was counting on that being the exact case when the 78 elec design was finalized and flt testing later supported that conclusion. What a shame all this has now come up and spoiled the debut of this unique, and what will be one day, great aircraft.

many ANA removals
 

The New York Times website currently shows as part of their January 30, 2013 content a story which leads with some details of (apparently for cause and unscheduled) removals of battery packs in the ANA fleet in the months before the recent unpleasantness.

This link is quite likely to a point behind their paywall, but even so may be useful to some of you who subscribe.

NYtimes battery removal article

It says that ten batteries were replaced. It mentions that in three cases the charger was replaced along with the battery. It quotes a GS Yuasa spokesman as stating that battery exchanges are part of normal aircraft operation, but gives no data from any source to provide context on whether this removal rate should be considered unusually high save for the GS Yuasa comment and a Boeing comment that the batteries were not lasting as long as intended.

This is hardly proof, but to my eye is an additional suggestion that the actual operational environment on the aircraft is not compatible with the actual battery packs as delivered. A (rather distant) analogy might be the way certain socket locations on certain aircraft have been known to exact a severely reduced lifetime from incandescent bulbs because of excess vibration.

WSJ Article
 

This past day (Tues) the Wall Street Journal reported on its front page that an industry standards organization (unnamed on front page) had disagreed with Boeing on the choice of battery then being considered for the 787. WSJ reported Boeing as saying that the organization did not appreciate the high level of Boeing expertise in this area, or words to that effect.

OE

@USMCProbe

Have a look to the pictures.
You will clearly see that the cells are put in serial. The thin wires at the joints come from the BMS which BALANCE charging current and monitor each cell.
So charged in serial, but adjusted individually.

Battery balancing - Wikipedia, the free encyclopedia


http://i337.photobucket.com/albums/n...SYuasa7871.jpg

http://www.flightglobal.com/Assets/G...x?ItemID=49342

To put the BMS in the same case beside the cells, is not a good idea IMHO...

Looking at the massive connector bars and the tiny wiring, by which percentage can individual charging be adjusted? And what is the "industry standard" with respect to individual charge adjustment, what does for example a Prius battery do?

Yep, good question.:ok:

I hope, it's not a design flaw....

Many Battery removals
 

Ten batteries were replaced at ANA after failing, more at JAL. Any info on other carriers?

This is highly unusual, more so as charge end appears to be very conservative at 4V.

4.1V-4.2V charge end is common in consumer and industrial applications.

Every 100 mV above 4V induces four times the stress, it gets worse above 4.3V though.

At 4V, these cells should last a very long time, many years.

The failed batteries must have been examined. Mighty strange that nothing seems to have been done about it.

@Volume
Less than 5%/3A max, was discussed from:
http://www.pprune.org/tech-log/50569...ml#post7656754
through:
http://www.pprune.org/tech-log/50569...ml#post7658425
and the next 10 or so posts

Quote from USMCProbe (my bold):
"Have a bunch of closely stacked cells being fast charged and discharged in series seems like it would be a really bad design."

Am currently pretty much out of my depth in this erudite, fascinating discussion. However, I cannot see how discharging the APU battery cells in parallel could provide enough "oomph", at about 4V, to start the APU.

As I understand it, the likes of RR_NDB are currently trying to work out how, in principle, the charging might be done in parallel?

Looks like more data is coming to light.

http://www.nytimes.com/2013/01/30/bu...anted=all&_r=0

The various regulations on the design of aviation batteries, posted elsewhere, seem clear to me and what I would expect, so the problem seems to be something fundamental in the battery type and behaviour over the full temperature operating range; or the behaviour of the battery, charger, and various loads as a combined system over the full temperature operating range.

Earlier there was a mention of Testability which in simple terms is the designed means to test the full parameters of a system and simulate the operational environment in which the system will live.

I'd be interested to know the range of operating temperatures within the EE bay and APU battery location from 37,000 ft ASL to sitting on the ground, the external ambient temperatures, and anything that could influence those temperatures in the EE bay/APU battery locations.

I would then want to know when loads are placed on the battery and the timings of that in the knowledge of the changing temperatures from ground to flight levels and vice versa. The batteries that haven't failed yet might have some of that data.

Presumably there is a test rig for this system that does some of this.

I'm sure that professional organisations working on this problem are equally interested.

saptzae:

The number replaced struck me as being rather odd as well. Why didn't this
raise concern ?. That level of replacement might suggest a problem in the
overall system design; pack temperature control in service, or perhaps vibration
related, though batteries failed in different locations.

The data sheet I have doesn't say how many charge / discharge cycles can be
expected. Perhaps you have some typical figures ?...

Regards,

Chris

Weakest link
 

@Chris Scott
All that works for the better part of 100 years, in every vehicle with an electrical system, including aircraft and submarines, as well as with big stationary 48V telco batteries and data center UPS installations. LiPo powered vehicles deserve special mention.

All those are serially charged and discharged. Balancing by bypass is used in all large installations, except NiCd perhaps, not sure there.

I think OP was referring really to the closely stacked cells, which probably is one of the weaker links in this system. Fortunately the box held up, although barely, but it did. Practical battery management is weaker still.

IMHO the weakest link so far is the response to those 10 batteries ANA had to replace. Alarm bells should have gone off no less than after having swapped 10 Engines.

I guess electrical systems must be harder to understand than turbines.

Shades of gray without black and white answers.

Chris Scott:

It's debatable if there would be any advantage, since better methods exist to
ensure that groups of cells in series are charged and balanced correctly. The
disadvantage is that to do it, you would need high current multipole relays
to switch all the cells from series to parallel configuration. As far as I know,
it's never done that way, either in industry, aviation or anywhere else. The cost
and weight penalty would be prohibitive. It would also mean that the battery
would be offline when charging, since the voltage would then be 4v, not the 32v
required.

For the 787 battery, all the cells are hardwired in series, charged and discharged
in series. What we know thus far is:

1) Cells are organised in series, with cell to cell connections via flat metal straps
bolted to the cell terminals. This is standard practice.

2) Cells are charged in series and there's no evidence of a separate charge circuit
per cell. For 8 cells, it would require 9 high current cables from the charger
and there are only two on this battery.

3) It's almost certain that each cell voltage is monitored.

4) It's not clear if there is a temperature sensor per cell, or one sensor for the
complete enclosure.

5) It's also not clear if electronic cell balancing is being done, though there are
probably enough wires to and from the cells to the pcb's, to do this.

Perhaps someone else can add to the above list if there's anything missing ?...

Regards,

Chris

@syseng68k

Along the lines of 4000 Cycles at 4V/80%, 1000 Cycles at 4.1V/90%, 250 Cycles at 4.2V/100%, 50 Cycles at 4.3V, 10 Cycles at 4.4V.

Li based cells keep almost all charge delivered. No gassing or heat. Cells age faster and faster above 90% charge (~4.1V). Above 4.2V cell deteriorates quickly. Heat develops by way of short or straight thermal runaway.

These batteries appear to be charged only 80% and not to see cycles like a phone, netbook or tesla roadster and will hardly be discharged below 60% capacity.

(Percentages and voltages may vary with technology and design)

All these batteries should last several years.

10 out of 30 bat's in a year? Ridiculous.

@syseng68k
Clear to me, two PCB's for exactly that purpose in the box. The upper for control and monitoring. The lower for balancing.

Does not seem to work as needed though, if it does, I would start looking for a hidden bypass somewhere powering the batteries.

Still, the monitoring should see and log all abnormal conditions and fire off warnings before batteries fail.

IMHO not a good place to put a BMS.

777 and others do the same. Even got a cooling fan.

Considering the number of connections, it is the best place to put it. Logs go to the charger and failsafe battery disconnection is also managed there, I hope.

When cells fail the PCB's (failed to) serve their purpose and are irrelevant after a few minutes of logging.

There will be a really interesting report one day.

GS Yuasa Li-ion battery cells selected to power International Space Station
 

GS Yuasa Li-ion battery cells selected to power International Space Station -- ROSWELL, Ga., Nov.*29, 2012 /PRNewswire/ --

Edit:
10 year design life. 16 cycles per day (50%* discharge): 58440

Guess they tune cells and charge just to 3.8V - 3.9V.

*90 min cycle with 35min eclipse, varies with load.

saptzae:

Not convinced :-). Normally, in projects where there is a mix of digital
and measurement quality analog electronics, the analog and digital
sections are separated onto separate pcbs, or the analog sections are
enclosed in screening cans to reduce digital noise. That would suggest
that the board nearest the cells is for the analog and the lower board
handles the digital. The fact thet there are several larger pin count
devices on the lower board suggests digital logic, possibly memory or
microprocessor class devices.

The other point is w/regard to the bypass, where worst case of 10w per
cell needs to provided for. Even using switch mode tech, you still
have to dissipate that power somewhere and there's little evidence of
parts to do this on the boards.

Agreed - the data log for the battery would be logically the first place
to look for the events that led to the failure, but where is it and
why wasn't this escalated back up the chain to Boeing ?. Such a
battery replacement rate should have been noticed at some level...

Regards,

Chris

That's alright. :)

To me it does not seem possible to do LiPo without, as the compounding imbalance will destroy the cells. http://focus.ti.com/download/trng/do...0and%20How.pdf

Edit: And there is an example with switch mode to pump power between cells.

saptzae:

Clever. I was assuming resistive cell balancing, but switched capacitors
would get the job done in a much smaller volume. So they are probably
doing cell balancing on one of those boards...

Regards,

Chris

Parallell charging SHOULD be adopted
 

Hi,

I will post ASAP the block diagram of a PARALLEL CHARGING WITH BITE designed during the last days.

All considerations like cost, reliability, MANDATORY K.I.S.S. approach, etc. was taken into account.

The use of SERIES CHARGING (evident from the pictures) to a battery with adjacent cells (thermally speaking) is imho a recipe for problems.

The location of the batteries in close vicinity to electronic modules and THE PLACEMENT OF TWO DENSE PCB´S in the same chamber of the cells SUGGEST the DECISION MAKING ON USE OF Li Ion was based on:

CELLS WIL NOT FAIL, WILL NOT go to THERMAL RUNAWAY AND BATTERIES WILL BETTER THAN PREVIOUS "STANDARDS" (Ni Cd).

This imho an ERROR, was partly made in "high rocks" of Boeing as we can imagine.

Sad and concerning.

PS

Parallel charging can be done with the cells connected in series. Additionally you may be able to "help" a less capable cell to deliver when battery is being used (as we know, rarely). Imagine 8 (identical) circuits (5 Amps avg, 20 Amps max) connected (floating, obviously) directly to ea. cell).
You can balance easily, manage and incorporate a very capable BITE to track cell performance)

saptzae:
Even my own limited experience with li suggests 100's of charge discharge
cycles and that's in consumer product. I know the lead acid gel batteries
that the telcos use have a 10 year design life and the earlier wet cells 20-30
years, even if that is under float conditions...

Regards,

Chris

mix of replacement reasons
 

That story is evolving this morning. A Seattle Times article quoted a Boeing source at more than 100 replacements. However, many of those are for reasons other than observed premature discrepant behavior in service.

Somewhere on those PCBs lives a function that "locks out" usage of the battery if it is excessively discharged, which then must be removed and returned to GS Yuasa. Apparently there are inadvertent operational errors somewhat akin to you or I leaving the dome light on in an older car which can trigger this removal reason. They also lock themselves out if "improperly disconnected", whatever that means. Another cause is going past the allowed age limit (remember that a lot of these hulls must have had their first batteries installed years ago).

While that may well bargain the hundred down to a much lower number of genuine interest to us, it obfuscates the question of how many genuine in-service discrepancies were seen. And the Boeing excuse list would not seem to cover those three reported cases in which the charger was changed along with the battery. Mis-diagnosis? Real problem?

I harbor a suspicion that somewhere in the removals is the signature of a genuine problem related to the major incidents, but that is just a mildly informed guess. I hope Boeing engineering is mining that data source more energetically than Boeing PR is trying to make it look OK.

Computing a component's suitability also depends on dependability.

Likewise cycles/time. If the suggestion from Boeing is that frequency of unsuitable performance issues were not somehow addressed ad nauseum in the certification process, they are smoking other than Lithium Cobalt Oxide.

There is no reasonable explanation for FAA not to know about the replacement rate in service. None. If it was logged, they know. If not, some one or other or many are in very deep doo doo.

SPOT ON!

I fully agree.

hetfield...

But that is not the case, the special language in the regs allow for fire, loss of electrolyte, and venting of toxic (hot) gases,

That means the materials encased were sacrificial, by definition. No consideration of isolation (engineering) was considered, if it was, it was rejected.
These are deliberate decisions, not blunders....

Do T7 have Li-Ion batteries?

I think many people are losing the sight of an important fact.

The CELLS are Japanese
The Charge/discharge regulators are American.
The Assemblers are French..they make "the battery"!

the whole "box of tricks" "SHOULD" be an integrated "plug 'n'play" fitment , fully self-contained ,to hook straight onto the hookup wires and perform

Clearly it doesn't. the fault lies in the Thales -supplied "Electricity reservoir box"

WRT charge/discharge.....APU should be started off bus fed by 4 main gen's before shut-down....APU feeds all demands in conjunction with main Gen's until engines are shut down.
IF Apu is shut down for ANY reason, the aircraft is then dependent on an outside GPU for the next half-hour ,before the APU can be started again. As soon as the main engines are restarted (GPU?) the power is restored, but temperature/time limitation means that the APU will be inop, under these conditions, for takeoff.

Even if the APU is off, and all gen's fail, there's STILL the RAT to supplement the main battery (which only needs 5 minutes reserve)....under those circumstances, the F.D. cerw will have a lot more to worry about than wether the APU has cooled sufficiently for a restart :}

There is little doubt that there is adequate redundancy and capacity in the system,- even with a U/S APU battery,the APU has 2 alternative starting-modes....so the circle completes......
The components in the Thales-assembled Battery do not perform as a safe, integrated ,rechargeable source of stored electricity.

IMHO, individual cell charge/discharge is inadequately controlled/monitored.

The fact that a marginally -safe technology is used , certainly doesn't help! :eek:

@cockney steve

What about the important BMS, built by "Kanto Aircraft Instrument"?

archae86:
Great info, thanks.

That's starting to give me feelings of dejavu, in that it's increasingly
common for equipment of all types to have no locally servicable parts.
The only recourse being to have the kit returned to the vendor, or their
designated agencies. Even where local servicing may be possible, service
information is often denied and it's designed to create an expensive
monopoly for the sole benefit of the vendor. In this case, why must the
battery go back to Yuasa and not the system integrator ?. At least, local
agencies should be able to interrogate the battery data log to determine
the detailed cause of the fault and thus eleminate spurious shutdowns due
to overzealous or faulty built in test.

Thats adds more weight to something discussed in a previous post, where I was
trying to work out why the data log for the burnt out battery appeared
to be unavailable. The suggestion was that the boards in the enclosure were
responsible for the data logging, (with good reason) and logged the whole life
of the battery. Because of board damage, the logged battery data would likely
be irrecoverable and make cause of failure very difficult to determine...

Regards,

Chris

Where components are made does not really matter (to me).

That several distributed suppliers must be integrated does not help. However, this is just "normal".

Concur, at the system level.

Concur

Right, but are ladders more than marginally -safe?, after all, one can fall down :confused:

What matters is management. That known deficiencies (10 bat's out of 30 replaced at ANA ) were not addressed in time is "not uncommon" mismanagement. Would they have been proactive, they could have started what the NTSB does now after the first three failures, and the fleet would not have been grounded.

That's what I am here for, to learn.

No, but they have corrosive alkaline and alu eating electrolyte, can overheat and short and boil, spilling the electrolyte all over.

My point is that the basic architecture is the same, implementation is common practice in aerospace.

And my point is, do not put a BMS into a case where a "thermal runaway" (Li-Ion) may happen...

@RR_NDB

Looking forward, I provide a free review.

@archae86
Sorry, 10 out of 30 at ANA is plenty to wake up to,with plenty of time. :=

saptzae:
Agreed. Anyone who's lifted the lid of a nicad after brisk charging will see
evidence of electrolyte and the fumes to go with it.

In this case, I wonder if the so called fan in that enclosure is there for cooling, or primarily to exhaust the corrosive gases ?.

Electronics and cells of any type don't mix and just because it's common practice,
doesn't mean that it's correct, nor that it isn't just wrong by design :ugh:. Everyone
else does it, so it must be ok, right ?.

Designers must question everything, especially their own assumptions and those of
the people they most respect...

Regards,

Chris

@syseng68k
Cooling, one would not want to exhaust alkaline gases into an alu fuselage.

As to the gases, like with Pb, it's primarily hydrogen/oxygen, which this poster learned in his early teens, when a car battery blew up in his night desk after being _accidentally_ shorted. (head light makes a good reading lamp).

Engineering practice generally has a sound basis. For example, practically every modern LiPo battery has electronics for balancing integrated.

Right, against convenience, one should not make assumptions at all, because assumption is the mother of _all_ failure.

See, think, validate goes both ways. We have still not seen all the facts and can not make assumptions on changes before understanding the cause.

First, according to the securaplane website, battery history is recorded by the chargers - not the BMS co-located with the battery cells.

(We've been dispelling this myth every other day as someone who hasn't done due reading brings it up again and others in the same state accept it as fact.)


Next, once a cell enters thermal runaway, the job of the BMS has officially ended once it orders the final disconnect of the pack - there is nothing more it can do as the pack is headed for failure at that point.

(It's worth noting that this, too, apparently worked as advertised as there was a "drop to near zero" recorded on the batt of the ANA aircraft.)


Putting the BMS inside the battery case makes it easier to calibrate and more reliable. You are dealing with DC levels into the millivolts, and at these low voltages even the best connectors can add undesired and unpredictable resistance.

Since the BMS isn't recording history, nothing is really lost if it gets badly degraded when a battery does a thermal runaway.

@rottenray

Fully concur, thank you :ok: