Heading North
 

Hi,

RetiredBA/BY:

Considering i am South of UK :) and you directed your comment to Heading South i will risk to comment something just before i leave:

A balancer is supposed to to this. You are ABSOLUTELY right. unfortunately the circuit being used IS NOT THE BEST for aviation use.

You gave a powerful argument to use another MUCH SAFER approach. Parallel charging of EACH CELL. The "solution" to avoid the condition that may explain very well both cases, i think you touched in your comment.

Synergy between your Valiant (electric) and the dangerous batteries you use today that i prohibited to enter our home in my kids RC models. :}

Did Boeing ever test the propagation in a battery installed in an actual airplane? As in, remove all safeguards from one cell, short circuit that cell so that it explodes/burns, then see what happens?

A cell only contains a very fixed amount of chemical energy in it, so it's difficult to imagine that "something happened" and emitted way more energy than expected.

cockpitvisit

I'm not sure that is available. Working backward, FAA accepted a single cell would not cause propagation (further runaway). So the assumption is BOEING designed, built, and tested a platform that demonstrated this to the satisfaction of BOEING. And to FAA.

But that is exactly what happened. "Propagation". Clearly, the fire, flares and heat spread to additional cells?

Volume

So, NTSB have narrowed it down. 1) Battery Design? should the batteries be Prismatic? Cylindrical? Should the batteries be in nonflammable liquid coolant? Active refrigeration or passive heat sink? Should the Enclosure be enlarged, mounted on a stand off platform? 2) Battery construction? High enough standards in layering the Cathode, Anode, Separator stack? Is the environment filtered? Are methods demanding enough as to paste application, folding, wiring, etc.? 3) Monitoring and Management.....

If the controls system caused, (allowed) the thermal runaway, then there is a huge problem. If the battery self ignited, there is a huge problem. If the two cannot be isolated one from the other, (via investigation), the problem is (may be) impossible.

But the third, and most important leg of this three legged stool is the FAA.

Believe it or not, if BOEING can still prove propagation is not possible, the System and Battery don't need to work any better than they did. All Boeing has to prove is one cell will not propagate into multi cell runaway.

The likelihood was recognized and assessed. What was unexpected was the misplaced confidence in the effects of thermal runaway being mitigable. (No Propagation to other cells).

It is this failure in design that is the problem.

If propagation to other cells had not occurred, there would have been no grounding, imo. There would have been smoke, perhaps leakage of electrolyte, and some singeing of the close in structures. But that is allowable, under the regs. (imo).

So it is the regs that offend, because they created the potential for this bizarre argument.

Allowing that single cell Thermal Runaway was acceptable at all turns out to be preposterous.

Which brings up the final turkey. #6 is not a cell, it is a BATTERY. One of eight.
The APU/MAIN BATTERY is (each of them) a group of eight batteries in a stainless steel case.

But FAA says it is allowable to call a BATTERY a CELL.

A ROSE..... by any other name? A lot of what will happen with the Regulations may hinge on nomenclature.....imo.

RetiredBA/BY, RR_NDB et al.

Albeit it might have come across as such I am not dismissing input from modeller airplane electrics - but there is one significant difference: I haven't seen many _paralleled_ LiPo's in RC aircraft. As I pilot RC aircraft as well I have the standard 1S, 2S, 3S and so on cells, but there are always n cells of 3.7V (nominal) _in series_. This is what the balancing circuit on the charger keeps an eye on - that the charge voltage stays equal at all cells.

The issue I seem to thnk about is more the paralleling of cells; as I mentioned before I consider unmatched cells (matching in terms of internal resistance, in 2nd approximation as well internal capacitance and inductance) not very suitable for high-demanding power applications. Whilst at relatively low current settings there are no issues, there certainly are some at high currents or changing loads (AC behaviour of a DC power cell). For as long as cells are just paralleled I agree with previous statements that the paralleling will provide rather stable DC characteristics. But then at AC conditions this looks certainly different.

Not to speak of the soldering quality of the connecting wires which in the past has not always convinced me of quality either. A minor non-perfect solder joint with some dirt in the solder tin can lead to funny effects, especially after some thermal cycling of the cell.

I fully agree with what has been said before, about the charging circuit not being ideal, about the AC conditions of a DC power source, about mechanical design issues when folding/winding up the cells, etc.
In the end, all these factors (and many others) are contributory to the risk and I wish these issues get solved.

It would not be the first time compliance is "demonstrated" based only on calculations, not actual tests or direct measurements.

At least with most RC model packs you can see if the cells have swelled up and not try to recharge them. Would be very interesting to know if any of the 787 cells replaced in service show any swelling.

Design limitation during cell charging?
 

Hi,

RetiredBA/BY and HeadingSouth:

Quote:

---

Still putting my money on a single cell failure and the charger trying to bring the pack upto full voltage, but isn't that what a balancer circuit is supposed to do ?

---

Most batteries are "series charged" i.e. the charger current flows through all cells from a DC source applying current to it´s plus terminal. (opposite as the current the battery delivers when supplying current to a load. E.g. the 787 APU starter system).

Lithium batteries are not charged directly from a bus. The designers prefer to charge them separately. In 787 the MAIN battery are kept charged separated from the bus by a diode: A "high speed switch" that is controlled by the bus voltage. When the bus drops below (estimated) 30 V the battery (if it is ON) supply current to the bus to maintain it´s voltage at required values. A low voltage in the DC bus would make the loads (electronics) fail.

Well, how Li ion (MAIN and APU) is being charged in 787? From the info we have, serially. I.e. the battery chargers are connected to it´s extreme (minus and plus terminals) and the cells are being monitored to avoid the DANGER of exceeding it´s voltage during the charging. Why? Simply if the current is the same in all cells (series connected) and the cells are different between each other (mismatches due several factors) the voltage in each one could be different, charging it (each cell) not equally. This is normally not a big problem in Lead Acid or even Ni Cd. In Lithium ion this is dangerous.

So the 787 designers (certainly) adopted the cell balancing scheme. This explain the harness over the cells. The several thin wires we imagine were used for individual cell voltage monitoring and to "bypass currents" if and when a given cell starts to present a growing voltage.

This approach requires software algorithms and is used for most of chargers probably also because is cost effective. But there are other ways to charge SAFELY the cells:

Imagine you have one (smaller) independent charger for each cell. This (1/8) little charger could easily (without software algorithms) charge each cell to it´s safe limit AND SIMULTANEOUSLY verify if the cell is “healthy”. A “sick” cell can be e.g. one that despite you pass through it the maximum current (no bypass at all) it´s voltage doesn´t shows the “normal envelope” (by different internal characteristics due ANY kind of fault)

Question:

The charger that were used in 787 fleet until it´s grounding is able to verify this cell condition? And proceed accordingly in order to avoid a thermal runaway?

IMO the best approach would be using a parallel charging scheme. Since ANA incident i am designing a solution (not only the charger) in order to eliminate (or reduce) most of the danger in using these wonderful cells.

The charger is part series charger (high current) and parallel charger (low current) and is being currently used (and tested). The software algorithms being used mainly to check if cells are healthy. Continuous cell characterization. (IMO an essential safety function). And to compare, track and record their electrical characteristics throughout useful life.

PS

Two CPT flying RC (carrying potential fire bombs) has a privileged view of the issues. :ok:

A few tupenny-ha'penny strain gauges could show swelling for any cell

An Arduino would be enough to monitor 'em!


:(

"AC" cell behaviour
 

Hi,

HeadingSouth


The big cells (LVP 65) cells were yet fully characterized WRT to transients, superimposed high frequency ripple, etc. ? Boeing selected this battery due his high performance. Highly optimized. Time tells...

Deregulation
 

:}

Boeing was lucky in both cases. In TAK the plane was going to have an inflight fire. Underneath cockpit, inside FWD EE bay. :E

Fortunately (probably) the short circuit to ground discharged and "opened" the battery circuit.

RR NDB

I don't follow. Some of what happened to JAL will bear on that statement. A complete report by the firefighting agency would be a good read.

I thorough reading of Lithium Ion regulations seems to permit some pretty nasty stuff, but I venture fire is not included.

The flight crew had only a BATT WARN, and the "smell" to go on. Is there a FIRE EE BAY alert we do not know about?

They had to descend from 31000 feet, which took some time. When was this predicted fire supposed to have started?

So it appears there may be a squishy area not addressed by regs? Of all the nasty stuff that is allowed, the assumption has to be that the airplane is capable of 180 (330) minutes of continued flight?

All these incidents have occurred aloft, save JAL. JAL problem may have begun while airborne?

What seems to be lucky, is the proximity of runways and/or Fire departments to allow emergency procedures.

787
 

RR_NDB

Thank you for taking the time to give your explanation, great stuff.

So far as I know model LiPos and LiIons are both charged serially, but the balancer takes a reading from each cell to ensure that cell voltage does not exceed the max value of 4.2. Individual cell voltage can be checked on the charger LED display and in a healthy pack the voltages stay VERY close together, suggesting that the balancer is working well.

Interestingly the LiFe charger that I have actually charges through the balance cable so I presume this is using parallel charging in the form you suggest. Battery reliability of my LiFes has been excellent even though it takes quite a load on engine start, certainly better than LiPos.

Of course we don't charge packs in flight in a model but obviously the 787 does, so is there any way the crew could be advised of a developing internal battery problem such as a defective cell, say through the EICAS, and could a defective cell be automatically isolated from the rest of the battery ?

I certainly like the positive (no pun intended) attributes of lithium for model use and with careful charging, and with a professional radio system which samples and records (both on downlink and internally on an SD card) battery voltage and currents every .1 second, they certainly show good performance, albeit with higher failure rate than NiCds or NiMhs.

I think parallel batteries have been mentioned earlier In some receivers I connect two LiIons to the receiver, (To create redundancy and a very secure power supply to a critical system, although I have never had a LiIon or LiPo failure in flight) one at each end of the receiver bus. One battery always seem to discharge considerably more than the other even though connected to the same bus without isolation.

Other receivers have a main and standby (with automatic switchover at a predefined voltage) for the redundancy but are never paralleled so the standby battery always stays fully charged.

Finally, someone queried whether the main battery could/should be used to start the APU on the 787. Isn't that one of its main functions ? I seem to remember frequently staring the APU on the ground on the 75 and 76 and in the event of double flameout and loss of both gens. in flight starting the APU from the battery would be a very high priority ! One might think that a main battery or separate APU batt should be able to do two or three APU starts with ease.

Anyway, I digress, back to the 787 !

P. S. I may have missed it but no one seems to have mentioned the electrical fire and emergency landing on a 787 during test work. Anyone know the cause of that fire?

On this machine, I believe this is not the case. The main battery looks after the mainframe computers (Common Core System), while the APU battery is used to start the APU. My (vague) recollection of the 757 system is that the batteries are paralleled for standby-power functions, but I can't recall what they do when starting the APU on the gound.

On your post
 

Hi,

RetiredBA/BY: Ref:18:13

1) Unfortunately these cells has not "graceful degradation" nor "fault tolerance" a good characteristic of Lead acid and to some extent, Ni Cd. So a message is just to inform the crew battery is INOP. I imagined along this days in a solution using a MAIN battery and a AUX battery. Will comment on advantages later. The AUX could be one to preserve the MAIN.

2) The problem to isolate a defective cell is: You need a costly relay and you reduce the voltage of the battery. It´s not practical.

Weight (cost and other) considerations led designers to have two independent. Certainly a good approach. But posing inherent operational limitations in a degraded scenario.

787 designers relied on this battery to provide adequate safety. Problem is the "ideal" battery failed. The a/c design was adequate. (IMO good)

The official cause was not a battery related issue. I´ve heard of a distribution panel destroyed by a tool left by a mechanic. :confused: The event was serious.

Will comment later on RC issue.

While there was the issue with the tool left behind, the general problem with the panels was the close prox that allowed arcing across parallel panels. That was a major redesign early on.

In regards to the batteries, in the fwd EE bay, it appears this is run like a UPS, where the power is constantly being drawn through the battery system, to provide a clean source for the avionics, else there would not have been an issue in-flight.

IF that is the case, the batteries are in constant use.

Mother nature helped POB of ANA 787 at TAK
 

Hi,

Bear,

Quote:

---

They had to descend from 31000 feet, which took some time. When was this predicted fire supposed to have started?

---

The event is recorded in FDR. Voltage fluctuated and went to zero. At this moment cells were no longer capable to deliver energy to battery terminals (inter cell connection was open between cell # 3 and # 4).

The danger of fire (outside battery case) IMO reduced after this moment. (Or a little later.) Battery was yet partially discharged, short to ground no longer existed and temperature of battery probably had an inflexion point minutes after inter cell opening.

This is a "model". Can be confirmed later or not. :8 IMO it´s probable.

I f you look to the time JAL 787 approaching Logan, taxiing, etc. it´s more probable thermal runaway started few minutes before firstly detected. The short circuit (NTSB put as triggering the thermal runaway) would be "visible" if battery voltage is measured before relay/contactor and recorded. I don´t know if it was.

Indeed. But i insist: ANA case was going to be much worse. The short to ground "saved" from a much more serious situation. Similar to BOS thermal runaway consequences. An inside FWD EE bay.

The basis for my post is:

There are many points in common between both cases:

• Same battery
• Not delivering current
• Being charged or being trickle charged
• Thermal runaway

The differences were:

ANA case:

• A short circuit to ground (from cell # 3 region)
• Inter cell connection “broke” (# 3 to # 4)
• Voltage was recorded (fluctuated and went to zero)
• Fire was contained

JAL case:

• Energy was enough to sustain thermal runaway


The rationale is:

ANA battery energy was lost in the external short circuit. The capability to expel fire was reduced and fire was contained inside the battery case. IMO both batteries were going to caught fire.

Batteries duty cycle
 

Hi,

FlightPathOBN:

In regards to the batteries, in the fwd EE bay, it appears this is run like a UPS, where the power is constantly being drawn through the battery system, to provide a clean source for the avionics, else there would not have been an issue in-flight. IF that is the case, the batteries are in constant use.

In your model what is the function of the diode module?

IMO the MAIN and APU batteries are kept OFF LINE.

Exceptions (batt.):

MAIN: DC bus below ~30 V (including obviously, negative going spikes)
APU: APU start or towing lights consumption (with no APU)

(Most of use after you enter the plane and starts to use electricity, if no GPU and before APU)

Hi RR

After the voltage dropped to zero. To conclude that loss of load on the battery prevented further fire, you have to assune the thermal runaway that had started and spread to other cells, would also cease because there was an open? At this point, how is voltage relevant?

I view the problem the other way, that the monitoring/controls system worked fine, the thermal runaway was due Battery issue, not control issue.

But it does not matter the cause, the runaway was an anticipated event. And a runaway, by definition, won't respond to further controls, hence "runaway"...

The nuts and bolts is the word, "contain". Again, had there been no propagation of thermal event, to other cells, the regs remain satisfied. We had to listen to Boeing repeat ad nauseum that the environmental protections functioned as designed. It seems they did, subject to a heated discussion between paid professionals....

You have to sense how frustrating it is for Boeing. The only thing that went wrong was that runaway propagated. JAL APUBatt burned for 75 minutes.

An hour and fifteen minutes! And yet the aft E/Ebay looked pretty ok...

That goddam pissant propagation! And even then, the house did not burn down.

But they have only themselves to blame. As it should be. They stand to make a half Trillion dollars on this mark.

I think they will.

:ok:

Thats not how I read it.

The BDM (Bat Diode Module) is there to stop the battery being 'charged' from the Hot Bat Bus in the event the Hot Bat Bus is connected to one of the [live] main DC Busses. It's just a non return valve as a back up to a bus breaker/contactor.

The Battery Charger is supplied from a different source (Capts Instrument Bus).

More or less correct.
The APU Bat will ONLY be used to start the APU if no other AC source is available.
The Main Bat again will only be supplying if main DC is not supplying.

The exceptions are: Towing Switch.
APU Bat supplies the Nav lights.
Main Bat supplies brakes, No.1 VHF Tx, etc.

On Bat refuelling.
Main Bat supplies fuel quantity indicating system and limited refuel valve operation.

For what it's worth my opinion is that the majority of battery replacements have been due to the towing and perhaps refuelling switches being left 'on' for too long.:hmm:

BDM, etc.
 

Hi,

TURIN:

Thats how I read it! Since first moment learned on use of BDM.

The Battery Charger is supplied from a different source (Capts Instrument Bus).

Yes. Designers put the important charger switch in LH side. APU switch was left to RH.

The BDM assures the only path to charge MAIN batt. is from it´s charger. Why? Because Li ion batteries are far more critical WRT to charging. When i first learned on it´s adoption for 747 i was concerned. The BDM "solved the concerning". You have a good approach.

Return valve is a good analogy.

Sure! APU batt. Is the secondary (redundant) resource.

I.e. when there is no DC (lower than a given value) in the BUS. I.e. when nothing (from AC source) is feeding DC bus. In this condition MAIN battery supplies (automatically) required energy to the bus through the BDM. (if batt. switch is ON)

:ok:


If "DC bus requires it" BDM allows flow from MAIN batt. to the bus.

:ok: Again, if DC BUS requires it from main batt. As i understand if DC BUS is receiving DC from other sources (from AC of GPU, APU or gennies) MAIN batt. remains OFF LINE.

Towing as i imagine is light consumption and "short" duration. It would be more probable more replacement of MAIN batteries. IMO much more probable to be inadvertently discharged below design limits. (designed to preserve and to avoid recharging a yet compromised battery)

In laptop batteries the voltage span from "empty to full" is about JUST ONE VOLT in a 12 V pack (3 cells in series). Therefore a VERY conservative use.

Fact is, these cells are wonderful but requires a lot of care. From design to end of life. And can be dangerous throughout it´s (entire) life. Someone here in pprune commented on fires of Li batteries dumped in trash in UK.