787 Batteries and Chargers - Part 1

[Ian W]

RR_NDB

What you mean by "incorrect disconnection"?

That was the quote in the WSJ story. I presume that there must be a procedure for disconnecting the battery that was not followed. It may be an area to be investigated.

[inetdog]

One of the descriptions stated that, among other things, opening a fuel fill door turns on the whole fuel monitoring, transfer pump, etc. system (on battery if no other power is connected) and leaves it on until the door is closed.
They felt that was one of the major causes of battery flattening.

[mm43]

Question?

What comes first? The Batt~DC Buss blocking diode then the Batt CB, or is the sequence reversed?

Am I right to assume that both the above are outside the battery pack?

[Old Engineer]

Help from Kirchoff

It's amazing how proper balancing is done with these tiny wires.

Comments about this have appeared several times. Perhaps I can describe a way in which this could be done.

We want to pass a slightly different charging current through each cell. Instead of 8 chargers each running at 40 amps, say, let's have one charger running at 40 amps, and perhaps one master control on this so that it can taper off at 20 or 10 amps. This will make the math of the example easy to follow.

Let's overlay a current loop thru each battery (two small wires to the two terminals of a cell, common and +/- DC). Let's use an overlay current that may be as great as +2A DC, or 0, or even -2A DC. We'll take +2A DC to mean the 2 amperes is added so as to flow in the same direction as the 40A chargilng current from the main source. Now that means that an overlay of -2A would flow in the opposite direction to the main charging current of 40 amps.

Now you say current cannot flow in two directions in the same wire. That is certainly true. However a guy name Kirchoff proved that you could do your calculations as if it did, in loops of your choosing. So even tho such calculations are in some ways not literal, they do give correct and quick answers. In fact, you can build circuits based on the imaginary math, and they will actually give real currents.

Let's say the target voltage on full charge is exactly 4 VDC per cell, and 40 amps for say an hour will bring a 65 Ah battery of 8 cells back to 32 terminal volts on the battery. This means it was discharged down by some number of ampere hours less than 40, but not exactly known.

Let's say we do know that on average, when the cells are new and uniform, and the same discharge has been made) a 41 A charge will bring the cells back to 4.1 V per cell; or of 38 A, back to 3.8 V, and so on. To know this we have attached a 3rd small wire to the output of each cell to measure cell voltage (between cell common - above - and the 3rd wire).

Let's say we have a cell that is only showing 2.9 V when six have all reached 3.0; and another that is in the hot center of the stack, and being warm, it charges more and shows 3.1 V. Because our electrical trim tab is so small, we must begin correction of the current on these two cells once a difference appears. So we apply +1 A overlay to the first problem cell through the small wires, and -1A to the second problem cell. So the 1st problem cell gets 41 A, and the 2d 39 A, net effective charging current.

By the way, the necessary wires to carry these currents can be related to the usual household wiring, where a #14 AWG in Cu carries 15 amps, a #12 20 amps, a #10 30 amps, a #8 40 amps. The plug in power harness from the GPU may need #8 cables for the main charging current and return.

In smaller wires, it is 10 amps for a #16 wire, 7 for a #18, and from that (ie, less 8 numbers) a #26 for two amps. The relations (in US, logarithmic among wires) are 3 numbers to double the copper area, 6 numbers to double the diameter; because of the surface area effect, 4 numbers to double the ampacity. A #10 is one-tenth inch in diameter. For Al wire, add 2 in size; altho I'd think Cu was more likely at a battery. Insulation, I'd think between 15 and 30 mils here.

Now, from all that, with the ruler scale on the battery pix, a pretty good idea of the battery wiring can be obtained, with a little help from CBrd considerations, as follows.

On a circuit board, a half amp can be handled with a top-hat power transitor; 1 or 2 amps with one mounted on a metal slug; but above that (say 5 amps) something with fins (not seen) is going to be needed; there may be some slugged devices. So it's possible eight 2-amp lines of the "trim" current supply are provided from the ground charger. My guess is yes and that the charger controls the current trim in accordance with the cell voltage telemetry transmitted digitally. All this may be in the "data" cable.

IIRC, an incident of operation with only the power cable attached was reported. The questions then are:

1. Is the ground charger multipurpose, whereby the data cable is not used for other battery types? What disconnect sequence? What is the effect of not connecting the data cable?

2. Is capabilty to vary the charging current by cell by 2 amps either way enough? Is it that or more?

3. Is this cell balancing done only on charging, or also discharging? It is obviously impractical in the scheme outlined during APU cranking.

4. Plus all the usual-- algorithims, floating discharge on bus, etc.

OE

[ross_M]

In fact, you can build circuits based on the imaginary math, and they will actually give real currents.

Loved that!

[TURIN]

A FEW POINTS FOR CLARIFICATION.

What comes first? The Batt~DC Buss blocking diode then the Batt CB, or is the sequence reversed?

From memory, no schematic available at the moment.

APU BAT
FO's Instrument Bus-CB-Bat Charger-Batttery-APU Hot Bat Bus-SPUC-SPU-ATRU-CMSC-APU VSMG.


MAIN BAT
CAPT's Instrument Bus-CB-Bat Charger-Battery-Diode Module-Hot Bat Bus-Bat Bus Contactor-DC Bus (Can't remember which one).

As you can see the diode module is there to ensure no backfeed from the Hot Bat Bus to the battery in the event the Bat Bus Contactor stays closed when normal power is supplying the DC Busses.

One of the descriptions stated that, among other things, opening a fuel fill door turns on the whole fuel monitoring, transfer pump, etc. system (on battery if no other power is connected) and leaves it on until the door is closed.
They felt that was one of the major causes of battery flattening.

Yup. The "battery only" refuelling option is there to get just enough fuel into the left main tank to allow the APU to be started. (Approx 15 mins).

The backup battery should not get much use. Starting up a "cold" airplane, the battery might be used just to get essential power up to do a APU start. Also cleaners might need a little light, so a mechanic might turn the batteries on. Opening and closing cargo doors. etc. Not high draw stuff, and not for very long.

The first bit is about right but lighting will come from the Virtual ground Handling Bus which will only be powered from External Power or Normal Busses.
Cargo doors are hydraulically operated from stand alone electrically powered hydraulic pump packs. Again power is from the Virtual Ground Handling Bus which will only be powered from External Power or Normal Busses.

I only saw a "dead" backup battery once. Someone left the battery switch on overnight, or maybe someone outside disconnected external power with the BATT switch on.

Me too.
However at least the old style battery could be recharged in situ. Not this beast.

One of the peculiar 'gotchas' at the moment is the workaround to ensure APU rotor bowing does not occur. It essentially means selecting the APU start switch to 'on' after the APU has shut down to re-open the APU inlet door for ventilation and then switching the Main and APU batteries off. It is possible I suppose that some of these flat battery problems could have been caused by this. Catch all phrase of mishandling?

[keesje]

If this batteries has an excellent measuring and control system, why didn't it cut off / send low warnings when it approached it critical low level? Or did it.

Anyway it seems Airbus better Indentified the risk of the use of these batteries at the start of the program, warned everybody, seperated the cells for cooling, took smaller ones and... developped a plan B in parallel just in case off..

Airbus says it has a Plan B for A350 jet batteries | Reuters

[cockney steve]

interesting comments in the last 2 posts. A former life saw me as a service-engineer for Medical equipment. (mainly Prem. baby incubators)

We recognised the capacity for abuse and misuse and as far as possible , designs were developed to circumvent "unplanned" "usage"

The saying was, " design to be idiot-proof and it'll be almost nurse-proof

Now Turin talks of an unorthodox procedure to overcome a major inherent design fault (distortion of Rotor due inadequate provision to dissipate soaked-heat.

(incidentally.SYSENG 68K , the BMW Mini-Cooper also uses a "run-on coolant pump and it was these devices causing unexplained spontaneous combustion of the cars,,,caused a major recall...Deja-vu!)

@ KEESJE. AIUI, the connection/balancing/discharging/monitoring ai all internal in the "battery" , it being a self-contained sub-system.
One therefore makes the assumption that all charge/discharge functions must go via the "charger"which is then a glorified power-supply/filter as seen by the battery. One also assumes the safest way to transfer the data from inside the Battery, to the data-storage within the charger, is simply to superimpose the digital(AC) info on the main cables and filter it off at the charger (destination)

That way, It's impossible to charge/discharge the battery cells without the controller talking to it's interface. any other "secondary-plug" system is, IMO, wide-open to human error as well as adding another load of wires/connectors into the "potential fault" chain.

No doubt these cells can be used safely in this particular application, but they need to be conservatively used with a much larger safety-margin than they already demonstrate.

Charge/discharge monitoring is, IMO, critical....in the "fuel-door" instance, the batt. should self-disconnect long before it reaches a critical level, just like your cell-phone battery does,- OK, you then have to get a ground-power connection to continue refuelling, but you can then safely restart the APU or a main engine to restore power to the battery-charger (assuming the GPU doesn,t perform this function.)

It needs to be understood that only the "middle" part of the stored energy is actually available for safe use.

Over-discharging effectively destroys the cell. charging to the limit brings a possibility of spontaneous combustion. therefore, like the pilots keep clear of "Coffin Corner" and" Stall" these batteries demand usage in the "safe" part of their envelope.

A design that allows cells to routinely fall below "lockout" level, is ,IMO incompetently and poorly engineered. (Unless it's a deliberate and cynical ploy to sell more batteries ) but no firm would put profit before dependability,would they?

[saptzae]

@mm43

Am I right to assume that both the above are outside the battery pack?

Diode module is external, for main battery only as TURIN explained above. Shorting this battery's connection without protection would lead to a firework. Looking at the picture, near the lower end of the right bus bar is a device, which may be a breaker.

@Old Engineer
Here it is how balancing is done in practice.
http://focus.ti.com/download/trng/do...0and%20How.pdf

IIRC, an incident of operation with only the power cable attached was reported.

if you are referring to the securaplane incident, the technician had not connected the management connector, which, according to one scenario, may have caused the conflagration. The report is linked in one of my earlier posts.

1. Is the ground charger multipurpose, whereby the data cable is not used for other battery types? What disconnect sequence? What is the effect of not connecting the data cable?

All Li based batteries have management electronics built close by, and are charged by a dedicated charger. 787 powers charger from an instrument bus as above.

2. Is capabilty to vary the charging current by cell by 2 amps either way enough? Is it that or more?

2A is plenty, 10s to 100s of mA typical.

3. Is this cell balancing done only on charging, or also discharging? It is obviously impractical in the scheme outlined during APU cranking.

Only on charging. Discharge must be stopped when the weakest cell is empty to prevent reverting it (below).

The critical reason for active balancing is to prevent cell over voltage.

4. Plus all the usual-- algorithims, floating discharge on bus, etc.

Undervoltage, monitoring/logging, emergency cutoff.

Li cells cannot be deep discharged below about 2.7V and must be replaced.

Pb cells can be deep discharged, as long as cells do not revert. And _must_ be charged without delay to prevent capacity loss by sulphate formation.

NiCd cells can be deep discharged and left without fail, as long as cells do not revert.

Reverting cells
With all batteries, reverse charge occurs when the weaker cell(s), after being totally discharged, get charged by the stronger cells in reverse polarity.

To understand, imagine an AA cell and an AAA cell in series on a load, what happens when AAA cell is discharged, while the AA cell still has not completely discharged. The voltage accross AAA cell will revert, destroying it.

[keesje]

A design that allows cells to routinely fall below "lockout" level, is ,IMO incompetently and poorly engineered. (Unless it's a deliberate and cynical ploy to sell more batteries ) but no firm would put profit before dependability,would they?

Steve, thnx. Exactly what I'm wondering about. Next to the battery uncontrolled heating problems.

[EEngr]

A design that allows cells to routinely fall below "lockout" level, is ,IMO incompetently and poorly engineered. (Unless it's a deliberate and cynical ploy to sell more batteries ) but no firm would put profit before dependability,would they?

Routinely, yes. But one of the functions of the battery systems is to provide power in an abnormal (emergency) situation. In this case, its better to kill the battery by running it dead than to lose the aircraft.

The problem now becomes: How do you differentiate between an abnormal condition and some maintenance person leaving a switch on at the end of shift? Without incorporating a bunch of air-ground and other logic with its failure modes?

[Pub User]

The problem now becomes: How do you differentiate between an abnormal condition and some maintenance person leaving a switch on at the end of shift? Without incorporating a bunch of air-ground and other logic with its failure modes?

Is that really such a problem for a machine with a core computer system? I would have thought it would just be a matter of a software tweak.

[Lyman]

Imaginary math can also cause a disastrous result.... One that may need some real math. Or engineering.

Hubris....

[saptzae]

@Pub User

Is that really such a problem for a machine with a core computer system? I would have thought it would just be a matter of a software tweak.

Practically all technical details were discussed already, at least twice. Time for a little philosophy.

• Assumption is the mother of all failure
• Fix one problem, create two anew
• Shades of gray, no black and white
• When to and when not to, that is the "to be or not to be" question
• I like to read much more than to write

[inetdog]

Over-discharging effectively destroys the cell. charging to the limit brings a possibility of spontaneous combustion. therefore, like the pilots keep clear of "Coffin Corner" and" Stall" these batteries demand usage in the "safe" part of their envelope.

A design that allows cells to routinely fall below "lockout" level, is ,IMO incompetently and poorly engineered. (Unless it's a deliberate and cynical ploy to sell more batteries ) but no firm would put profit before dependability,would they?

One of the potential advantage of Lithium batteries is that they have about 20% of their energy storage capacity below the point of no return state of charge (SOC). This means that in a true emergency, you still have 1/5 of the nominal battery energy available as long as you are willing to replace the battery after the incident.
But this raises the problem of how to make use of this reserve. You have to either set the Low Voltage Cut Off (LVCO) inside the battery pack to allow discharge into the reserve if nothing outside the battery itself stops the current flow at a higher voltage, or you have to allow for some sort of signal to override the LVCO in the case of a true emergency.
I have no idea which approach the system designers took, if indeed they planned for the use of the reserve capacity.
But even if they set the LVCO above the point of no return, I can see that they would want to latch this condition in case the discharge to that point may have actually taken one or more cells into the danger zone. Possibly after the battery is returned to the factory, it can in fact be recharged after testing. But in terms of certification, etc. it would probably make more sense just to do a quick forensic analysis and then discard it.
That strategy relies on over-discharge being an infrequent event, and that does not seem to be the case here.

One therefore makes the assumption that all charge/discharge functions must go via the "charger"which is then a glorified power-supply/filter as seen by the battery.

But, as described, the charging algorithm, specifically the fast charge without overcharge part, was designed by the charger manufacturer, and would require some significant exchange of data between the charger and the battery's internal management system. Did Yuasa design part of the overall charge control (not just balancing) into the battery pack based on instructions from the integrator? Or does the charger receive full data from the battery and decide what voltage to apply to the pack as a whole? Is the information flow one-way or bidirectional?

[saptzae]

Coffin corner

I mentioned it before, the batteries are rated at 4V, so about 80% of LiCo limit charge at 4.2V

@inetdog
Concur with your 20% spare margin at the expense of destroying batteries.

But, I don't buy this, inline what we have seen and documented in this thread, and according to the NYT report, ANA clearly found several operational failures which had nothing todo with "operational misuse".
http://www.nytimes.com/2013/01/30/bu...anted=all&_r=0

After all, what is the foundation of this business? Following procedures.

[Lyman]

But that is the frustrating thing. Thus far, no one is privy to Boeing's "intentions" with the new technology.

It strikes me that is the basis of discussion at this point. Muddying "Emergency" (back up) duty with normal ops is the challenge, one apparently not met.

The mitigating conditions speak of pretty aggressive outcomes, fire, smoke, collateral damage to other systems, etc. For an Authority, FAA, JTSB, to allow for such outcomes in an emergency is one thing, but an expectation of such dangerous behavior becoming a "normal" expectation, (non emergency, non "back-up"), cannot possibly be the purpose of what essentially could be interpreted as a "waiver" for other than "abnormal" ops.

Could it?

So, here is the predicted outcome: Airworthiness is the call of the FAA, so if not met, the responsibility can be laid at the Feet of FAA, letting Boeing off the hook from some pretty severe sanctions. But, the FAA has no Money, so Boeing will pick up the tab for "Continued Airworthiness".

That is the way it is set up, grousing about it is useless. I hope it happens quickly and safely.

[hetfield]

@Old Engineer

Thx for the explanation

[Uncle Fred]

Just have to say Gents (or Gentesses as the case might be) that this is PPrune at its best.

I fly airplanes, not build them, and thus have to march quickly to try to keep pace with the more detailed engineering explanations. Yet there seems to have gathered a good group who aspire to make the unknown, known and it benefits us all. Many thanks for the reasoned and searching dialogue. Good work.

Never would have thought that a discussion of batteries would be so interesting. Maybe next it will be tyres or fan blades...

[RR_NDB]

Batteries are increasingly complex

Hi,

Uncle Fred:

Never would have thought that a discussion of batteries would be so interesting.

A long time ago i started to respect these devices that many people consider not so important. In one opportunity in winter time i kept a Ld acid battery being night long charging (~ 2 Amps) and before leaving with my VW beetle when disconnecting the leads a loud hydrogen explosion (inside the car) made me respect the issue a litle more.

...group who aspire to make the unknown, known and it benefits us all.

Precisely, this is the motivation behind our diligent work. These devices are not simple, Boeing, NTSB and many professionals in the industry are learning (technically speaking) and trying to do the best for Aviation Safety.

The message to the high rocks of the entities involved including FAA is clear. The issue deserves kind attention.

Thanks for your incentive.