Tuning Up the 

ALPA teams up with the University of North Dakota to launch an 
Advanced Accident Investigation Course for aviation accident investigators.

By Jan W. Steenblik, Technical Editor
Air Line Pilot, September 2003, p.26

Capt. John O’Connor (Atlantic Coast) was puzzled. He and the other members of the Structures Group had roamed the wreckage site, giving the scene a quick lookover, then searching for what First Officer Conny Kleissas (US Airways, Furl.) called "the four corners of the airplane."

The wings of the Navion Rangemaster—a four-seat, low-wing piston single—were lying in plain view on the grass next to the University of North Dakota ramp at Grand Forks Airport. So were the crushed fuselage, lying on its side, and the twisted tail section, and the front pair of seats.

But where were the elevator balance weights? Capt. O’Connor and his team had found one inside the rectangle of yellow plastic tape that marked the staked-out wreckage area. The other wasn’t—suddenly Capt. O’Connor noticed that something was lying in the high grass a few yards beyond the tape. The missing balance weight!

"Just like in real life," F/O Kleissas grinned. "Never assume that the ‘crime scene’ tape—which the local law enforcement officers will put up before you arrive at the accident site—contains all the wreckage."

"Thinking outside the box" is a tiresome cliché, but Capt. O’Connor had just gotten a hands-on lesson in accident investigation in the truth at the core of the cliché. So simple, so basic —and so important.

Capt. O’Connor was a member of the first class of pilots to go through the ALPA Advanced Accident Investigation Course (AAC), held June 3-5 in Grand Forks, N.D., with the help and cooperation of the University of North Dakota’s internationally renowned aviation school, the John D. Odegard School of Aerospace Sciences. Two years in the making, the AAC class was designed to give ALPA accident investigators—and key UND Aerospace faculty—the kind of education that only a well-designed mock accident investigation can provide.

For the 19 ALPA pilots in the first course, a prerequisite was attending ALPA’s Accident Investigation Course, which the Association presents twice each year. Some of the 19 already had participated in NTSB investigations of airline incidents and accidents. Most of the 19 were the chief accident investigator for their pilot group’s master executive council.

The day before the course began, the course instructors—all members of ALPA’s Accident Investigation Board (AIB)—plus Dana Siewert, director of aviation safety for UND Aerospace, and Frank Argenziano, special projects coordinator and facilities manager for the school, placed the wreckage on the field near the UND ramp and put up the yellow police tape.

Capt. Todd Gunther (Comair), chairman of the ALPA AIB, explained, "By participating in the NTSB investigation of the Emery Flight 17 accident at Sacramento, I met the guy who was storing the wreckage of that airplane. I told him we needed an airplane for our advanced course, and he donated the Rangemaster to UND."

During the mock Safety Board "organizational meeting" the next morning, Capt. Lindsay Fenwick (Northwest), chairman of the ALPA Accident Analysis Group, informed the students that he would be playing the role of the NTSB investigator-in-charge (IIC). First Officer Robert Slowik (Northwest), ALPA’s training plans coordinator, would play the FAA representative.

Six student groups, each made up of four students, would rotate through six typical NTSB groups—structures, survival factors, operations, maintenance records, cockpit voice recorder, and air traffic control. On a real NTSB investigation, a representative of a participating "party," such as ALPA, usually would be assigned to one group for the duration of the investigation.

Each of the students was required to sign a copy of the NTSB confidentiality agreement that party representatives must sign at the beginning of an actual investigation. With that, they were out the door and on to their first group exercises.

Structures and Survival Factors

F/O Kleissas told the first Structures Group that its work would be "straightforward wreckage documentation—simple, but critically important." He stressed that members of the Structures Group, in particular, need to take meticulous field notes because, after the field phase of the investigation is over and the wreckage is removed from the accident site, they cannot go back to recreate the accident site. "If your observations do not get onto paper," he warned, "they don’t count."

Another warning: "We never want to be looking through the wreckage by ourselves," F/O Kleissas pointed out. "We want to be completely above reproach. For example, we want to never be accused of entering a cockpit and lowering the [landing] gear handle.

"Never, ever pick up pieces and try to match them up. This is like an archaeological site—everything exists in situ, in a context."

As First Officer Ron Bach (United), Capt. O’Connor, and F/O Bill De Groh (American Eagle) hiked to the wreckage, F/O Kleissas continued, "Let’s go for the incredibly obvious first. The world is fraught with possibility right now; keep an open mind, and never fall in love with a theory.

"We’ll be looking for, among other things, fire patterns, soot, ground scars, crush patterns, the condition of leading edges. Locate and identify the control surfaces, the control systems, the landing gear position. Note the forces—tension, compression, shear, and torsion—that bent the metal."

The students opted to use a stake line with distance measurements from the stake line to plot the wreckage, rather than setting up a grid pattern. As they mapped the scattered pieces of the shattered airplane, they pored through a parts catalog for the airplane to positively identify the pieces.

Meanwhile, Capt. Gunther, who has several major airline accident investigations under his belt, was leading his quartet of students—the Survival Factors Group—around the wreckage site. F/O Slowik showed up in his role as the FAA representative, "discovering" prescription medications and alcoholic beverages in the back pocket of one seat.

Two of the survival factors students were swaddled in the white "bunny suits" and yellow boots that must be worn in a known or suspected biohazard area.

Capt. Gunther shared his considerable knowledge of a grim subject—how seats and seat tracks reveal impact forces, how to tell if a seatbelt was fastened during the accident, how to tell who was handling the controls at time of impact.

Ops Group

Capt. Paul Brady (American Eagle), another AIB member, led the simulated Operations Group.

"We’re trying to ascertain how the airplane got from California to Grand Forks Airport," he began. "We’re going to try to answer the questions in ‘The Red Book’ [the ALPA Accident Investigation Handbook,] regarding operations. The NTSB uses a checklist for ops; you should, too."

They began with crew histories—duty time, medical history, flight time, employment.

"We found alcohol and prescription drugs in the airplane," offered Capt. Victor Cabot (American Eagle), having already been through the Survival Factors Group session in the morning. To determine the possible effects the medications might have had on the pilot, the pilots called Micah Parkinson, a UND aviation student serving as a "gofer" for the course and, at the moment, impersonating a physician on the other end of the telephone line.

The class participants discussed what they knew so far about the airplane’s oxygen system (oxygen bottle found, fully charged—two nasal cannulae found, one still in package), talked about the pilot’s communications with ATC and compliance with regulations, and reviewed questions answered and unanswered about the weather during the flight and the pilot’s flight planning.

"We need to contact the ATC facility and the Flight Service Station for radar data, PIREPs, and turbulence reports," Capt. Brady pointed out. "We need to check into possible induction system icing, possible VFR into IMC, thunderstorm penetration."

Witness interviews, he said, are "the meat of the Ops Group." The Ops Group’s witness list would include not only the pilot’s wife, for whom they developed a list of questions, but also the pilot’s flight instructor and his mechanic.

By coincidence, in real life, Capt. Brady knew the mechanic who had worked on the airplane. He was thus able to play the role of the mechanic in a mock interview.

Some of the Ops Group’s work inevitably overlapped that of the other groups—for example, no carburetor heat was installed on this airplane. "We need to follow up with the Maintenance Records Group," Capt. Brady acknowledged. Regarding weight and balance, had the Structures Group found any heavy object(s) that the pilot might have been transporting for his business?

The questions multiplied, divided, grew branches; some answers came readily, but not as many as questions, which spread across the chalkboard and down the pilots’ notepads—just like the first days of an actual accident investigation.

Maintenance Records

Capt. Dan Sicchio (US Airways), another member of ALPA’s AIB, welcomed his fellow pilots to the Maintenance Records class.

"Obviously, this is a general aviation accident," Capt. Sicchio acknowledged, "but many of the same processes in tracking down maintenance records and practices apply to an investigation of an airline accident or incident.

"We have three maintenance logs—engine, airframe, and prop—and one computer today. We’ll be looking for discrepancies in the logbooks and using the computer to search for airworthiness directives (ADs) and service difficult reports (SDRs). With computers, in an airline accident investigation, we’re not just limited to the company’s records."

Capt. Dave Miller (Spirit) spoke up. "I’m an A&P and an AI, so I’d like to start with the logbooks."

Capt. Sicchio said, "If you can put somebody with an A&P on the NTSB’s Maintenance Records Group, all the better."

Capt. Jack Wilkes (Alaska) recalled that, during the NTSB investigation of the crash of Alaska Airlines Flight 261 off the California coast in January 2001, ALPA put a line pilot who is also a physician on the Maintenance Records Group. "He turned out to be absolutely the best guy we could have put on that group," Capt. Wilkes asserted.

"You’ll find doctors, engineers, people from many different backgrounds" on the NTSB Maintenance Records Group, Capt. Sicchio agreed.

By the end of the session, Capt. Miller reported a number of discrepancies in the logbooks—"dates don’t match, events don’t match"—and Capts. Wilkes and Louis Dashiell (Atlantic Coast) had used the computer to dig up SDRs and ADs pertaining to the airplane type.

On the fine art of interviewing people who might be able to contribute useful information to an investigation, Capt. Sicchio said, "That’s when you do your best work. You don’t stop until you run out of ‘whys.’ The idea is not to act like Perry Mason or to make an effort to prove a point; instead, put the interviewee at ease. If you let him tell his story without fear of being hammered, you’ll get what you’re looking for."

In an airline accident investigation, Capt. Sicchio added, "when multiple mechanics or technicians are involved, it might be best to start with the maintenance supervisor—in part, to try to determine what role the corporation played in the whole story. For example, is the company pushing the mechanics to get the work done, signed off, and out the door?"

CVR Group

Capt. Pete Frey (Delta), a member of ALPA’s AIB, conducted the CVR Group class. Because the airplane actually involved in the accident scenario recreated for the course was not an airliner, it did not carry a CVR. Therefore ALPA provided CVR tapes from an airline incident and an airline accident involving carriers no longer operating.

Capt. Frey laid out the rules: No writing or recording materials, including laptop computers, would be allowed on the table, except for the legal pads and pencils provided by the Safety Board. The members of the CVR Group, he added, must not talk about the CVR recording outside the room.

Corey Stephens, a staff engineer in the Engineering and Accident Investigation unit of ALPA’s Engineering and Air Safety Department, added that NTSB staff members "will tell you that you can’t talk to anybody about the CVR recording. However, this does not mean that you cannot talk about it with the other ALPA representatives participating in the investigation."

The pilots donned headsets to listen to the first audio demo—the CVR recording from the crash of Air Florida ("Palm") Flight 90 shortly after takeoff from Washington National Airport on Jan. 13, 1982. Despite the cockpit’s having a good-quality cockpit area mike, a fair amount of static crackled in the background, and the ATC chatter was not very clear.

The second audio demo involved a British VC-10 that suffered an engine fire shortly after taking off from John F. Kennedy International Airport in New York. The hot mikes used in that cockpit produced a wonderfully clear recording, including the sound of the captain’s increased breathing rate during the emergency. The principal exercise for the group was transcribing the VC-10 CVR tape, discussing the words and sounds as they went along.

The value of having a current line pilot on the NTSB CVR Group was exemplified by a discussion of Morse code heard on the tape. Capt. Wilkes suggested that the recording showed that the first officer tuned in the ILS audible identifier, entered it on the captain’s side, and shortly thereafter tuned in the VOR audible identifier on his own side. The two signals overlapped, one louder than the other, one coming into each student’s left earphone, the other into his right. Stereo and multiple channels helped the ALPA representatives recreate the sequence of actions in the VC-10 cockpit.

Stephens also emphasized the importance of ALPA’s Critical Incident Response Program (CIRP) in helping accident investigators deal with the emotional trauma of their work.

"We’ve seen three types of outcomes in pilots who have participated in accident investigations—(1) the pilot has come through it with CIRP support and been okay, (2) the pilot has come through it without CIRP support, but quit ALPA safety work, and (3) a couple of instances when, without CIRP support, the pilot has quit flying."

ATC Group

Playing his role as the NTSB group chairman for the ATC Group, Capt. Ray Gelinas (Air Canada Jazz), another AIB member, was indignant that ALPA tried to replace their representatives from the day before. He also said that ALPA representatives were seen talking to a CNN reporter, thus disqualifying themselves from participating further in the investigation.

Capt. Gorden Burgess (Northwest) admitted to talking to the reporter, but swore he told the reporter only that he couldn’t talk to him. After discussions with the IIC, Capt. Gelinas let Capt. Burgess stay.

UND ATC instructor Chad Didier played the role of the Grand Forks Approach Control supervisor who was on duty at the time of the accident. Group 3 interviewed him regarding ATC procedures, the local facility’s equipment, and whether he heard an ELT transmit after the crash.

Next up for an interview by the ATC Group was Jason Sileo, a UND ATC student acting as the Grand Forks approach controller. At one point, Sileo said he’d been advised by his attorney to not answer questions.

UND student Shelbee Sanito played the tower controller in the next interview. Capt. Gelinas pointed out that it would be "very unusual" for a controller to be sitting in a real NTSB interview without an attorney from the National Air Traffic Controllers Association (NATCA) at his or her side.

In the interviews, Capt. Gelinas used the 39 items on the NTSB ATC Group Chairman’s checklist, which can also be found in "The Red Book."

After the mock interviews, the pilots participated in the exercise of checking the time line from the draft ATC transcript against stopwatches and the ATC recording. On a second pass, they checked the draft transcript thoroughly for accuracy. Finally, to complete the ATC Group work for the investigation, they worked together on Group notes to be given to the IIC.


On both Tuesday and Wednesday nights, after a long day simulating the field investigation, the Course attendees gathered in a large classroom for the nightly NTSB progress meeting, followed by the ALPA-only meeting —with cold pizza and warm soda, just like in real life. Each night, before the NTSB meeting, they watched a videotape of "NTSB member" Keith Hagy (actually, assistant director of ALPA’s Engineering and Air Safety Department) hosting a press conference to release the limited information available at that point in the field investigation.

They then listened as Capt. Fenwick, in his role as the NTSB IIC, asked the group chairmen to summarize their groups’ work for the day. Capt. Fenwick advised the course participants that, in an actual NTSB progress meeting, the ALPA representatives and party coordinators must stay mum; only the IIC and the group chairmen are typically allowed to speak.

The third day of the course was devoted to an extensive debriefing about the Course itself—especially important because this was the first time ALPA and UND had conducted this type of joint training.

Capt. Gunther advised that, in an actual NTSB accident investigation, "you’ll usually have five to seven days and be in only one group, but you’ll still have to hump it from sunup to sundown and go to progress meetings every night. After you’ve turned in your field notes and the wreckage is removed, all your evidence is gone."

He added, "Don’t always assume that the wreckage is going to be in the same location and orientation as the day before. Don’t assume that the yellow tape encompasses the entire wreckage—oxygen bottles, balance weights, engines, and other heavy parts can travel far from the rest of the wreckage."

Capt. Fenwick noted, "Having ALPA representatives on the CVR and FDR groups is absolutely critical."

Capt. Frey added, regarding the CVR group, "Each group averaged about 400 to 500 words [transcribed] in about an hour and forty minutes of work. Over the course of two days in an actual investigation, you can expect to transcribe about 2,500 words."

Asked about NTSB analysis of sounds other than voices, Capt. Frey replied that the Safety Board has staff engineers who conduct acoustic spectral analyses.

Capt. Sicchio said, "The folks here at UND were helpful in providing some computer-based maintenance records. As Frank [Argenziano] demonstrated, for a maintenance records group, the first day of its investigation should involve learning the maintenance records system being used."

Participating in the NTSB Maintenance Records Group on the Alaska Airlines Flight 261 accident investigation, said Capt. Fenwick, was a full-time job for a couple of years for some of the Group members. Capt. Wilkes added, "All three years—and it’s still going on."

The Course participants discussed whether the wreckage should include an actual engine and propeller, which were not included in the course because of the difficulty of moving the engine. Capt. Gunther said he didn’t feel it was necessary, because during actual accident investigations, "we’ve suddenly realized, ‘Hey, we’re missing major portions of the airplane!’"

Capt. Brady stressed that the goal of the course "is not to determine the probable causes of the accident—it’s to learn the process."

Capt. Gelinas complimented the UND ATC students for their help and participation. "They received their scripts only on Monday," he pointed out. "They were nervous during the first couple of group interviews, but polished by the final ones. The ultimate goal is for the UND students to script this scenario and run through it in their 360-degree tower simulator."

Capt. Gelinas added that, regarding the previous day’s role-playing, when Capt. Burgess admitted that he’d talked to a CNN reporter, in real life, he’d be kicked off the NTSB group. The entire ALPA group would be sent packing, "with irreparable damage to the Association’s relationship with the Safety Board for years to come," Capt. Gelinas asserted.

Capt. Miller said it would have helped him to have been told a couple of weeks earlier to review "The Red Book." Again, Capt. Gelinas said the Course simulated real life: "Swissair 111 crashed," he recalled, "and an hour later, my phone started ringing off the hook. For the next 18 months, I didn’t have a life. I recommend you review ‘The Red Book’ while you’re enroute to the accident site."

Capt. Miller asked how "high-tech," how "wired" the field investigations have become. Capt. Gelinas warned that new laptop computers are being built with chips for wireless Internet connections. He said that, without proper security controls on the laptop computer, "you could be sitting in the airport terminal, typing away, sending e-mail, and the CNN guy across the aisle is surfing the web, but he’s also picking up your stuff."

Regarding a suggestion that instructors telephone course participants at 3 a.m., to further heighten the realism of the mock investigation, Capt. Miller elicited hearty laughter from the group when he commented, "I don’t think you need to train to be miserable."

True enough; but this vanguard of 19 ALPA members now comprises one of the best-trained group of aviation accident investigators anywhere—ready to serve their fellow pilots and the traveling public when the real call comes at 3 a.m.

The misery just comes with the territory.