In-Flight Fires Wreak Havoc With Systems Reliability

Awkward emergency equipment and erratic displays complicate coping
There is nothing like the whiff of something burning to have an effect on a pilot akin to a jolt of chilled water to the heart. Fire, and especially an electrical fire - with its potential to directly attack the "nervous system" of a modern jet - is at the very top of the typical pilot's primal fears.

In a recent survey of nearly 100 reports of in-flight fire and smoke, the pilots' hit upon consistent themes: notably the need to cut power to the perceived source of an electrical malfunction, the imperative to land as soon as possible, and the need for more demanding simulator training while wearing emergency equipment.

As the saying goes, fire in an airplane is fundamentally different from fire in a building on the ground because the airborne occupants cannot get out. Buildings can be evacuated. Airplanes must descend and land, oftentimes with fire raging, or smoldering dangerously, with power to key systems shut off.

In-flight smoke and fire events occur with sufficient frequency to generate a thick stack of confidential reports over a three-year period (2000-2003) submitted by pilots to the Aviation Safety Reporting System (ASRS). The database is maintained by the National Aeronautics and Space Administration (NASA) on behalf of the Federal Aviation Administration (FAA).

The rich and varied collection of smoke and fire events has been distilled to a selection of cases showing the many ways the burning odors are first noticed, the initial wisps of smoke first appear, the lights flicker, or the instruments suddenly display a problem message and pilots are forced to shift gears from a routine mindset to the hyper-vigilance of coping with an emergency.

The pilot's reports are de-identified in the ASRS system, and the cloak of anonymity appears to diminish any reluctance they may have in confessing to their own errors. The confidentiality also tends to create an environment that encourages plain speaking. The virtue of ASRS is that a vivid word picture may be worth more than a stack of statistics. That value is plainly evident in the abbreviated vignettes that follow. From these sometimes harrowing accounts, consistent themes emerge:

It is often difficult, if not impossible, to locate the source of smoke or fire.
Circuit breakers cannot be quickly located, extending the dangerous time that power can feed an electrical fire.
Instruments malfunction, flicker, blank out, give false or possibly misleading readings. The erratic pattern of failures can confound trouble-shooting. Of course, seemingly unrelated system indications can be a symptom of cascading failure as electrical arcing and fire eats its way through bundles of wiring or closely packed avionics. Fire compromises the "trust your instruments" drilled into every pilot. Rather, "trust but verify" is perhaps the more appropriate guidance.
Personal protective equipment can be stored in an awkward location, not easily retrieved in an emergency, and communications while wearing the equipment can be difficult - at the very moment when intra-crew and external communications with the ground are critical.
There is enormous time pressure when faced with in-flight fire. The workload can get very high, very quickly.
Events progress quickly from the first indication of a problem to an emergency landing.
Overweight landings are frequent.
Stress and a fixation on fire can lead to "tunnel thinking" and errors in other areas, such as altitude deviations or misreading instructions from air traffic control.
Passenger and flight attendant anxiety can contribute to the stress level in the cockpit.
A third person in the cockpit, be it a relief pilot called forward from the cabin or a mechanic riding in the cockpit jumpseat, can aid significantly in executing checklists and trouble-shooting.
Smoke often comes from bad seals, leaks and such in engines, auxiliary power units, and air conditioning equipment. Smoke detectors in air conditioning ducts would help greatly in distinguishing these oftentimes less dangerous cases from the ravages of full-blown electrical fires.
Support personnel on the ground may lack experience, may not be the most helpful and, in overseas flights, may not speak adequate English.
The proliferation of in-flight entertainment systems has increased the potential for smoke and fire events.
Recurrent training needs to be more realistic, to include forcing aircrews to communicate while wearing their breathing equipment. The potential for "long range intercom" is increased, wherein a pilot mistakenly radios a message meant for fellow aircrew. In the Swissair Flight 111 case, this happened as the pilots facing the fire approached the moments of greatest stress.
These issues are brought to life in the accounts that follow:

Smoking battery

Case: B777 in cruise at 37,000 feet declared an emergency due to a main battery overheating with smoke and fumes.

Details: "Just prior to ... an EICAS [engine indication and crew alert system] message battery charger main ... changes to main battery overheat. SATCOM [satellite communications] patch to maintenance to discuss concerns over message and lack of information to deal with message. Curious of battery location, maintenance calls back with main battery location ... FO [first officer] flying jet. Captain elects to check E&E [electronics and equipment] compartment to visually check battery. Captain finds battery warm and all appears normal. New call to maintenance to discuss ways of dealing with situation if it occurs again. Believe situation under control.

"Continue [to] monitor electrical page. At ... main battery charger reengages voltage 27 charging amps 27, rapidly rises to 26V 62 amps.

"Captain elects to go to E&E compartment due to previous knowledge of location and pull CB P320 K-9. Upon entering E&E, battery is rapidly emitting smoke. No time for PBE [personal breathing equipment]. Time critical. Captain pulls P320 K-9. Captain back in cockpit. Test, brief to #1 FA [flight attendant], PA [public address] to pax [passengers], mayday call to ... nearest suitable [airport].

"Request ARFF [airport rescue and firefighting] with infrared capability. Battery overheat message back. Cleared direct [to airport]. FO flying. Captain check lists smoke fumes fire electrical, due to fumes in cabin.

"Normal landing. Stop on runway. Fire crew checks aircraft with infrared and visual of battery in compartment.

"Reporter revealed the following information: The second trip to the equipment bay was forced due to the rapid rise in amperage 62 amps and the requirement to pull the charger circuit breaker. The reporter said on entering the equipment bay the battery was emitting smoke and the circuit breaker [CB] immediately pulled. A short time later, the battery overheat message was displayed on the [EICAM] status page. Reporter states the first [EICAM] message was a low level status message that required no action by the crew. The reporter said no procedures were found in the flight operations manual on an overheated battery and no CB location charts are carried in the cockpit any more. After the second trip by the captain to the equipment bay to pull the battery CB, the battery overheat message came on. When on the ground, maintenance decided to swap the APU [auxiliary power unit] battery with the main ship's battery. When removing the main ship's battery the technician incurred a skin burn to his arm from the battery case."

Issues/observations:

A two-man crew with the captain not on oxygen and entering a smoke-filled E&E compartment?
Checklist deficiency, plus no procedures in the flight operations manual?
No knowledge of battery location?
Critical [battery charger] CB located in E&E compartment and requiring entry for deactivation?
Availability of portable breathing apparatus and full-face smoke mask?
No flight attendant trained to do this E&E entry job?
CB location charts not carried?
Reliance upon establishing radio contact for advice from maintenance?
Impact upon ETOPS [extended operations]?
Can the main battery be disabled and the APU battery utilized as an alternate last resort source in flight?
What might have happened had the CB not been pulled?
Multiple breaker trips

Case: A320 crew had multiple circuit breakers trip and cockpit smoke after takeoff.

Details: "Climbing through 10,000 feet ECAM [electronic caution alert module] circuit breaker W-Z. FO was flying. I looked back at CB panels and saw static, PHC [probe heat computer], angle of attack, pitot, TAT [true air temperature] DHC3 were popped out. I took aircraft over and had FO try to call maintenance controller. Shortly after, we smelled and saw electric smoke in cockpit. I declared an emergency, asked for clearance back to LAX [Los Angeles]. Did QRC [quick reference checklist] smoke and fumes check. ATC [air traffic control] asked if I would go to Long Beach. It looked like low overcast at Long Beach ... I said LAX, ASAP [as soon as possible] ... Light smoke was gone, so I had FO lift his mask to see if he smelled smoke. The smoke and burning smell was gone. I told FA what had happened, that everything seemed to be okay and we would return to LAX for a normal overweight landing ... Landing was normal. Fire trucks followed us back to gate. Note: with oxygen mask on, the sun's glare on the mask made it very hard to see [autopilot selector panel, normally on A320 glareshield], PD [primary flight display], ND [navigation displays].

Issue: Mask design. Is emergency equipment optimally user-friendly in a time of high workload and stress, to include pilots who wear spectacles?

Highly stressful scenario

Case: B747-200 cargo flight, smoke in the cockpit and an altitude deviation.

Details: "Shortly after descending through 10,000 feet ... the FE [flight engineer] reported smoke in the cabin. Bluish-colored smoke was indeed evident ... the FE went briefly downstairs to make certain there was not a fire in the main cargo area. When the FE returned the FO and I were already using oxygen masks. I told the approach controller ... that I wanted the emergency equipment standing by. We received some delaying vectors as we were descending to the assigned 4,000 ft. [altitude], and there was considerable amounts of conversation both on the interphone within the cockpit, as well as with ATC. Jumping from one audio source to another (interphone button on the audio selector panel, then pushing the Com #1 button for ATC) was a significant factor, due to running emergency, approach and landing checklists, as well as normal commands and responses (flap selections, landing gear, etc), we leveled off at 4,000 ft. briefly but the PF [pilot flying] was hand-flying and inadvertently descended to 3,600 feet ... during a high workload time.

"On several occasions, I communicated on the 'wrong audio button' and may have used language on one occasion that was inappropriate for standard ATC communications.

"The altitude discrepancy was quickly corrected. I believe we also missed a heading assignment sometime during this high workload time frame.

"The difficulty of communicating with oxygen masks on, using at least 2 sources of audio, while performing normal, abnormal and emergency checklists under a highly stressful scenario of smoke and possibility of fire while airborne cannot be overemphasized."

Issues:

Familiarization and acclimation training. Perhaps crews could regularly train for the real thing by using masks when not under the stress of an actual event.
Is emergency equipment maximally user-friendly in a time of high workload and stress?
Altitude bust a possible byproduct of 'task fixation' in the emergency.
'I missed it!'

Case: MD-11 has smoke in cockpit.

Details: "Smoke of unknown origin while in cruise. Emergency landing in Atlanta.

"When asked by the FO what to enter on the FMS [flight management system] fuel dump page, I responded with the default of 430K pounds (no change). However, I added that this would give us 80K pounds of fuel remaining. We were both on oxygen and my [audio] pick-up was weak unless I held it close to my lips. All the FO heard was the end of my statement, so he set 48,000 pounds in the dump window. I missed it! We landed 45,000 pounds overweight, but well within runway allowable, thanks to relief pilot, who caught error.

"Lesson: answer the question and keep it simple when in this type of situation. As in a lot of scenarios such as this, communications was a key factor in the outcome. Continued practice during simulator training working various abnormal situations with the oxygen masks on would be helpful."

Issues:

Recall the military aphorism to train like you fight and fight like you've trained.
Is emergency equipment maximally user-friendly in a time of high workload and stress?
'No company procedure'

Case: A310 cargo flight with a malfunctioning transformer rectifier on #2 engine producing smoke.

Details: "I was the captain. Departed Memphis. Climbing through 13,000 feet, we heard a 'pop' and the FO's instrument panel went blank for about 2-3 seconds, then returned to normal. The R-hand ECAM indicated a DC electrical advisory. The transformer rectifier #2 indicated 30 volts and 0 amps. There is no company procedure for this situation. Approximately 5 seconds later, we smelled burning electrical fumes. Went on oxygen, declared an emergency for return to Memphis. The 250-knot speed limitation was surpassed due to the emergency.

"[After landing] company flight safety personnel interrupted us as we were attempting to complete our duties while completing the maintenance logbook. We were given a new flight release and told that the airplane next door would be ready soon.

"Overall recommendations: 1) crew be permitted to finish their duties and get appropriate medical attention as required ... 2) the airplane was equipped with the old style oxygen masks and goggles. The mask covers the entire nose and my glasses do not fit correctly, making it difficult to see. I had my arms and hands at all angles trying to focus on the little print on the approach charts. I strongly recommend that all aircraft be outfitted with the full-face oxygen masks ASAP.

"Call back conversation: The PIC [pilot in command] further stated that the position of the smoke masks hampered their attempt to don it. The A300 masks are located by the pilot's knee but the A310's masks are on the side behind the pilot, making it difficult to obtain. The transformer rectifier had 'fried itself' after a maintenance review. The maintenance person said that 'there was nothing you could do' in these cases and that is why there is not a checklist procedure outlining steps for a problematic transformer rectifier. Reference may be made to this anomaly within the miscellaneous section of 'cautions' in the flight handbook. The person interrupting maintenance debrief and [interrupting] the PIC was a new safety person that was on his first assignment and was acting in an over-eager way. At last contact, the PIC was told [by the company] 'they are working on the problem of mask design."

Issues:

The absence of a specific company procedure may be typical of cascading electrical failure scenarios.
Momentary or permanent loss of instruments also can attend a cascading electrical failure.
Crew was hampered visually by ill-fitting smoke goggles.
Note inexperienced ground personnel, a trend that may continue as cash-strapped airlines opt for greater use of contract ground crews. Aircrews in an emergency cannot assume that ground personnel are experienced.
'Misled by smoke'

Case: Canadair regional jet crew reported dense cabin smoke just after takeoff.

Details: "The captain thought, due to the smoke immediately after takeoff, that air conditioning smoke was a possibility ... we turned off the #2 [air conditioning] pack ...Moments later, a 'toilet smoke' caution was received ... The captain advised the flight attendants that we were indicating smoke in the toilet and asked her to verify. She said that it was coming out of all the overhead vents and that smoke was not originating from the toilet. Captain [pilot flying] declared an emergency and advised ATC ... of need to get the aircraft on the ground. The QRH directed me to open several circuit breakers. As I searched for the circuit breakers, the captain advised me that we had been cleared to land. Shortly after, the toilet smoke message cleared ... Captain landed the aircraft and cleared the runway ... [FA] reported a large amount of smoke coming form the aft portion of the aircraft after landing. Captain was hand flying the aircraft and, with no smoke in the cockpit, and FO doing the QRH [quick reference handbook], never donned his oxygen mask. I should absolutely have done this, as smoke should always be taken seriously.

"Total time from takeoff to landing was approximately 4-5 minutes.

"I feel that I became too focused on finding the circuit breakers and that communications between me and the captain broke down. I found myself unaware of our location and when I realized where we were, we were already on short final. The entire event lasted no more than 5 minutes from the report of smoke in the cabin to landing.

"The reporter said the chief pilot has released no information on the cause of smoke in the cabin.

"Though initial findings by maintenance indicated the #2 [air conditioning] pack did completely fail, I was misled by smoke bad enough to set off the toilet smoke caution message. With only a few minutes in the air, I did not have time to completely finish the QRH process."

Issues:

Things happen fast, as evidenced by the short elapsed time typical of events such as this from onset to landing.
The challenge of isolating and reliably identifying the source of smoke.
Trust your instruments?
Difficulty locating the correct CB.
'Fire was coming out of the ceiling'

Case: Piper PA-31 (6-8 place piston twin) fire in overhead rear cabin.

Details: "Climbing through 4,000 feet ... visually saw fire in aft cabin. Fire was coming out of the ceiling ... I immediately began an emergency descent into Opa Locka, Fla. Declared an emergency. I then shut off all electrical. Fire then spread to the seat below as burnt roofing fell.

"Callback conversation with reporter revealed ... the aircraft was equipped with an after market passenger entertainment system. The entertainment system cabin speaker caught fire and ignited the cabin interior headliner. The company has removed all the entertainment systems on all aircraft. The pilot put out the fire by beating on it."

Issues:

Inadequate separation of heat sources from flammable materials.
Yet another unsafe installation of an in-flight entertainment [IFE] system.
'Checking for hot spots'

Case: B757 diverts after a strong odor of sulfur fumes follows a muffled 'bang' in the aft cargo compartment.

Details: "In flight odor resembling sulfur was smelled. The flight attendants were concerned and we asked pax if anyone had lit a match. The smell got worse, following the meal service and now we began searching overhead bins and seat areas ... Unable to find it. We notified the cockpit. FO came out and also smelled odor ... FO returned to cockpit and we went to last coach row. As we sat down, there was a loud bang and rumbling sound from the area of aft cargo. Following the sound, the entire cabin filled with a very strong burning sulfur smell. The fumes ... were so strong that the cockpit crew had to go on oxygen, including the FA that was there ... The cockpit crew had the cabin crew turn off all of the cabin electrical items, with the cabin crew then checking for hot spots on the floor."

Issues:

Suspected improvised explosive device placed on board by a suspicious passenger who had "a handheld device resembling a TV remote control which he said was his remote for his computer that was in his checked baggage."
Difficulty isolating source.
Anxiety level among FA's, all other pax save this one 'cool' pax.
'No indication'

Case: B767-200 crew had #1 engine failure.

Details: "Climbing through FL [flight level] 300, crew heard a loud bang. Aircraft yawed to the left. Almost instantaneous burning smell and some smoke in the cockpit ... captain accomplished engine surge checklist and directed FO to declare an emergency and begin a descent. Engine stabilized at idle. Donned oxygen mask. There were no EICAS messages and no indication of fire.

"We landed with flaps 20 degrees and auto brakes #2. Still no indication of fire. "

Issues:

Nil instrument readings.
Rapid pace of events.
'Detectors did not alert'

Case: MD-11 smoke in cruise.

Details: "A slight haze/smoke was noticed in the cockpit. A visual inspection of main cabin revealed moderate smoke. An immediate diversion ... was initiated. Landing was uneventful. Note: smoke detectors did not alert crew to smoke.

"Once the decision was made to divert, it was approximately 30 minutes to landing. The cabin was depressurized at FL 250. 350 KIAS [knots indicated air speed] was maintained until 15 NM on final. Speed was maintained above 250 knots below 10,000 feet because of the unknown source of the smoke and fumes.

"High speed approaches had been previously trained for in simulator sessions."

Issues:

Difficulty isolating/locating source of smoke.
Reliability of smoke detectors.
Virtue of simulator training in high speed approaches.
Note crew dispensed with 250-kt. speed limit below 10,000 feet.
'Smoke pouring out'

Case: B737-300 had heavy smoke entering the aircraft following takeoff, after being de-iced.

Details: "Got de-iced, Taxied to runway. On takeoff after gear retraction, saw thick white smoke pouring out above and behind the FO - so thick we could not see the instrument panel.

"Continued flying, donned mask, declared emergency. Landed ... uneventfully."

Issues:

Crew's actions exemplified the hierarchy of aviate, communicate, navigate.
Case illustrates virtue of emergency vision assurance system [EVAS] when thick smoke obscures instrument panel (see ASW, Dec. 21, 1998).
'Passengers were becoming very nervous'

Case: MD-80 had smoke, fumes in cabin after takeoff.

Details: "We could not isolate the problem and the haze and smoke appeared to be increasing. Flight attendant said that the pax were becoming very nervous and the sound in a very senior flight attendant's voice indicated to me that the situation was deteriorating. At this point we declared an emergency and returned to JFK.

"Pax already were uptight and anxious due to an earlier security problem during boarding (we departed 1.5 hours late).

"Maintenance found an oil leak in the APU [auxiliary power unit] that was contaminating the pneumatic system with APU turbine oil. This then caused the reported smoke and fumes."

Issues:

Inability to isolate source of smoke.
Rising level of anxiety in the cabin.
Fire in the lavatory

Case: B767 lavatory fire.

Details: "In cruise at FL 370, the flight attendant called from the aft cabin and stated there was a fire in the left aft lavatory and another flight attendant was fighting the active flames with the Halon extinguisher. I declared an emergency and requested priority handling to divert ... We lowered the gear and extended flaps to burn fuel and minimize the overweight landing. I determined it prudent to land 1,000 lbs. over the maximum landing weight for the Cat. II autoland rather than delay longer."

Issues:

Time pressure to land ASAP.
Flight attendants not required to extinguish a realistic fire in training and often face the real thing for the first time in flight.
'An amber X'

Case: B717 engine malfunction.

Details: "Climbing to FL 330, an amber X appeared on the left oil pressure indication. Shortly thereafter, the indication reappeared, showing 0 psi along with 'L oil pressure low' alert. Then the X reappeared, before switching back to reading 0 psi ... we suspected a bad indication ... Maintenance control concurred with our suspicion of a bad indication and advised us to continue to [Atlanta]. As I, the FO and PF, began to advance the left engine throttle, while carefully monitoring all engine indications, we ... heard several loud bangs, along with puffs of smoke in the cockpit.

"The captain declared an emergency [and] took control of the aircraft."

Issue: Trust your instruments?

'Smoke and sparks'

Case: Entertainment system fire on B757 on climbout.

Details: "We were notified by FA the video unit in the cabin was on fire and smoking ... shortly after the notification, several lights illuminated in the cockpit including landing gear, flaps, pressurization and autopilot lights.

"Power was cut off to the video system at the source and CB's were pulled in the cockpit.

"Autopilot and autothrottles would not work properly. We made an uneventful return ...

"Safety concerns are for a video unit fire/smolder to cause this many collateral issues is not good. Because this unit did not draw more power, the CBs did not pop, and perhaps a thermal shutdown feature would be appropriate for aviation video systems."

Issue: Perhaps not really understanding the mechanics of arcing? Arcing involves the carbonization of some types of wire insulation and, carbon being conductive, the charring provides an alternate path for an electric current. The conductor may not draw sufficient current to thermally trip the breaker, yet the whole wiring bundle may still burn (particularly if it has a vertical orientation). With insulation thus destroyed, conductors may touch and fuse together. If the damage progresses that far, CBs will trip by shorting out.

Language barrier

Case: B767 cargo fire warning.

Details: "On takeoff ... we received a momentary main cargo fire warning. This included an EICAS message, fire bell and fire lights at 80 knots on takeoff roll. This warning was extremely brief.

"Once airborne, we coordinated with our company dispatcher via ACARS [airborne communications addressing and reporting system] and the decision was made to divert to Tokyo [Narita] airport. We declared an emergency, dumped as much fuel as possible, and then made on overweight landing. After landing, the main cargo fire warning illuminated again, and remained on until the aircraft was shut down.

"Due to the language barriers of being in the Japan ATC system, communicating our abnormal situation was nearly impossible. It took over 6 transmissions to get them to understand that we were declaring an emergency and that we needed to land ASAP. Most importantly, once we landed, it was impossible to talk with the fire crews. We could not understand them, and they could not speak any English. Therefore, the situation became very hectic and chaotic. Not being able to communicate clearly with ATC or [ARFF] personnel would have made this a very dire situation, had we been on fire."

Issues:

Erratic instruments.
Language problem. Basic proficiency in routine messages for non-native English speakers does not mean the ability to communicate effectively in abnormal situations.
'A large blowtorch type flame erupted'

Case: B737-300 arcing at cockpit window heat connection.

Details: "While in cruise [on a functional check flight] and approximately one hour after turning on the window heat, I heard a loud cracking or popping noise behind me. I looked back to the CB panel and then turned around and noticed that the electrical terminal connector on my window [R1] ... produced a sparkler type firework.

"I immediately donned my oxygen mask and goggles and saw that the captain and mechanic jump seater had also done the same thing. Captain declared an emergency and commenced a descent and turn back to the coast.

"The cockpit door had remained open the entire flight since there were only three people on board. This probably helped to dissipate the smoke and fumes quicker.

"It is my belief if I had been 'heads down' typing on the FMC panel, the flame or sparks would have made contact with my head and hair. As it was, I had spots of black ash on my shirt and pants. Pilots who apply alcohol based hair sprays or gels should be made aware of the potential for fire and sparks igniting their hair.

"The R1 windshield ... was spitting out black pieces all over the right cockpit area. The next thing we noticed was a substantial amount of smoke.

"A large blowtorch type flame erupted and then almost seemed to recede into the upper corner of the windshield/foam-core in that area followed immediately by another shorter flame that began to continuously burn.

"I turned to the mechanic and told him 'do whatever you need to do to get this under control.' He was checking CBs on both sides of the cockpit [and] a couple minutes later said the power was off to the window heat. He had the fire extinguisher out of its harness and ready to be used if necessary."

"The red rubber boot covering the window heat power terminal was charred to a cinder. Maintenance believes corrosion of the power terminal caused a short circuit."

Issues:

Note seductive, teasing nature of in-flight fire, flaring, receding, only to return again.
Implications for clearing smoke when cockpit doors are locked on revenue flights.
Can secure cockpit doors be opened from the cabin once pilots start shutting off electrical busses? Most cockpit doors feature a manual reversionary opening mode (after electrical failure or turning off busses) that allow pilots to open up from the inside. What about when the captain collapses from smoke in the E&E bay and there's no autopilot (due to electrical failures)?
Mechanic's actions almost like having a flight engineer in the cockpit.
'Retracting and extending on their own'

Case: A320 in cruise experiences severe arcing.

Details: "I began to notice the cockpit lights (overhead and panel lights background) flicker as if there were a power interruption. There weren't any ECAM messages.

"The purser called and informed us that the right side cabin lights were flickering on and off. Additionally, the video screens on the right side were retracting and extending on their own. At the same time the FO's navigation display and PFD [primary flight display] flickered on and off, including moments of loss of flight data ... The ECAM began to cycle messages about the number 2 yaw damper, pack control, pitch trim, flight warning computer, autopilot and possibly others. The purser then informed us he smelled smoke.

"At this time, we began an immediate diversion.

"At no time during this event did the ECAM indicate this was a generator loss of phase problem ... Maintenance found a feeder wire pin in the firewall connector burned with evidence of severe arcing. The reporter stated the failure of the ECAM system to warn of a right bus intermittent loss of phase was ... discussed with the company."

Issue: One cannot expect a failing electrical system to have the integrity to drive the ECAM warning system and "tell on itself." The situation is analogous to the fire truck that is itself on fire. Pilots may be back to "20 guesses" about the true nature of the problem.