Go Back  PPRuNe Forums > Flight Deck Forums > Rumours & News
Reload this Page >

LATAM B773 complete electrical failure?

Rumours & News Reporting Points that may affect our jobs or lives as professional pilots. Also, items that may be of interest to professional pilots.

LATAM B773 complete electrical failure?

Reply

Old 24th Dec 2018, 03:24
  #61 (permalink)  
 
Join Date: Jan 2011
Location: Seattle
Posts: 602
Originally Posted by tdracer View Post
The 777 can sync the AC buses when needed.
The 777 does parallel generators. But only for the duration of a power transfer (shutting down or spinning up a generator). The IDGs (and APU controller) are equipped with a speed control input that bring two sources into sync by throttling one up or down. Once synced, the new source is paralleled with the system by closing the proper breaker. Once the new source is on line, a tie breaker is opened to isolate the two sources and their respective loads. Buses do not operate continuously tied together.

On the various 747 models, the generators did remain in parallel continuously, each feeding into the four main buses tied together. 747-400 buses will revert to isolated operation for certain autoland configurations requiring isolated and redundant autopilot channels. On 757 and 767, the electrical systems remain isolated. All transfers are break before make, resulting in a short but annoying power interruption.

During normal operation, the 747 is the only model where an electrical fault can be 'seen' momentarily by all buses. Until the protection systems act to isolate the faulted section. Because of this parallel operation and the need to identify the location of and isolate a fault, the 747 has perhaps the most complex protection system. The heart of this system is something called differential protection, which can identify where in the system a fault has occurred and open the correct breakers to isolate it. All other models have much simpler (if any) differential protection systems and may trip more breakers than needed upon sensing fault currents passing through.

Due to the isolated (at most times) operation of the 777 electrical system, no single fault should affect more than the half of the system (left or right) where it occurs. So this particular incident is troubling. One possible cause is that (for some reason) one of the generator control units failed to transfer load to its generator upon engine start. Leaving the left and right main buses powered from one generator through a bus tie breaker. A subsequent fault, or even an overload condition due to one generator available resulted in the good generator tripping off line. Leaving only standby (battery and RAT) sources to power critical loads.
EEngr is offline  
Reply With Quote
Old 24th Dec 2018, 06:05
  #62 (permalink)  
 
Join Date: Dec 2010
Location: Balikpapan, INDONESIA
Age: 66
Posts: 361
Originally Posted by jimtx View Post
My memory is hazy but we synced the generators on the 727 using phase lights. If you rushed the sync you could get a little shudder in the A/C.
As part of my expansion of CV during the Arab Oil Emargo, where no flying gig was available, I did a stint in a nuclear/fossil power plant. They told a story of a many ton generator being twisted off mounts when somehow put online out of phase. Off topic sorry.
Yes, on large alternators that sort of thing can and has happened.
On large diesel sets the most common failure mode from trying to sync out of phase was either to shear the alternator to flywheel coupling bolts or to snap the crankshaft.

Fortunately that is now a thing of the past since auto-sync became common.
WingNut60 is offline  
Reply With Quote
Old 24th Dec 2018, 10:16
  #63 (permalink)  
 
Join Date: Sep 2016
Location: Earth
Posts: 10
Thinking about this some more, I bet when the full story is out inaction or inappropriate action by the crew contributed to the severity of this situation ...
There is nothing worse than having successfully dealt with an abnormal system failure to hear management and engineers tell you that, “that sort of failure cannot possibly happen” and then ask, what did you do wrong?”

Cascade system failures do, despite all the ‘bulletproof’ inbuilt protection, still happen. And when they do, and the cause of failure is discovered, the same people who told you they couldn’t and blamed the crew, never apologise to your face.


Small cog is offline  
Reply With Quote
Old 24th Dec 2018, 17:09
  #64 (permalink)  
 
Join Date: Jan 2004
Location: West Coast Canada
Posts: 3,314
Small Cog

Unanticipated fail modes leading to cascading failures can occur but I would suggest that historically the record shows most of them were exacerbated by inappropriate crew actions. It will be interesting to see the official report on this incident.

Like I said above when the only thing on the PFD and MFD is the makers name, a good outcome is going to depend on the basic flying skills of the crew, something this crew obviously had but sadly is no longer a given on modern flight decks

Big Pistons Forever is offline  
Reply With Quote
Old 24th Dec 2018, 20:11
  #65 (permalink)  
 
Join Date: Nov 2005
Location: Potomac Heights
Posts: 421
Originally Posted by belfrybat View Post
According to the commander, interviewed on TV, power came back after the engines were shut down. Presumably from the APU, on the footage the cabin lights were visible.
I guess we'll know if this was the case if the new procedure from Boeing to address this fault is:
  1. Power up APU;
  2. Power down both donks;
  3. Wait for APU-powered electrical busses to come back on line;
  4. Pray that both donks (or at least one) can be restarted.
SeenItAll is online now  
Reply With Quote
Old 25th Dec 2018, 03:09
  #66 (permalink)  
 
Join Date: Jan 2011
Location: Seattle
Posts: 602
That fault report (thanks JumpJumpJump) is interesting reading.
Question: How do the time stamps relate to the actual flight? Quite a few faults are logged around 20-Dec : 02:55 (Zulu, I assume) and then a few more at 20-Dec : 03:46. This appears to be a few minutes after landing. But since this incident seems to have gone on for around 3 hours, the earlier logged times don't appear to correspond with the onset of the problems.
EEngr is offline  
Reply With Quote
Old 25th Dec 2018, 18:45
  #67 (permalink)  
 
Join Date: Apr 2003
Location: 24.7098N 46.7252E
Posts: 36
Lots of discussion about the AC busses, looks like an ELMS failure to me. Electrical Load Management System going haywire.

I'm pretty sure the ACE's (Flight Control System) are powered by 28v DC from PMG's (Permanent Magnet Generators on each of the engines) which is independent of all the AC power distribution systems. The fly by wire system would likely have been unaffected by the loss of AC power on the three busses.

It is interesting to hear that the RAT did not reconnect AC power to the busses. Were both bus tie circuit breakers tripped? or was the ELMS load shedding everything? How is it the transfer bus remained unpowered with the Rat deployed???

Hat's off to the crew, very impressive presence of mind and aviation skills.
speed2height is offline  
Reply With Quote
Old 25th Dec 2018, 19:06
  #68 (permalink)  
 
Join Date: Sep 2011
Location: FL390
Posts: 249
If like the 787 then ACEs on independent PMG, but PFC on normal electrical network. Hence loss of autopilot and other degradations.
Fursty Ferret is offline  
Reply With Quote
Old 31st Dec 2018, 21:39
  #69 (permalink)  
 
Join Date: Sep 2007
Location: Florida
Posts: 81
Any more news on this incident?
viking767 is offline  
Reply With Quote
Old 1st Jan 2019, 12:11
  #70 (permalink)  
 
Join Date: Apr 2006
Location: uk
Posts: 720
Megan touched on the question of ETOPS (23rd Dec); I'm guessing this was not an ETOPS flight; correct?

But even if not, was the aircraft maintained and equipped as an ETOPS aircraft?

I wonder how things would have gone if an unconnected (or indeed connected) engine failure had been added to the crew's problems, shortly after the electrical failures at 0254 and 0255?
old,not bold is online now  
Reply With Quote
Old 1st Jan 2019, 13:36
  #71 (permalink)  
 
Join Date: May 2018
Location: London/Fort Worth
Posts: 0
All 777's are ETOPS qualified - its not like the old days where you had some 757's that were ETOPS and some that were not and they had different components on them.
BAengineer is offline  
Reply With Quote
Old 1st Jan 2019, 22:28
  #72 (permalink)  
 
Join Date: Apr 2000
Location: london
Posts: 29
Two decades years of certifying 777 electrical systems and I have never seen anything like this before.

Having read the limited public informational available (particularly the AHM report), and just thinking aloud, I can't help thinking...

The 777 has only a single Back Up Converter controlling both Back Up Generators and both Transfer Busses.

..and wondering...

Would this incident still have happened if each Back Up Generator had it's own, completely independent, Converter?

I am not speculating on the cause - I just don't have access to the detailed information to answer this question.
max motor is offline  
Reply With Quote
Old 2nd Jan 2019, 05:13
  #73 (permalink)  
 
Join Date: Mar 2018
Location: Canberra
Posts: 76
Originally Posted by max motor View Post
Would this incident still have happened if each Back Up Generator had it's own, completely independent, Converter?
I am not speculating on the cause - I just don't have access to the detailed information to answer this question.
At least they have the entire undamaged aircraft, so hopefully the real cause will be found and remediated.
Dee Vee is offline  
Reply With Quote
Old 2nd Jan 2019, 17:28
  #74 (permalink)  
 
Join Date: Jan 2011
Location: Seattle
Posts: 602
Originally Posted by max motor View Post
Would this incident still have happened if each Back Up Generator had it's own, completely independent, Converter?
It does appear that the backup generator/converter system was involved and may have been the root cause of this sequence of events. If so, there are two possible solutions: Provide separation within the backup system so as to eliminate a single fault from taking both main AC buses with it. Or make ELMS smarter so as to clear faults with greater selectivity and not disconnect unfaulted sources and buses.

It all comes down to certification costs. While redesigning the backup system to provide additional channel separation would seem to be a more sound solution, there is a strong motive to 'fix it in software'. So I'm betting on a patch to ELMS.
EEngr is offline  
Reply With Quote
Old 2nd Jan 2019, 21:08
  #75 (permalink)  
 
Join Date: Mar 2018
Location: Canberra
Posts: 76
Originally Posted by EEngr View Post
It all comes down to certification costs. While redesigning the backup system to provide additional channel separation would seem to be a more sound solution, there is a strong motive to 'fix it in software'. So I'm betting on a patch to ELMS.
Surely it comes down to safety.

If the design is faulty it needs to be fixed, "putting a patch on it" but leaving the root cause to rear its ugly head some time in the future, isn't an option.

Luckily this time no lives were lost, that might not be the case if it happened again.
Dee Vee is offline  
Reply With Quote
Old 3rd Jan 2019, 06:34
  #76 (permalink)  
 
Join Date: Feb 2006
Location: Europe
Posts: 66
Originally Posted by Dee Vee View Post

"putting a patch on it" but leaving the root cause to rear its ugly head some time in the future, isn't an option.
Please remind me how the B787 battery issue was "solved". Was the root cause addressed or a "safe(?)" box put over the battery?
Bidule is offline  
Reply With Quote
Old 3rd Jan 2019, 08:08
  #77 (permalink)  
 
Join Date: Jul 2013
Location: Everett, WA
Age: 63
Posts: 1,915
Originally Posted by Bidule View Post
Please remind me how the B787 battery issue was "solved". Was the root cause addressed or a "safe(?)" box put over the battery?
Well, the battery was completely redesigned. The battery charging system was completely redesigned. Then, because the damage to the event batteries was so severe they were unable to conclusively determine root cause, as a failsafe they put the battery inside a steel box that would contain any possible battery failure.
Appears to have worked - prior to the redesign there were two failures on 50 aircraft in a few months. Seven years later, with about 600 aircraft in service, there has been a grand total of one battery failure - which due to the battery redesign was contained to a single cell.
So yes, they fixed the 787 battery problem by putting a box over the battery
Constantly repeating the myth that all Boeing did was put the battery in a box doesn't make the myth true.
tdracer is offline  
Reply With Quote
Old 3rd Jan 2019, 15:36
  #78 (permalink)  
 
Join Date: Jan 2011
Location: Seattle
Posts: 602
Originally Posted by Dee Vee View Post
"putting a patch on it" but leaving the root cause to rear its ugly head
Fixing ELMS logic to clear faults more selectively (i.e. not dumping all the sources) is a legitimate fix. One may not be able to identify all possible future fault conditions. Even by redesigning the backup converter. The redesign of which might create a whole new batch of unforeseen failure modes. So handling them properly would be a good approach.
EEngr is offline  
Reply With Quote
Old 3rd Jan 2019, 23:22
  #79 (permalink)  
 
Join Date: Mar 2018
Location: Canberra
Posts: 76
Originally Posted by EEngr View Post
Fixing ELMS logic to clear faults more selectively (i.e. not dumping all the sources) is a legitimate fix.
Its providing more (possibly better, possibly worse) complexity, but not addressing the SPOF, therefore not a legitimate "fix", but merely an enhancement that hasn't removed what can still turn out to be an un-recoverable failure.
Dee Vee is offline  
Reply With Quote
Old 6th Jan 2019, 02:32
  #80 (permalink)  
 
Join Date: Jan 2011
Location: Seattle
Posts: 602
Originally Posted by Dee Vee View Post
but not addressing the SPOF,
Had ELMS not dumped two perfectly good generator sources, there would have been no single point of failure. At most, a change in the ETOPS capabilities.
EEngr is offline  
Reply With Quote

Thread Tools
Search this Thread

Contact Us Archive Advertising Cookie Policy Privacy Statement Terms of Service