Pan American World Airways 160
Taken in isolation, this accident appears to be another unfortunate mishandling of Hazardous Materials, and there is no doubt that if it were not for the HAZMAT, the event would have never happened. But there is a case to be made that had the crew worked together more efficiently, they could have landed successfully. But their airline did not foster that kind of environment.
This was another incident in a series of 13, 11 of which pointed to a problem with the Crew Resource Management Culture at Pan American World Airways at the time. They were able to reverse this culture and became one of the safest airlines in the world.
Photo: N458PA, (www.baaa-acro.com)
Click photo for a larger image
- Date: 3 November 1973
- Time: 09:39
- Type: Boeing 707-321C
- Operator: Pan American World Airways
- Registration: N458PA
- Fatalities: 3 of 3 crew, 0 of 0 passengers
- Aircraft Fate: Destroyed
- Phase: Approach
- Airports: (Departure) New York-John F. Kennedy International Airport, NY (JFK/KJFK), USA; (Destination) Glasgow-Prestwick Airport (PIK/EGPK), United Kingdom
There are a few things to keep in mind when reading this NTSB accident report.
First, the occurred on November 3, 1973. The aviation community's consensus about cabin fires was to fight the fire immediately, land when you can. The industry's thought process would make a radical change — land as soon as possible, fight the fire if you can — after Air Canada 797, but that was still ten years into the future. More about this: Cabin Fire.
Second, Pan American World Airways at this time still embraced a culture that gave the captain total power and subordinate crewmembers were reluctant to challenge decisions. In that environment, I could see the second officer acting independently on things he considered purely flight engineering.
Finally, the NTSB report doesn't pay enough attention to the enormity of the task of flying this airplane without the yaw dampers. The report also implies cockpit visibility was so bad that frequencies couldn't be changed and perhaps the position of the speedbrake handle couldn't be seen either. But reading through the Cockpit Voice Recorder transcript seems to contradict that.
Note: the NTSB report uses local time for the body of the report, but Zulu time for the Cockpit Voice Recorder. I've converted everything to local time.
[NTSB Report, ¶1.1]
- Clipper 160 reported level at FL 310 at 0850. As the flight approached Sherbrooke VORTAC 100 miles east of Montreal, Canada, at about 0904, it advised Pan American Operations (PANOP) in New York that smoke had accumulated in the "lower 41" electrical compartment, and that the flight was diverting to Boston.
- At 0908, Clipper 160 advised Montreal Center that they were level at FL 310 and wanted to return to JFK. Montreal Center cleared Clipper 160 for a right turn to a heading of 180°.
- At 0910, Clipper 160 advised PANOP that it was returning to New York and that the smoke seemed to be "getting a little thicker in here." At 0911, the crew advised PANOP that they were now going to Boston and that "this smoke is getting too thick. They also requested that emergency equipment be available when they arrived at Boston. During this conversation, the comment was made that the "cockpit's full back there."
[NTSB Report, ¶2.11]
- According to the CVR, the crew donned oxygen masks at 0911 and put on their smoke goggles at 0912.
- At 0914, they asked to remain on the current radio frequency because "its too hard to change." This remark infers that the smoke in the cockpit was so dense that they had difficulty seeing the frequencies on the control panels. The crew, however, did not at any time become alarmed by the situation.
- Other conversations recorded on the CVR indicate that the crew was firmly convinced it was an electrical problem.
This "infers" runs contrary to the CVR, which shows the crew tested their radio altimeters just prior to requesting the single frequency and set their barometric altimeters after. It is more likely they wanted the single frequency as a precaution, not that they had limited cockpit visibility at the time.
[NTSB Report, p. 66 (Cockpit Voice Recorder)]
I/P-3: I can't find a thing wrong back here
I/P-1 What's that?
I/P-3 I can't find anything wrong
I/P-1 Okay, uh, maybe it's in a package
I/P-3 Could be
The crew was unaware they were carrying HAZMAT and believed they were dealing with an electrical problem. But shutting electrical equipment down wasn't helping and the captain, at least, hypothesized that it wasn't an electrical problem after all, the flight engineer acknowledged.
[NTSB Report, ¶1.1]
- During its return to Boston, the flight was given preferential air traffic control treatment, although it had not declared an emergency.
The crew was asked if they wanted to declare an emergency and declined. There was a real reluctance back then to admit you had a problem you could not handle. Since Boston Center was giving them traffic priority and any assistance asked, this did not have an impact on the events.
[NTSB Report, p. 73 (Cockpit Voice Recorder)]
I/P-3 Pressure altimeters
I/P-1 Twenty-nine seven-five Is okay
I/P-2 Set right
The crew was able to set their altimeters, about this time they spotted an airport on the ground, the confirmed three green landing gear lights, and started looking for the runway.
[NTSB Report, ¶1.1]
- After issuing appropriate descent clearances en route so that fuel could be burned off more rapidly at lower altitudes, at 0926:30 Boston Center advised Boston Arrival Radar (AR-2) that the flight was at 2,000 feet.
[NTSB Report, p. 75 (Cockpit Voice Recorder)]
I/P-1 I don't smell that smoke as much now, there doesn't seem to be as much, does it?
I/P-3 Ah - ah, it doesn't seen to be as much
I/P-3 It doesn't t seem to be as much
[NTSB Report, ¶1.1]
- At 0929, Clipper 160 asked Boston Center for the flight's distance from Boston, and added, "The DME's don't seem to be working." The Center answered, "You're passing abeam, Pease Air Force Base, right now, sir, and you're about 40 to 45 miles to the northwest of Boston."
Most airlines at the time had believed smoke and fumes in the cockpit didn't require an immediate landing at the nearest suitable runway. In hindsight, however, we know that they had several suitable runways that would have allowed a safe landing before the airplane became unflyable.
[NTSB Report, p. 77 (Cockpit Voice Recorder)]
I/P-3 That's worse. I don't see
I/P-1 It's getting worse?
I/P-3 Ah, I turned the, ah, equipment cooler off and that - - - that made it worse
I/P-1 All of a sudden it is getting worse in here
I/P-3 Tell ya what, turn the radar off, the Doppler's off - anything yuh don't need, let's shut em down
We find out (see the Analysis, below) that the flight engineer turned more than just the equipment cooler off.
[NTSB Report, ¶1.1]
- The first communication between Clipper 160 and AR-2 was at 0931:21. The flight was cleared "direct Boston, maintain 2,000." AR-2 asked if the flight was declaring an emergency; the reply was "negative on the emergency, and may we have runway 33 left?" The AR-2 controller approved the request, and the flight proceeded to Boston as cleared. At approximately the same time, the captain instructed the crew to "shut down everything you don't need."
At first I thought either the body of the report or the CVR transcript is wrong. With the former it is the captain who says shut everything down, with the latter it is the flight engineer. But looking at the listed causes, I am wondering if the NTSB bowed to political pressure from Pan American World Airways to paint the pilots in a positive light.
[NTSB Report, ¶2.11]
- At 0931, shortly before the CVR ceased to function, the captain noted that the smoke was suddenly getting worse and advised the crew to "shut down everything you don't need."
The report is trying to paint a picture of an airplane losing electrical systems as a result of some kind of fire while the cockpit visibility is going down to zero. Neither was true, as becomes evident in the Analysis, below. (The CVR stopped because the Flight Engineer took its electrical power away.)
[NTSB Report, p. 81 (Cockpit Voice Recorder)
I/P-3 Okay, landing gear
I/P-1 Three green
I/P-2 Down three green
Even at this late stage, the crew did not have problems identifying the landing gear lights.
[NTSB Report, ¶1.1]
- At 0934:20, AR-2 asked, "Clipper 160, what do you show for a compass heading right now?" Clipper 160 answered, "Compass heading at this time is 205." AR-2 then asked, "will you accept a vector for a visual approach to a 5-mile final for runway 33 left, or do you want to be extended out further?" The crew replied, "Negative, we want to get it on the ground as soon as possible."
- At 0935:46, the AR-2 controller stated, "Clipper 160, advise anytime you have the airport in sight." Clipper 160 did not reply. At 0937:04, the AR-2 controller made the following transmission: "Clipper 160, this is Boston approach control. If you read, squawk ident on any transponder. I see your transponder just became inoperative. Continue inbound now for runway 33 left, you're No. 1. There is a Lufthansa 747 on a 3-mile final for runway 27, the spacing is good. Remain on this frequency, Clipper 160."
- At 0938:31, the AR-2 controller, who was talking to another flight, stated: "...this Clipper has lost his transponder and nobody's working him, and he's been given a clearance to land in the blind. He's just about 4 miles east of Boston now."
- At 0940:23, the AR-2 controller transmitted the following message: "All aircraft on the frequency, the airport is closed at Boston." The AR-2 controller transmitted the message, because ATC personnel had seen Clipper 160 crash. Witnesses saw the left cockpit window open and smoke come through the window. Aeronautically qualified witnesses saw the aircraft approach runway 27 at a faster-than-normal speed and saw it enter roll and yaw maneuvers. These maneuvers increased in severity until the aircraft assumed a final nose-high attitude. The nose-high attitude was followed by an abrupt nosedown attitude, and the left wing and nose contacted the ground simultaneously. The aircraft was nearly vertical at impact.
- The aircraft struck the ground about 262 feet from the right edge of the approach end of runway 33.
Carrying improperly packaged hazardous materials is a recipe for disaster, no doubt about it. Flying for a Part 121 airline should bring with it the necessary protocols to make that safe and the fact that sometimes doesn't happen should chasten those of us operating under "no fly" rules for HAZMAT; these things can kill you. More about this: HAZMAT.
This airplane did not crash because of the smoke and fumes, however. There is ample evidence that the crew was coping with the smoke in the cockpit and were well on their way to a successful emergency landing. What doomed them was poor Crew Resource Management. In particular, it was the flight engineer's actions that impacted the flyability of the aircraft without the captain's knowledge.
Every now and then I come away from reading an accident report wondering about the motivation behind the report. It should be to prevent recurrence. In this case, the focus needed to be on the HAZMAT and it was. So we got recommendations on how to change those procedures. There should also be an examination about the Boeing 707 smoke evacuation procedures and there was. So in both of these cases, the result of the report should be to make flying safer for everyone. But the report went out of its way to say the crew was debilitated by the smoke and just gives the actions of the flight engineer a mention in one sentence in the list of causes. Why? This was the eleventh crash of a Pan American Boeing 707, the ninth which was plausibly caused by pilot error. It could be there was intense political pressure from the well connected CEO of Pan American to downplay the role of the pilots in this one. The FAA did finally act, but it was six months after this crash, in which time there were two more. More about this: CRM Culture.
The hazardous materials
[NTSB Report, ¶2.11]
- There is no evidence to indicate that any member of the flightcrew was aware of the restricted articles on board. It is possible that the cabin cargo areas would have been immediately suspect as a smoke source had the flightcrew been aware of the quantity, nature, and location of the chemicals on board; however, the smoke migration pattern, which caused smoke to emerge from lower 41 compartment would have further confused the crew as to the origin of the smoke and thus would have seriously impeded timely and accurate assessments.
[NTSB Report, ¶1.12 to 1.16]
- The restricted articles were loaded on four pallets which were placed into the aircraft at pallet positions 1, 6, 7, and 9. The various chemicals were not segregated from each other or from other articles being shipped. According to cargo loading personnel, the cargo on pallets 1, 6, 7, and 9 was arranged to provide the required crew access to the hazardous cargo.
- Numerous hazardous materials on Clipper 160 had not been packaged according to regulations.
- The nitric acid bottles were found packed in marked wooden boxes and were cushioned by sawdust. There were no inside metal containers.
The electrical system
[NTSB Report, ¶1.12 to 1.16]
- The aircraft's electrical system and components disclosed no evidence of in-flight fire or preimpact overload or overheat conditions.
[NTSB Report, ¶2.11]
- Since the smoke detector indicators apparently failed to provide an early and positive indication of the source of the smoke, the flightcrew assumed that the smoke in the lower 41 was from an electrical or avionic source. This assumption probably influenced the subsequent actions of the flightcrew more than any other factor.
Problems with Boeing and Pan Am Smoke Evacuation Procedures
[NTSB Report, ¶2.11]
- The Safety Board believes that had an electrical problem in lower 41 actually been the source of the smoke as the flightcrew suspected, the logical decision from a safety and logistic viewpoint would have been to land at the nearest airport where Pan American maintenance personnel and facilities were available to accomplish required maintenance, return airplane to service, and to continue the flight. In this case, the nearest airport with such Pan American facilities was Logan International Airport at Boston.
- Apparently, the problem was underestimated or misunderstood by the crew of Clipper 160. Late during the approach to Boston, conditions in the cockpit rapidly deteriorated. Serious impairment of visibility inside the cockpit and drastic impairment of outside visibility prompted the opening of the cockpit window. Since opening the window was not prohibited, this action taken by the crew is understandable. The procedure was prescribed by Boeing and Pan American at the time of the accident. However, as discovered during smoke evacuation tests after the accident, opening the cockpit window allows even more smoke into the cockpit when the source of the smoke is continuing and originates in the cabin.
- The "electrical smoke and fire emergency procedure" requires that the radios be changed to the No. 2 position before the essential bus is isolated. Since the radio was not changed, only the flight engineer knew what had occurred when the essential bus was isolated. Why the flight engineer did not return the power to the bus could not be determined.
The world of aviation has changed since this crash and very few would advocate overflying viable landing spots with an electrical problem causing smoke and fumes.
[NTSB Report, ¶2.14]
- Initial testimony by the FAA and the Boeing Company indicated that existing procedures for evacuating smoke were adequate if followed to completion. However, data developed during and subsequent to the smoke evacuation tests disclosed that the smoke test conducted during the initial certification of the Boeing 707 did not take into consideration a continuing source of smoke. In view of these findings, the Safety Board believes that the procedures in effect at the time of accident were not effective in controlling or evacuating smoke. On the contrary, it appears that smoke origin and circulation made it virtually impossible to determine accurately the source of the smoke.
The impact of the smoke on the crew's ability to fly
- Recovered oxygen system components disclosed that required oxygen was on board the aircraft; however, the functional capability of the system or the degree to which the system had been used during the emergency could not be determined. The CVR transcript indicates that the flightcrew donned oxygen masks during the emergency. There was no evidence to suggest that the walk-around oxygen bottle had been used.
- The three crewmembers were killed in the crash. Toxicological tests on the deceased disclosed no evidence of carbon monoxide, hydrogen cyanide, alcohol, or drugs.
- The smoke goggles used by Pan American World Airways were examined. The goggles were found to fit loosely around the temporal region of the head, especially if the crewmember is wearing glasses. The goggles were rigid and would not mold readily to facial contours.
There is no doubt a cockpit filled with smoke will impact any crewmember's ability to think but these three were remarkably calm all the way to the end. Given the toxicological tests, I don't think the smoke was the primary cause of this accident.
The Flight Engineer's role
[NTSB Report, ¶2.11]
- The final actions taken by the flight engineer, as prescribed by procedures if smoke continues, include the positioning of the "essential power selector" in the "external power" position. If the selector is positioned to "external power," the yaw damper becomes inoperative.
- The FDR parameters and the CVR disclosed that the wing flaps had been lowered. There is evidence that spoilers had been extended for about 4-1/2 minutes and probably had remained selected at the extended position when the speed was reduced for final approach.
- Performance data for the Boeing 707-321C show that lateral control capability may be extremely limited, if not impossible, with an inoperative yaw damper, extended spoilers, and lowered flaps.
- The evidence suggests that the captain was not aware that the flight engineer's actions had rendered the yaw damper inoperative. In addition, the position of the spoiler control lever may not have been visible through the smoke in the cockpit.
- One of the critical factors in the final accident sequence was the flight engineer's execution of emergency procedures while other crewmembers were not aware of his actions. Various switch settings found on the flight engineer's panel after the crash and information from the CVR indicate that the flight engineer performed the "smoke evacuation emergency procedure" and was in the process of performing the prescribed steps of the "electrical smoke and fire procedures," as prescribed in the Boeing 707 flight manual. The latter procedure requires that the essential bus power switch be placed in the "ground power position," thus removing all power from the systems on the essential bus. Included on the essential bus are: The captain's flight instruments, the No. 1 VHF radio, the cockpit voice recorder, intercom, the yaw damper, and the No. 1 transponder. If these systems are deactivated without the captain's knowledge, the captain may conclude that the smoke problem in the lower 41 compartment had worsened.
- Flight recorder data indicates that a stable approach was never established. The airspeed, altitude, and heading traces fluctuated constantly throughout the approach. Under conditions in which the flight parameters are constantly changing, careful monitoring by the crew is necessary in order to avoid entering a dangerous flight regime. However, since the crew of Clipper 160 could not communicate verbally with each other and probably could not see the instruments because of dense smoke, they could not monitor airspeed and altitude during the final phase of the approach. This could easily lead to a stall or an uncontrollable maneuver at an altitude too low for recovery. Heading excursions during the final moments of flight also indicate that the crew may have had difficulty seeing the runway because of the dense smoke in the cockpit.
- According to FDR traces, the airspeed deteriorated from about 160 to 122 kn. during the last portion of the flight. Stall speed for the aircraft's configuration at the time of the accident was 118 kn. in wings level, unaccelerated flight. Since the FDR indicates a continuous heading change, the aircraft must have been in a bank or a yaw. If the aircraft stalled during such a maneuver, considerable altitude would have been required to recover safely.
The impact of the "external power" position was much more than just the yaw damper, but just losing the yaw damper alone made landing this airplane very difficult.
The report does not go into lateral control capability much deeper than this sentence, but it bears much more examination. The Boeing 707, without the yaw dampers, tends to Dutch roll. The aircraft becomes a handful, even under the calmest weather conditions. Unless you had trained to fly the airplane without the yaw dampers, your chances of landing safely without the yaw damper are greatly reduced. I've never flown an airplane with worse Dutch roll than the Air Force Boeing 707 (known as the C-135B and other variants) which did not have yaw dampers. The Air Force KC-135A was almost as bad. More about this: Stability and Control.
Eight of the ten Pan American Boeing 707 accidents that preceded this one had Crew Resource Management issues that involved an overbearing captain and perhaps an obsequious crew as the problem. The culture at Pan Am had not yet changed, but the make up of the crews were starting to. The captain on Pan Am 160 was 53 and had been with the company since 1951. The first officer was 34 and was hired in 1966. The flight engineer was 37 and was hired in 1967. A common survival technique with a very senior captain and younger cockpit crewmembers is to adopt the path of least resistance. For the first officer this means doing things the captain's way and only acting independently if given no other choice. A flight engineer has much more latitude, given that much of his or her responsibilities are outside the captain's field of view. There is no mention of the flight engineer coordinating these checklists or the changing of the essential bus power switch on the cockpit voice recorder.
[NTSB Report, ¶1.1]
- Aeronautically qualified witnesses saw the aircraft approach runway 27 at a faster-than-normal speed and saw it enter roll and yaw maneuvers. These maneuvers increased in severity until the aircraft assumed a final nose-high attitude. The nose-high attitude was followed by an abrupt nosedown attitude, and the left wing and nose contacted the ground simultaneously. The aircraft was nearly vertical at impact.
It appears that the person who wrote §2 of the NTSB report didn't talk to the person who wrote §1. The report's analysis discounts the high speed indicated by the recovered airspeed indicator, saying it was probably an error and that the aircraft may have stalled. The report's narrative cites "aeronautically qualified witnesses" that point to a pronounced Dutch roll. The Analysis also mentions "continuous heading change" at this point; in a Boeing 707 that also leads me to believe the crash was caused by a pronounced Dutch roll.
I would rewrite the findings as follows:
- Boeing failed to adequately consider smoke evacuation procedures for more than just momentary episodes of cabin smoke.
- Pan American World Airways failed to ensure proper packaging of hazardous materials and failed to inform the crew of their inclusion on the aircraft's loading manifest.
- Pan American World Airways failed to adequately train its crews on the requirement for all aircraft configuration and systems changes that impact safety of flight be fully coordinated with each cockpit crewmember.
- The flight engineer failed to coordinate essential electrical system changes with the captain and first officer.
[NTSB Report, ¶2.2(b)]
- The National Transportation Safety Board determines that the probable cause of the accident was the presence of smoke in the cockpit which was continuously generated and uncontrollable. The smoke led to an emergency situation that culminated in loss of control of the aircraft during final approach, when the crew in uncoordinated action deactivated the yaw damper in conjunction with incompatible positioning of flight spoilers and wing flaps.
- The Safety Board further determines that the dense smoke in the cockpit seriously impaired the flightcrew's vision and ability to function effectively during the emergency. Although the source of the smoke could not be established conclusively, the Safety Board believes that the spontaneous chemical reaction between leaking nitric acid, improperly packaged and stowed, and the improper sawdust packing surrounding the acid's package initiated the accident sequence.
- A contributing factor was the general lack of compliance with existing regulations governing the transportation of hazardous materials which resulted from the complexity of the regulations, the industry-wide lack of familiarity with the regulations at the working level, the overlapping jurisdictions, and the inadequacy of government surveillance.
NTSB Aircraft Accident Report, AAR-74-16, Pan American World Airways, Inc., Boeing 707-321C, N458PA, Boston, Massachusetts, November 3, 1973.