The Crash Detectives

Investigating the World's Most Mysterious Air Disasters

Christine Negroni

• Publisher: Atlantic Books
• Year: 2016
• ISBN:
  9781782396420
• Topic: Technology & Engineering
• Subtopic: Transportation & Navigation
• Index: Technology & Engineering

In The Crash Detectives, veteran aviation journalist and air safety investigator Christine Negroni takes the reader inside crash investigations from the early days of the jet age to the present, including the search for answers about what happened to the missing Malaysia Airlines Flight 370.

As Negroni dissects each accident, she explores the common themes and, most importantly, what has been learned from them to make planes safer. Indeed, as Negroni shows, virtually every aspect of modern pilot training, airline operation and aircraft design has been shaped by lessons learned from disaster. Along the way, she also details some miraculous saves, when quick-thinking pilots averted catastrophe and kept hundreds of people alive.

Tying in aviation science, performance psychology and extensive interviews with pilots, engineers, human factors specialists, crash survivors and others involved in accidents all over the world, The Crash Detectives is an alternately terrifying and inspiring book that might just cure your fear of flying, and will definitely make you a more informed passenger.

PART ONE

Mystery

I have approximate answers and possible beliefs and different degrees of uncertainty about different things.

— NOBEL PRIZE–WINNING THEORETICAL
PHYSICIST RICHARD FEYNMAN

The Clipper

On the last leg of a journey halfway around the world, Pan American Airways captain Leo Terletsky began to worry. And when Captain Terletsky worried, everybody else on the flight deck worried, too. “His anxiety caused him to shout at copilots, issue orders and immediately countermand them. He infected his crews with his own anxiety,” wrote Horace Brock, who flew a few times with Terletsky and didn’t much like it.

At noon on July 29, 1938, there was plenty to be anxious about. Having spent fifty-six hours over five days flying Pan Am passengers from San Francisco to the Far East, Terletsky and his nine-man crew were deep into bad weather as the Martin 130 flying boat approached the Philippine archipelago.

The plane was “sandwiched between two layers of clouds,” explained Pete Frey, a captain with a large American carrier and a safety investigator with his union, who reviewed for me the weather reports submitted by the crew on that miserable summer morning. The stratocumulus clouds that Terletsky encountered are often at the beginning or end of worse weather, including rain and turbulence. Terletsky was dealing with both as he threaded the four-engine plane through the bands of clouds above and below where he was flying at 9,100 feet, 586 miles east of Manila. As Frey explained it to me, the rocky ride was not the crew’s biggest problem.

“They are inside the clouds half the time, flying on instruments. This would make navigating by observing landmarks below impossible. It would also make getting a fix from the sun or other celestial object impossible as well.

“They are navigating with dead reckoning,” Frey said. Dead reckoning is the most basic form of navigation: essentially a mathematical calculation involving weather, wind, time, speed, and direction. “You make an assumption of wind correction and simply hold a heading and course for a fixed period of time. At the end, you hope to be where you planned,” Frey explained. Yet considering their inability to see the earth below, the crew would have had little on which to base their position; or as Frey imagined flying under the circumstances, “You are lost.”

Around noon local time, radio operator William McCarty, thirty-three, sat at his desk behind the copilot, tapping the keys of his Morse code machine. He was sending a message to the Pan Am ground station on the Philippine island of Panay. Even if the crew was uncertain of its position, Pan Am’s ground personnel would try to use radio wave direction finding to pinpoint the flying boat’s location. They could also provide the crew with information about the weather ahead.

McCarty reported the weather and the winds, the temperature, and the crew’s approximation of where they were, along with their speed. Morse code could get through even when the plane’s radio signal was not strong enough to transmit a voice. By the time McCarty was done, about ten minutes had passed, and Edouard Fernandez, the operator at Panay, wanted to pass the weather news on to the crew. McCarty asked him to wait. “Stand by for one minute before sending as I am having trouble with rain static.” When Fernandez tried later to contact the Clipper, there was no reply.

Nothing was ever again heard from the Hawaii Clipper. No piece of the plane, no human remains, no luggage or cargo, and no airplane fluid or fuel would show up. As with Malaysia 370 seventy-six years later, only the evidence still on the ground would be available for investigators to consider. They could scour the maintenance records and operational history of the plane and review the performance and training of the crew along with the information sent by McCarty during the flight, but it might not be enough to determine conclusively what happened. It could be illuminating; it might be baffling. It turned out to be both.

Thin Air

No one knows for sure what happened aboard Malaysia 370. The scenario I am about to describe is based on a framework of events put forward by Malaysian and Australian investigators and other sources who participated in gathering or analyzing the known data. To this I have applied Occam’s razor, the principle that suggests that if there are many possible explanations for something, the simplest is the most likely.

Shortly after midnight on March 8, 2014, and seemingly without warning, what had been an entirely normal flight devolved into an illogical series of events. That kind of wacky has been seen before when pilots are afflicted with altitude sickness, known as hypoxia.

An inability to get enough oxygen into the lungs to sustain cogent thought happens when planes lose pressurization, and that can happen for a variety of reasons. It can be triggered by an electrical problem or some mechanical difficulty. Pilots sometimes fail to turn the pressurization on at the beginning of the flight, but even when the pressurization is working as it should, there’s no way to keep a plane pressurized if there is a hole in the fuselage or if leaks at the seals of doorways, windows, or drains from the galley and bathrooms allow the denser air to escape.

If the pilots on Malaysia Flight 370 experienced oxygen deprivation because something happened to cause the plane to lose pressurization, they would have behaved irrationally, perhaps turning a moderate problem into a catastrophic one. The passengers and crew would have become feebleminded and helpless.

At the time of the MH-370 disaster, people were boarding airplanes around the world at a rate of eight million a day. Few air travelers then (or now) gave a thought to the fact that outside those aluminum walls the air is too thin to sustain coherent thought for more than a few seconds. Life itself is extinguished in minutes. While the percentage of oxygen in the air (21 percent) is the same as on the ground, the volume of air expands at higher altitudes. We rely on air density for the pressure that drives oxygen into our bodies. Miles above the earth and absent this pressure, oxygen will rush out like air racing out of a balloon.

What keeps us air travelers alive and, for the most part, in our right minds is a relatively simple process that pumps air into the plane as it ascends, like air filling a bicycle tire. The air comes off the engines and is distributed via ducts throughout the plane. In most airliners, the cabin pressure is set to mimic the pressure density of about eight thousand feet. So to your body, flying is like being in Aspen, Colorado, or Addis Ababa, Ethiopia.

When it is time to land, the valves that closed on takeoff to maintain that air density in the cabin begin to open, allowing it to escape gradually until the inside of the plane is equalized with the outside, or generally, to 190 feet above the altitude of the airport. You’ll know this process is happening when your ears start to pop in the last twenty to thirty minutes of your flight. If this extra pressure weren’t allowed to vent, the door of the plane might explode outward. It happened as recently as 2000, when an American Airlines Airbus A300 made an emergency landing at Miami International. Insulation blankets blocked the outflow valves, so the differential pressure inside the cabin was still high even after landing. It is not clear if the flight attendants realized it, because they had other problems. A smoke alarm had triggered, and they were worried about a fire. So they were trying to evacuate the plane, but the doors would not budge. Finally, thirty-four-year-old senior flight attendant José Chiu pushed hard enough, and the door blew out. Chiu was jettisoned off the airplane and killed.

On most flights the automated system works as designed. Still, at least forty to fifty times a year, an airliner somewhere in the world will encounter a rapid decompression, according to a study for the Aviation Medical Society of Australia and New Zealand. James Stabile Jr., whose company, Aeronautical Data Systems, provides oxygen-related technology, said that when slow depressurizations are figured in, the rate increases even more. And because not all events require that regulators be notified, the problem is “grossly underreported.”

When planes fail to pressurize after takeoff or lose cabin altitude in flight, it is potentially life-threatening. The reason we don’t see tragedies more often is because pilots are taught what to do. First, they put on their emergency oxygen masks. Then they verify that the system is on. There are numerous cases of pilots discovering that they failed to set cabin altitude upon takeoff, which I liken to finding the laundry I loaded in the washer unwashed hours later because I forgot to start the machine.

If pressurization was set correctly and is still not working, pilots immediately begin a rapid descent to an altitude where supplemental oxygen is not necessary. When pilots do not follow these steps, the situation spins out of control quickly.

To be clear, pilots don’t intentionally ignore the procedures. When they do, it is usually because their mental processes are already compromised by oxygen starvation. Sometimes the effect is unfathomable; pilots faced with an alert that the cabin altitude is exceeding twelve thousand feet have been known to mistakenly open the outflow valves, completely depressurizing the cabin and ratcheting up the problem.

On an American Trans Air flight in 1996, a mind-boggling sequence of events brought a Boeing 727 a hairbreadth from catastrophe. The miracle is that despite the lunacy in the cockpit, the plane landed safely.

ATA Flight 406 departed Chicago’s Midway Airport bound for St. Petersburg, Florida. At thirty-three thousand feet, a warning horn sounded because the altitude in the cabin was registering fourteen thousand feet. First Officer Kerry Green was flying. He immediately put on his emergency oxygen mask. Capt. Millard Doyle did not, opting to try to diagnose the problem. He instructed the flight engineer, Timothy Feiring, who was sitting behind and to his right, to silence the alarm. Doubtless already feeling the effects of altitude that was steadily increasing, Feiring could not find the control button, and more time passed.

As he looked around, the captain evidently thought he’d discovered the source of the problem, an air-conditioning pack switch that was off, and he pointed it out to Feiring. Then he turned his attention to the flight attendant in the cockpit, asking her if the passenger oxygen masks had dropped.

They had, she replied, and promptly collapsed in the doorway. Now Captain Doyle reached for his own mask, but it was too late. Disoriented and uncoordinated, he could not place it over his head, and he passed out, too.

Two of four people in the cockpit were now incapacitated, and Feiring was having trouble thinking. He mistakenly opened an outflow valve, creating a rapid and total decompression of the airplane.

He put on his mask and then got up to attend to the unconscious flight attendant, placing the flight observer mask on her face, but dislodging his own in the process. He passed out, falling over the center console between the two pilots’ seats.

Through all this, First Officer Green, with his mask on, was taking the plane down to a lower altitude at a speed of about four to five thousand feet per minute.

Back in the passenger seats, the cabin crew had not been given any instructions from the cockpit, but the flight attendant seated at the front of the plane made a pantomime with her mask to demonstrate what the passengers should do. Some travelers followed her example; others did not. Through it all, the flight attendants reported that the plane was pitching up and down and side to side, and there was a brief, incomprehensible announcement from the cockpit.

Passenger Stephen Murphy of San Diego thought he was going to die and remembers feeling a sense of peace as he recited his prayers. Then the woman seated behind him started having convulsions, and the man across the aisle began to claw at his ears.

“What bothered me was there was nothing I could do for him. It’s not like you see on TV; people don’t grab portable oxygen bottles and walk around the cabin helping people,” Murphy told me years later. “Had I had my full senses, I’d like to think I could have helped somebody. But based on what was going on, I didn’t. I knew I couldn’t.”

On the flight deck, Green was trembling, a common symptom of hypoxia. Something was wrong with the microphone in his mask, and he had to pry the compressible seal away from his face to contact air traffic control.

When the oxygen mask Feiring had placed on her face rejuvenated the flight attendant, she got up and returned the favor, replacing the mask that had come off him as he moved away from the flight engineer console. She also got a mask on Captain Doyle. Soon they both came to. American Trans Air Flight 406 landed safely in Indianapolis, but the episode could have ended in catastrophe.

The story, equal parts chilling and absurd, tells me that knowing what to do does not mean pilots will actually do it if their ability to think has begun to deteriorate.

Nine years after American Trans Air 406, on August 14, 2005, a Boeing 737 took off from Cyprus on a flight to Athens, but it never arrived. Helios Flight 522 ran out of fuel and crashed into a mountain south of the airport after flying on autopilot for more than two hours—long after the pilots and nearly everyone else on board had fallen into deep and prolonged unconsciousness. They had been starved of oxygen, presumably because the pilots failed to pressurize the aircraft after takeoff. The pilots were hypoxic before they realized what had gone wrong.

The Helios 522 disaster started about five and a half minutes after takeoff, as the plane climbed through twelve thousand feet. A warning horn alerted the pilots that the altitude in the cabin had exceeded ten thousand feet.

Less than two minutes later, the passenger oxygen masks dropped, but Capt. Hans-Jürgen Merten and First Officer Pambos Charalambous did not put on their masks, deciding instead to try to figure out what was wrong: a classic case of impaired judgment due to hypoxia.

For nearly eight minutes, Captain Merten, a pilot with five thousand hours of experience on the 737, conversed with the Helios operations center in Cyprus in an exchange that grew increasingly confusing to the men on the ground. One thing was certain. The horn warning of altitude did not direct the pilots to focus on the cabin altitude, and here’s why: the alarm’s insistent staccato is also used on the runway when an airplane is incorrectly set for takeoff. At that time in the flight, the same alarm is called a takeoff configuration warning. This case of one alarm for two hazards relies on the pilots’ knowing to which hazard they are being alerted.

On the ground, it seems straightforward. The takeoff configuration alarm will sound only prior to takeoff. The distinction is not so obvious, however, when the pilot’s ability to think is already fading. And we know this because, when the alarm on Helios 522 went off, Merten told his airline’s operations desk that the takeoff configuration horn was sounding. He did not associate the warning with cabin altitude. That mistake has been repeated on passenger flights around the world, including ten instances over ten years found in the files of the NASA Aviation Safety Reporting System, or ASARS.

“The simplicity of the error” is what struck Bob Benzon, an accident investigator with the National Transportation Safety Board at the time, who was helping the Greeks on the Helios accident. “There were one hundred twenty-one people who died on a modern airliner, and all through a simple error. That was the thing,” he said.

Six years earlier Benzon had been assigned to investigate a similar accident, involving a private jet and a popular American athlete. Payne Stewart was one of the most famous golfers on the pro circuit, beloved for the weird collection of tam-o’-shanters and knickerbockers he wore at tournaments. He suffered hypoxia on October 25, 1999, in the early stages of a four-hour flight...