757 All Generator Out

 
A twin engine Boeing 757 experienced a temporary loss of both generators on January 14, 1985 during a flight by Monarch Airlines (registry G-MOND) from Teneriefe to Luton. While the initial pilots report was that only the left bus tripped off line, in a later debriefing the crew reported that both generators dropped off line. The airplane diverted to Lisbon. A flaps-up landing was made with no blown tires or fuse plugs. Upon inspection after landing, there were signs that there had been a wire bundle fire in the left-hand side of the electronics bay. New wires were spliced in and the airplane was ferried to Luton.

The 757 is equipped with three 90 kVA generators, one on each main engine and one powered by the Auxiliary Power Unit, a tail mounted gas turbine engine. Normally the two main engine generators are operating, as was the case here. The APU is only put into in-flight service when one of the main engine generators is inoperative. The design objective in a jet transport electrical system is that no single failure should lead to loss of all generators, short of loss of all engines. In addition to finding the cause of the wire bundle fire, another significant question here is how that could lead to loss of both generators. Distribution wiring in an airplane is normally protected by circuit breakers which should have cleared the faulted wiring without resulting in the tripping of a generator.

What the Pilot Reported

“Cruising at FL.390 centre autopilot engaged. Two loud bangs and the left bus tripped off line and smoke entered cabin at the front row of pax. Cabin altitude started rising so emergency was declared and emergency descent initiated. Set course for Lisbon. Approx. 20 miles from Lisbon the speed was reduced, the gear lowered and flap 1 selected. No indication of flap movement on gauge so flapless approach and landing carried out. A reset was tried during initial descent on the L bus tie breaker, but this caused the return of smoke and fumes so the left bus was left isolated.”

At a crew debriefing on January 16 at Luton, the crew reported that both generators tripped off line. There were numerous visual and aural warnings occurring at the same time as the loud bang. These included leading edge and trailing edge flap warnings. Cabin altitude started to climb and warnings sounded, but descent had already begun so that oxygen mask drop did not occur. The Captain flew the aircraft even though he had less instruments than the First Officer. Other reports were:

  • 1) Capt. station box inop on intercom and No. 2 VHF. No. 1 VHF not tried.
  • 2) No HSI or ADI on Captain’s side.
  • 3) No heading on F/O RDMI.
  • 4) Standby horizon, altimeter and ASI and Captain’s RDMI used to fly the aircraft.
  • 5) Other instruments on Captain’s side showed “off” flags.
  • 6) Both EICAS screens went blank. Bottom EICAS returned when No. 2 generator reinstated.
  • 7) No. 1 Hostess reported smoke issuing from area of forward louvers.
  • 8) APU was started approximately 25,000 ft. Started on first attempt.
  • 9) No. 2 (i.e. right) generator was reset.
  • 10) BTB tried – smoke again reported.
  • 11) Control of cabin regained.
  • 12) Smoke removal/fire drill carried out.
  • 13) Normal flaps selected to position No. 1 – no movement indicated. Alternate selected regardless of the LE and TE warnings which were still evident. No indication of movement but left selection to 1.
  • 14) Flaps-up landing procedure carried out.
  • 15) No auto speed brake on landing.
  • 16) Warnings evident after reinstating No. 2 generator:
    • a) L Bus Off
    • b) L Gen Off
    • c) L Bus Isolation
    • d) R Utility bus was selected off.

The Wire Bundle Fire

Upon inspection, wire bundle damage due to arcing was found in the EE bay between the left generator main power shield and the P70 electrical distribution panel. Arcing occurred only wire-to-wire and did not involve arcing to structure. The damage was at a point where the wire bundle passes beneath the toilet drain line. The immediate area of damage was heavily saturated with toilet fluid. The damaged wire bundle contained the following conductors:
  • 1) Three #10 gauge polyimide insulated copper conductors, supplying the left utility bus, powered by a 35A circuit breaker from the left main bus.
  • 2) Three #8 gauge Nomex braid covered polyimide insulated copper conductors supplying left main bus power to the P70 distribution panel, protected by a 50A circuit breaker.
  • 3) Three #4 gauge polyimide insulated aluminum conductors powering the No.1 Center electric motor driven hydraulic pump, protected by an ELCU. The ELCU is a remotely resettable motor starter and circuit protective device equivalent to a circuit breaker.
The 35A c/b and the 50A c/b were found to have tripped. All three of the #10 gauge wires and all three of the #8 gauge wires were completely severed by arcing. The ELCU which protects the #4 gauge wiring was not known to have tripped. Two of these three wires were severed by the arcing. The wires, after having been removed from the airplane, were sent to an independent testing laboratory in England. It was found that the insulation on the #10 gauge wire had been damaged by the wire marking process. The damage had penetrated the insulating material and allowed the fluid from the toilet leak to come in contact with the electric conductor. The result was wire-to-wire arcing that ultimately resulted in the damage listed above.

Other Findings

The generator control unit contains built-in-test-equipment (BITE) that records events which cause generator tripping. A readout of the memory showed that both generators had tripped due to differential fault protection (DP). This type of protection compares the total generator current at the generator with the current at the main bus by the means of current transformers. These currents are normally equal. Any significant difference in these currents will trigger a DP trip. DP is intended to protect the aircraft from faults in the heavy electrical conductors between the generators and the main bus. Faults beyond the main bus, as was the case in this event, should not trigger the differential fault protection. It appears that the arcing nature of the fault somehow triggered the differential protection.

What Probably Happened

The exact details are fragmented, as is usually the case when something unexpected happens. Since it was differential protection that first tripped the left generator, the bus tie breaker that would allow left bus to be transferred to the right generator was locked open (by design). The left bus goes dead and the Left Bus-Tie isolation light came on. When this occurs the normal operating procedure is to attempt one reset of the left bus-tie isolation switch. We can assume then that the pilot reset the switch in an attempt to re-establish power to the left bus from the right generator. At this time arcing was also re-established and the right generator tripped off, also on differential protection. Then attempting to reset both generators, the right generator came back on, but the left generator remained off.

Since the Captain reported attempting to re-establish power to the left main bus had caused the resumption of cabin smoke it is likely that he left the switch in the isolate position. With the left main bus dead and the bus tie switch isolated the Captains Instrument Bus is shut down. This would account for the loss of the Captains flight instruments.


The cabin pressurization control has two automatic channels, one powered from the left bus and the other powered by the right bus. While cabin pressurization can be controlled manually from battery power alone, there was temporarily a condition of descending cabin pressure. Once the right generator was re-established, cabin pressurization returned to automatic control.

Flap position indication is lost when the left main is dead. This would lead the crew to assume that the flaps were inoperative, but in actuality they would have still been operable. The L.E. and T.E warning lights would be caused by a dead left DC bus. Normally the loss of the left DC bus would actuate the DC bus tie to power the left DC bus from the right DC bus. However, the DC tie is locked out by operation of either bus tie isolation switch to the isolate position, as was probably the case.

Laboratory Arcing Tests

Boeing attempted to duplicate the conditions in it’s laboratory.
Faulting two or three wires together without arcing did not trip the generator, but tripped the circuit protective device instead as it should have. However by creating an electric arc using a salt water solution, the Monarch incident was duplicated. It was never firmly established just why arcing current trips the generator differential fault protection. However, examination of the current traces from the tests shows that there is a tendency for the sine wave of current to contain some DC current component. It would take relatively little DC content in the current to cause saturation of the differential protection current transformers. It is a little known fact that the core of the differential current transformer in the generator is smaller than the core of the current transformer in the generator power shield (EE bay). A small component of DC current would tend to saturate the smaller core before the larger core. I believe that this is the primary cause of tripping the differential protection from arcing faults. It would be prudent in future designs to insure that the specifications for the differential current transformers match the core sizes as well as the turns ratio.

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s