Go Back  PPRuNe Forums > Flight Deck Forums > Tech Log
Reload this Page >

787 Batteries and Chargers - Part 1

Tech Log The very best in practical technical discussion on the web

787 Batteries and Chargers - Part 1

Old 30th Mar 2013, 21:31
  #1061 (permalink)  
Join Date: Dec 2006
Location: Florida and wherever my laptop is
Posts: 1,350

Does the 787 system have such parasitic loads? Stupid question actually, because the batteries own built in charge monitoring system is one instance. But then, does the 787 system have such a 'float' mode to avoid this state of charge cycling and the inevitable low points in the capacity curve?

Its possible that the only way to fly with a LiCoO2 is either to provide the float voltage (unsafe) or to accept the lower point on the available charge curve as the 'best' avalable charge for design/certification purposes and thereby have to oversize the battery to allow for it. One would need to peek inside the head of the systems engineer responsible to know how this decision was made, I'm sure.
I don't quite follow the logic. The charge monitoring system should not act as a significant continual drain the battery. All other 'parasitic loads' would surely be met by the generators that would be running if the battery is being or has been charged.

It does seem that assuming that the design assumes a NiCd charging profile rather than that required for a LiCo battery is assuming that Thales and Boeing have both made a rather basic error. Do you really think that is the case?
Ian W is offline  
Old 30th Mar 2013, 23:07
  #1062 (permalink)  
Join Date: Jan 2011
Location: Seattle
Posts: 677
The charge monitoring system should not act as a significant continual drain the battery.
Define 'significant'. While it is possible to build a microcontroller-based system that has sub-milliampere current demands, I'm not certain of the technology of the monitoring equipment inside the battery. Its possible that on the shelf it runs in a sleep mode, with a few tens of microamperes load. But when installed and communicating with the charging system and other airplane equipment, it draws far more.

All other 'parasitic loads' would surely be met by the generators that would be running if the battery is being or has been charged.
On the 767 and 747-400 there are a few loads powered off the battery bus at all times when the bus is energized. Although this didn't amount to more than a few amps, it was supplied during normal operations by the battery charger's T-R (constant voltage) mode. These sorts of loads together with the battery internals are what I consider to be parasitic. That is, always connected (unless the bus is switched off) and not switched over to other sources.

This bus topology may have been carried over to the 787 from models with NiCd batteries, where floating the battery was a non-issue.
EEngr is offline  
Old 1st Apr 2013, 03:29
  #1063 (permalink)  
Join Date: Jun 2009
Location: NNW of Antipodes
Age: 78
Posts: 1,330
@Kenneth House

Your succinct description of the correct method of charging Li-ion batteries is appreciated.

However, the apparent failure of the Charger/BMU to cease charging the battery at CV (4.025V) once CC/20 (65A/20=3.25A) was reached, must have been occurring often. Deformation of the affected cells must also have been occurring incrementally throughout those periods. So that begs the reasoning behind Boeing's destructive test program, which if followed to the letter, has failed to take account of compounding cell deformation over a longer period.

Page 33 of the NTSB's Interim Factual Report includes the following Boeing test procedure for FAA approval -
Battery testing consisted of full-performance, environmental qualification, and destructive tests. The destructive tests included external short circuit (low and moderate impedance shorts at battery terminals), overcharge (charge battery at 36 volts for 25 hours), high-temperature storage (185º F for 18 hours), and over-discharge (discharge battery to zero volts) tests. Boeing indicated that the tests found no evidence of cell-to-cell propagation failure or fire.
Note that no mention of any cell deformation is made, rather NO "cell-to-cell propagation failure or fire" met the spec..

Last edited by mm43; 1st Apr 2013 at 03:37.
mm43 is offline  
Old 1st Apr 2013, 07:31
  #1064 (permalink)  
Join Date: Oct 2007
Location: London
Age: 77
Posts: 479
Lithium charging

Kenneth House,

You mention a cutoff voltage of 4.025 v should that not read 4.25 v?
Or is it that 4.025 is chosen as a safety buffer from the absolute maximum lithium voltage of 4.25?

Last edited by RetiredBA/BY; 1st Apr 2013 at 07:53.
RetiredBA/BY is offline  
Old 1st Apr 2013, 14:07
  #1065 (permalink)  
Join Date: Mar 2013
Location: huntsvegas
Posts: 41
But where is the test data?


The characteristic curve graph on page 2 of the LVP65 datasheet indicates 4.025 volts. i would attach an image showing this but don't seem to have image upload privileges.


i did see that statement in the report about passing an overvoltage charging test and was quite puzzled. i would love to see that test data. So far i am at a loss to understand or explain how the battery could survive a 36-volt overcharge (4.5vpc) for 25 hours without seeing the corresponding current profile--voltage without current is meaningless.

On page 21 of the latest JTSB report there is a graphic showing both battery and cell voltage ranges for the 4 BMUs. i have not translated it yet but it shows some captions at the 36.5 volt pack level and also for the 4.55 volt cell level.

Could it be that the contactor opens when the terminal voltage exceeds one of the BMU limits on the high side just as it does at the low side? -- in which case the 36 volt overcharging claim is a legally true statement, but lacks technical integrity.

Why the curious 25 hour limit? A 2.6 amp contactor coil current would use up the entire pack capacity in 25 hours (2.6 x 25 = 65 Amp-hrs), in which case the contactor would return to it's normally closed position and a large potential would be applied across the depleted cells with a huge charging current exceeding 1C.

i have taken thousands of laboratory measurements of voltage, current and temperature on several dozen lithium iron phosphate (3.2 v nominal) cells while under high-rate (1C) charging and discharging. Even though the Yuasa cells are a different chemistry, i am quite certain that a true 36 volt charging condition (4.5 volts per cell) would rather quickly be spectacular to watch from a suitable bunker.
kenneth house is offline  
Old 1st Apr 2013, 15:33
  #1066 (permalink)  
More bang for your buck
Join Date: Nov 2005
Location: land of the clanger
Age: 79
Posts: 3,513
kenneth house for reference: http://www.pprune.org/spectators-bal...une-guide.html

Interesting series of posts by the way.
green granite is offline  
Old 1st Apr 2013, 18:57
  #1067 (permalink)  
Join Date: Mar 2011
Location: engineer at large
Posts: 1,409
From Boeing Land..

Interesting just how many panels have roasted, unrelated to any battery issues of course!

UPDATE: Boeing 787 flight test postponed
A planned Boeing 787 Dreamliner flight to test new power panel engineering was postponed Saturday
FlightPathOBN is offline  
Old 1st Apr 2013, 20:33
  #1068 (permalink)  
Join Date: Jun 2009
Location: NNW of Antipodes
Age: 78
Posts: 1,330
@Kenneth House
Why the curious 25 hour limit? A 2.6 amp contactor coil current would use up the entire pack capacity in 25 hours (2.6 x 25 = 65 Amp-hrs), in which case the contactor would return to it's normally closed position and a large potential would be applied across the depleted cells with a huge charging current exceeding 1C.
Perhaps we need to look more closely at how the contactor is being controlled via BMU/Charger interaction. I note your neat deduction of the contactor coil current, but suspect there is some "smoke and mirrors" involved in its operation. Could well be that the contactor physically latches in the open position and a low current control circuit trips the latch when it is required to return to the NC position. The BMU detects the Charger over voltage condition and either opens the contactor or prevents it from closing.

If the above is the case when applying the over-voltage test, then you are correct, the only condition fulfilled in such a test was the ability of the Charger/BMU to prevent an over-voltage charging condition, effectively indefinitely. To apply 36V to the battery without any BMU involvement would IMHO certainly result in thermal runaway.
mm43 is offline  
Old 1st Apr 2013, 22:08
  #1069 (permalink)  

Ut Sementem Feeceris
Join Date: Jul 1999
Location: UK
Posts: 3,166
McNerney also said that Boeing was taking the opportunity of the battery fix downtime “to tighten up some things and make sure we’re in good shape as we get this plane back into service.”

Is this supposed to inspire confidence in the product?
A4 is online now  
Old 2nd Apr 2013, 11:18
  #1070 (permalink)  
Join Date: Jan 2008
Location: lancs.UK
Age: 74
Posts: 1,191

"While attention is diverted elsewhere, we'll rectify some of the other deficiencies before they bite us on the ass as well"
cockney steve is offline  
Old 2nd Apr 2013, 13:17
  #1071 (permalink)  
Join Date: Feb 2001
Location: Singapore
Posts: 89
The other problems appear to be the power panels...!
ITman is offline  
Old 2nd Apr 2013, 15:24
  #1072 (permalink)  
Join Date: Mar 2011
Location: engineer at large
Posts: 1,409
Sorry about the link...the Seattle Times has gone to some crap subscription program.

The items in the article elaborated on the issues with the power panels, and gave more detail of several flights, including ferry delivery flights, that had to return due to error messages in the flight deck. Panels were found to have shorted out, with one charred.

There is of course, the story regarding the panel 'fire' on flight test..

A failure in the P100 electrical panel of Boeing’s second flight-test 787 Dreamliner led to Tuesday’s fire, which involved an insulation blanket, Boeing said Thursday.

The company said it was still investigating how the incident happened and what steps might be needed before returning 787s to flight testing.

“We have determined that a failure in the P100 panel led to a fire involving an insulation blanket,” Boeing said. “The insulation self-extinguished once the fault in the P100 panel cleared.

“Damage to the ZA002 P100 panel is significant. Initial inspections, however, do not show extensive damage to the surrounding structure or other systems. We have not completed our inspections of that area of the airplane.”

Boeing has removed the panel and nearby insulation material and is sending a replacement panel to Laredo, Texas, where the jet landed Tuesday, the company said.

The panel receives power from the left engine and distributes it to aircraft systems. Backup systems — including power from the right engine, the Ram Air Turbine, the auxiliary power unit or the battery — are designed to automatically engage in such incidents and did so Tuesday, allowing the crew to maintain positive control of the airplane and land it safely, Boeing said.

“Molten metal has been observed near the P100 panel, which is not unexpected in the presence of high heat. The presence of this material does not reveal anything meaningful to the investigation,” Boeing said.

Inspection of the surrounding area will take several days and is ongoing. It is too early to determine if there is significant damage to any structure or adjacent systems.

As part of our investigation, we will conduct a detailed inspection of the panel and insulation material to determine if they enhance our understanding of the incident.
We continue to evaluate data to understand this incident. At the same time, we are working through a repair plan. In addition, we are determining the appropriate steps required to return the rest of the flight test fleet to flying status.

Last edited by FlightPathOBN; 2nd Apr 2013 at 15:27.
FlightPathOBN is offline  
Old 2nd Apr 2013, 16:28
  #1073 (permalink)  
Join Date: Mar 2013
Location: huntsvegas
Posts: 41
Contactor and BMU operation

According to the NTSB Airworthiness Report on page 37:

The contactor is identified as part number 104CZ02Y01 from the French company Zodiac Aerospace. It is a normally-closed solenoid with single-pole single-throw contacts and does not have a latching feature. On page 32 it states that
the contactor is normally closed and was designed to open in the event of an overcharge situation.

Also from the same report on page 32 is some description of the battery monitoring unit (BMU):

In the event that monitoring thresholds are exceeded BMU1 and 2 will send a signal to the battery charger unit to discontinue charging (overcharge and over-discharge monitor).

BMU3 contains controls for the contactor and provides further monitoring for battery and cell overcharge.

BMU4 monitors for cell over-discharge, high current charge, and contains inhibitions signals that will latch in the event of an over-discharge or a battery high current charge.

In the event that the monitoring thresholds are exceeded, BMU3 and 4 will send a signal to the battery charger unit to discontinue charging.

There was mention in the report that the contactor had dual coils, but i have not found a datasheet for the device to see the electrical schematic.

Last edited by kenneth house; 2nd Apr 2013 at 19:31.
kenneth house is offline  
Old 2nd Apr 2013, 20:41
  #1074 (permalink)  
Join Date: Jun 2009
Location: Oxford, England
Posts: 297
kenneth house:

There was mention in the report that the contactor had dual coils, but i
have not found a datasheet for the device to see the electrical schematic.
Can't say for this one, but it's quite common for contactors of high
rating to have dual coils. The contact pressure required needs a strong
spring to keep contact resistance low, which in turn needs high magnetic
field strength to operate against the spring. The first coil is low
resistance, high current to close the magnetic circuit, while the second
or holding coil is much higher resistance and is enabled by an auxilliary
contact. Typically, the two windings would be in series at rest, with the
high resistance winding shorted out, then unshorted as the solenoid

There was some discussion of the contactor earlier in this thread.
Post #902, for example...

The really worrying thing is that the battery subsystem appears to have
been designed on the assumption that the cells would never leak electrolyte.
If they do, then the bmu boards are quickly toast. The contactor, which
is driven from one of those boards, may never operate, leaving the
battery connected as it self destructs.

I wonder if the long term fix will be to put the boards in a sealed
enclosure within the box, or will they fudge that as well ?. Until there's
more detailed info on the solution, there can be little confidence that
they have done it right. An armoured box doesn't qualify as a solution, imo,
though presentation often wins over substance these days, it seems ...

Last edited by syseng68k; 2nd Apr 2013 at 20:45.
syseng68k is offline  
Old 3rd Apr 2013, 18:36
  #1075 (permalink)  
Join Date: Jun 2009
Location: NNW of Antipodes
Age: 78
Posts: 1,330
@ Kenneth House
There was mention in the report that the contactor had dual coils, but i have not found a datasheet for the device to see the electrical schematic.
Likewise, my search for any info on Zodiac Aerospace electrical data has also drawn a blank.

Thales, as the main contractor appears to have a large number of sub-contractors under its wings. OEMServices seems to be the umbrella vehicle, and Meggitt, Securaplane and Zodiac Aerospace provide the expertise which Thales has sold to Boeing.

The split high/low resistance solenoid winding as detailed by syseng68k would appear to equate to the dual coil mentioned by the NTSB. I suspect that quite some more detail will be revealed when the NTSB undertakes its public investigative hearing (April 23-24) into how the Li-ion battery/charger came to be approved.

The following from the Thales Group website -
Thales offers an innovative and highly secure Li-Ion emergency low voltage system for commercial aircraft. It is the first commercial application of Li-Ion technology anywhere in the world. Thales Li-Ion sub-system comprises:
- a Li-Ion battery
- a Li-Ion an[sic] Ni-Cd battery charger.
- may yet prove to be an embarrassment to not only Thales.

Last edited by mm43; 3rd Apr 2013 at 18:38.
mm43 is offline  
Old 4th Apr 2013, 00:05
  #1076 (permalink)  
Join Date: Jul 2007
Location: Virginia, USA
Age: 83
Posts: 77
I spent the late hours of last night reading the entire corporate page of ZodiacAerospace (it was an e-copy of their annual report). Zodiac apparently made the Zodiac boats that Cousteau was always speaking of. These had flotation sides that were tubes of cloth-covered rubber material. Fast forward and Zodiac Aerospace is making large numbers of aircraft seats, with knowledge of how to affix cloth to a rubberized backing with rubber cement a key factor here, receiving much attention.

At some point, ZA appears to have acquired a company they made into a division called ECE (could not discove its antecedent name). ECE made one major product (taking the single product mentioned in the annual report), namely "contactors". ECE, with the backing of ZA, widened its line to aircraft electrical distribution systems, including panelboards, and some high intensity LED cabin llighting.

ECE now seems to be principally manufacturing in Tunisia (certainly their contactors). Cabin seats are manufactured in several countries (not France), of which Tunisia is one of ZA's larger producton locations.

So ZA began as a maker of rubber boats, and Boeing began as a maker of aircraft. I recall that certain Rolls-Royce turbines of great output had turbine shaft bearings make by a German supplier who had been a maker of motorcycles.

In the matter of the matter of the contactor part number, I think the whole part number would have to be something like "Contactor, aircraft, 104CZ02Y01". The number form has tripped a dim memory-- it seems to derive from the [US] Federal Specification system which saw such developement in WWII. So, 104C identifies the basic contactor wanted, and Y01 is an add-on set of repeating contacts (the Y01 is what I remember ...IIRC of course).

I forget what Z02 was, but it's some sort of add-on device that could also be ordered preinstalled. The repeating contacts would confirm to the supervisory system board that the contactor had actually moved as commanded. I forget what Z02 was, but I want to say Z was for locks. On one job where there were a lot of these puppies (GE had beautiful and massive slick paper catalogs of all this good stuff), I recall on a double-end substation, we had a load bus divided in the center (by some distance), so we had three large breakers (1200 amps) locked by special keys-- only two nonreproducible keys. So the power sources could not be interconnected; also the phase cable across the center separation were additionally protected by cable limiters (a type of fuse) at one end.

This all sounds OT, but it is exactly the configuration of the two battery souces in the 787. Or should be-- that is why I am puzzled that someone said the forward battery could assist the APU battery in starting the APU. Well, of course its DC and you have that diode isolation (that never shorts through, of course). Not at the same time, I hope.

You could also get field kits for more elaborate add-on functions, but these were not available factory-assembled. Possibly mechanical latches or blocking devices were in this category. One thing I recall was an electric drive to reset the bigger breakers, otherwise a crank was used.

So who knows what this contactor situation really was?


PS-- What is the difference between the battery bus and the hot battery bus?
Old Engineer is offline  
Old 4th Apr 2013, 00:12
  #1077 (permalink)  
Join Date: Feb 2001
Location: right here inside my head
Age: 62
Posts: 178
OE. The hot battery bus is "hot", or powered, even while the battery may be switched off. Not so with the battery bus.
3holelover is offline  
Old 4th Apr 2013, 04:10
  #1078 (permalink)  
Join Date: Mar 2013
Location: huntsvegas
Posts: 41
peeling a couple of onions

@ OE, syseng, and mm

Thank you all for the good zodiac research and explanations about the contactor and the second coil/auxillary contact. i tried to sign up on their customer support site to download technical info but haven't been given access yet.

In the latest JTSB report on page 10 is a schematic showing the Hot Battery Bus (HBB) downstream of the Bigass Diode Module (BDM) on the Main Battery positive terminal. The HBB connects to the Captains 28V DC Bus (CAPT) thru the Main Battery Relay (MBR). What is operational function and criticality of the Captains Bus?

Also does anyone know if the FDR data for the ANA/Japanese event is available online similar to that found on the NTSB docket for the Boston event?--it would be interesting to see what the Main battery current signal shows before and while the pack was shorted to ground thru the case.

i thought it was an interesting finding by JTSB that cells 4 and 5 in the Main pack were basically intact with no broken current collectors inside and the pressure relief valves had not opened, unlike cells 1-3 and 6-8.

Some of the lithium battery experts in the electric car world have set up trickle charging experiments to determine how long it takes for a cell to go thermal from overcharging--but none are yet using a LVP-65 Yuasa cell.
kenneth house is offline  
Old 4th Apr 2013, 05:25
  #1079 (permalink)  
Join Date: Jul 2007
Location: Virginia, USA
Age: 83
Posts: 77
@3-HL Thanks for the answer. The "Hot Battery Bus" I would think of as the "tie bus." A tie bus was usually broken in the center, but kept powered from either end with the load roughly divided in half. The tie could only be closed if one of the sources was open. The double ends were sourced from two independent power company lines. They are a bit unusual, and at the customer level usually operated manually.

How's that song go... "Ya gotta know the jargon..." No, "territory." OE

PS-- GE offers only mechanical latching at 400 amps and up. It's electric release of latching. Of course, that's 3-pole stuff from GE. Below 400, latching comes as a kit, field installed, with a separate part number and price.
Old Engineer is offline  
Old 5th Apr 2013, 05:12
  #1080 (permalink)  
Join Date: Jul 2007
Location: Virginia, USA
Age: 83
Posts: 77
Size of APU alternators; amount of frequency variation

Hamilton Sundstrand is the maker (possibly European sales use a different source) of the alternators driven by the 787 APU. These are 225 KVA units (each; I put that = 180 KW) intended to have a phase to phase voltage of 230. That make the phase current about 600 amps.

Hamilton also makes the blower/compressors for cabin air. They note these as being able to operate on a variable frequency feed with a range of 360 to 800 cycles. That gives us the alternator output frequency specification. Boeing has described it as a variable freqquency system.

Hamilton also makes the turbine main engine-driven alternators. There are four of these and I believe they are also each of the save 225 KVA output. They might differ in the gear ratio to the respective turbines.

Between 1974 and 1978, the labs at Curtis-Wright, with major assistance from GE, worked to build and test a turbine-driven alternator with a rating of 150 KVA. This had a permanent-magnet rotor with salient (misspelled 'sahlen' somewhere) poles of samurium-cobalt-- 14 rotor poles and a 9-phase stator winding, cycloconverted and filtered to a well-formed 3-phase output. Curtis-W labs had earlier tested an equal-sized and similarly constructed machine which had a wound rotor (i.e. suppled with external DC).

Of interest is the rotating speed range, from 12000 to 21000 rpm (that is, within 2:1). I'll have to clearify which unit had that speed as the last report page was a errata sheet dealing with that. I will explain the significance of all these variances, down the road.

What I am looking for is if and how these components and their operating ranges might cause variation of the voltage on the "hot" battery bus. Since we have minimal system details, we can just design what major detail is missing.

For starters, a major question is whether the APU altenators have a wound-rotor design (has then to be a DC winding). Anyone know?

Old Engineer is offline  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service - Do Not Sell My Personal Information -

Copyright © 2021 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.