Grisly precision

Inside the strange world of the NTSB.

 

- - - - - - - - - - - -

By Phaedra Hise

salon.com

Nov. 18, 1999

 

A plane crash, whether a large commercial airliner or a tiny home-built ultralight, sets into motion a flurry of events that ends with a National Transportation Safety Board investigator pacing around the wreckage. The men and women of the NTSB have a rare breed of government job. They find the cause, often beginning with little more than a handful of crushed aluminum, of almost every aviation crash they investigate.

 

If the Federal Bureau of Investigation takes over the EgyptAir Flight 990 investigation later this week, it won't mean much of a change in the daily job of NTSB investigators. The change is mainly a matter of politics -- the FBI will decide which leads to pursue, while NTSB experts remain on the case, analyzing clues and submitting a standard final report. The FBI has followed the investigation closely, which it does with any crash that starts to look fishy. The NTSB normally gives up control of an investigation in one of two situations: When an American-based plane crashes in a foreign country, or in cases of suspected criminal activity.

 

The NTSB is a uniquely independent organization. It was formed as an off-shoot of the Civil Aeronautics Board. The CAB had a crash investigation arm, but it was smeared with a little conflict of interest since the agency also wrote aviation regulations. Conceivably, the people who write the rules won't have much incentive to question whether or not those rules are working in the wake of an accident. So in 1967, the NTSB was created and charged with investigating every single civil air crash, as well as major railway, marine, highway and pipeline accidents.

 

Everyone else in aviation has an ax to grind. Manufacturers want to avoid liability issues, the FAA wants to keep its procedures looking good, airports want the airlines to be happy, and the airlines want to keep operations running on time at low costs. The NTSB was sent into the game for the same reason that baseball appointed a commissioner: to put an impartial arbiter in control. It's a completely independent agency, paid for by tax dollars (under 20 cents per person, by its estimate), but which answers to no branch of government. To further protect the agency, none of the NTSB report material can be cited in legal proceedings.

 

So what happens after a crash?

 

Someone, either a local police officer or an FAA controller, reports the incident to one of the six regional or four field NTSB offices. The investigator "on call" gets paged, and rushes to the crash site. If the crash is major, like the recent EgyptAir Flight 990 disaster, then a duty officer in Washington gets a simultaneous call, and usually reassigns the case to someone more senior than the on-call investigator. A big crash, if eventually solved, can make an investigator's career.

 

Because there are fewer than 60 regional aviation investigators working 2,000 events a year, small teams lead each case and pull in outside parties. The FAA, and engine and airframe manufacturers normally show up for small accidents, but in major investigations the parties multiply. A catastrophic and confusing crash like EgyptAir relies on teams of experts including instrument manufacturers, the Coast Guard, psychological analysts, airline management and overseas government officials.

 

The parties that work with the NTSB provide mounds of data for investigators to sift through. Almost everything a pilot does is documented by some federal agency. His doctors file records of his annual physical exams; his instructors keep data on his training and flight testing; the FAA records his voice when he radios to Air Traffic Control. The airplane is tightly watched also -- the FAA has radar maps of its position and historical data on all maintenance and owners since the plane rolled off the assembly line. The NTSB talks to witnesses, friends and relations, experts.

 

Recovering the flight data and voice recorder boxes is a major coup in resolving any crash, but normally the data is used more to verify a hunch than to solve the mystery. That's one reason the EgyptAir 990 crash was initially so frustrating. Experienced investigators are usually able to draw pretty accurate conclusions based on initial evidence. With EgyptAir, however, none of the early clues added up.

 

Crashes on land initially yield more material to work with. Investigators say they get used to the distinctive crash stench of aviation fuel mixed with charred seat foam. They're busy studying the sheared tree tops, the broken limbs revealing the plane's last trajectory. Or they get wrapped up in analyzing the wreckage -- the bent propeller indicating that the engine kept running down to the ground, the frozen instruments revealing the plane's final air speed and angle. They spend a few days at the site, ordering toxicology reports on the pilot and testing the control cables.

 

An ocean crash is slower, more agonizing. The wreckage may never even turn up. Investigators begin by commandeering radar reports and medical records. They float out on Coast Guard boats to grapple for chunks of luggage floating in 5-foot ocean swells. They calculate survival rates and ponder the gray print outs from side-scan radar.

 

What kind of person digs this job?

 

A pilot, obviously, one who likes to work puzzles. Flight experience is the first requirement for hire, and it's helpful to have a background in the hard sciences. And it's important not to be wigged by the occasional gristly bits hanging in trees. One NTSB manager who spent 18 years leading investigations said, "It wasn't a burnout-type job for me, but it bothers some people." One former investigator, who worked crashes for five years, says the turnover rate has recently increased, especially among the entry-level investigators analyzing the more frequent small plane wrecks.

 

Taking a year to file the final report is typical. On average, NTSB officials say, reports are finalized within 13 months, although the agency's online database reveals pages of non-finalized crash reports several years old.

 

The report summarizes the crash's probable causes and any resulting recommendations for change. Based on NTSB recommendations, the Federal Aviation Administration mandates all kinds of plane and flight operations requirements. Recently these have included anti-collision warning alarms, improved de-icing systems, better rudder design (for the Boeing 737) and design changes to those pesky thrust reversers (on the Boeing 767).

 

The FAA takes some heat for not responding promptly. Pilots complain that it's a "tombstone agency," one that adopts change only after people die in plane crashes. For example, the NTSB long advocated putting smoke detectors in aircraft cargo holds, but it was the ValuJet crash that made the FAA sit up and seriously consider the proposal. Currently, the NTSB wants all infants to fly in approved safety seats. But the FAA sits in an awkward position -- the agency is charged with both promoting and policing aviation. Since the airlines don't want child seats, the FAA isn't likely to approve them. Until more infants die in crashes, that is, and the NTSB files more reports.