BC&A / Accident-Prone Flight Organizations
May 30, 2007 :: Accident-Prone Flight OrganizationsMay 10, 2007
By Patrick R. Veillette, Ph.D.
In his submission to NASA's Aviation Safety Reporting System, the pilot complained that his employer "has become increasingly confrontational, condescending and oppressive toward the pilot group. . . . This kind of atmosphere is fatiguing, numbing, unhealthy and downright dangerous to work in."
Welcome to aviation. If you've been flying long enough, you've probably worked for such an operator. And you've probably been told, "If you don't make this flight I'll find someone who will." And of course let's not forget the management practice of paying the pilot only after the successful completion of the flight; no undue pressure there, especially when the rent is due.
Two prominent aviation safety authorities have expressed what is clearly obvious to any pilot who has endured such management pressures. Dr. John Lauber, when serving as an NTSB member, once said, "Behind every 'pilot error' accident was a management accident." Meanwhile, Dr. Richard Jensen, retired director of the Ohio State Aviation Psychology Laboratory and former editor of The International Journal of Aviation Psychology, states in his book Pilot Judgment, "Until recently, responsibility for human error was placed almost exclusively on the pilot. We are finding that the organizational environment may be even more responsible for these 'pilot errors' than the pilot."
Consider the official record of an aviation organization that operated two Beech BE-58P Barons, an Aero Commander 500B, two DHC-6 Twin Otters and a Basler-converted DC-3TP during the mid-1990s. The combined flight time in an average year for the organization was around 1,400 hours. Piloting the aircraft were a half dozen full-timers, supplemented by some seasonal pilots. The vast majority of the flying was VFR-only. Using the NTSB's estimates of an average of 3.47 accidents per 100,000 hours for the FAR Part 135 industry, an aviation unit flying 1,400 hours per year should be able to fly 20.6 years before having a single accident. And yet this relatively small aviation unit experienced six mishaps within a 15-year time period.
The first of the accidents occurred on Aug. 5, 1980, when the unit's AC-500B was dispatched on a routine flight from Salmon to McCall, Idaho. Witnesses observed the aircraft take off from the Salmon airport and enter the nearby canyon at a relatively low altitude. The aircraft was not heard from thereafter. It wasn't until 11:20 a.m. on the next day that the wreckage was found, along with a severely injured passenger and the fatally injured pilot. The internal investigation found that significant wind shear at low altitude was the proximate cause of the accident. The accident investigators also found that pilot testing and certifying procedures were inadequate to meet the intent and needs of this pilot for proficiency. The pilot had no previous history of flying multiengine airplanes in mountainous terrain. After just three practice flights, the pilot had been authorized to perform personnel transport, low-altitude reconnaissance, back country/remote mountain airstrip operations and fire reconnaissance, all involving highly specialized and highly risky operations.
The second accident occurred on June 24, 1988. A senior check airman who had flown seven different types of aircraft in six weeks piloted one of the P Barons to McCall, then hopped into the unit's Beech 99 and flew a flight around the local area dropping a set of parachutists. During this segment, the pilot pulled the circuit breaker on the gear warning horn since it was a nuisance during drops. However, when he returned to McCall, the pilot failed to extend the landing gear after entering the traffic pattern and the disabled "gear not down and locked" horn was silent. The pilot didn't realize his oversight until the propellers began scraping on the runway. In the final report, the investigators would note that the check airman disregarded a number of the unit's own policies, including failure to use the proper checklist and violating agency rules restricting its pilots to three types of aircraft.
Later that summer, while taxiing out of the McCall ramp and doing the aircraft's "before takeoff" checklist, another pilot in the unit was interrupted by a radio call from a dispatcher and inadvertently left the right fuel selector in the cross-feed position. After several hours of flight, the right engine quit, and the left engine failed shortly thereafter. The pilot put the airplane down on a sandbar in the Snake River and it promptly sank in 10 feet of water. When the accident investigation team pulled the aircraft out of the water, they found the fuel cross-feed switches in an incorrect position.
On Jan. 26, 1989, the unit's BE-58P Baron departed Ogden for a practice instrument proficiency training flight. At the conclusion of the flight the aircraft was going to be delivered for a scheduled 100-hour maintenance inspection. The PIC was a 6,800-hour pilot accompanied by a check airman from the unit. They performed a number of approaches and missed approaches at nearby airports and then landed at Salt Lake City International. Even though the runway was nearly 9,600 feet long and the maintenance shop was at its far end, the PIC decided to raise the flaps. Unfortunately, he raised the landing gear lever instead, resulting in retraction of the landing gear while the twin Beech was still rolling down the runway. The aircraft skidded on its underside for another 700 feet before coming to rest on the edge of the runway. The propellers, engines, landing gear and doors, and underside of the fuselage were substantially damaged. The accident investigation team determined pilot error to be the cause of the accident.
Then on April 8, 1991, the very same pilot pulled the Baron out of the hangar using a gas-powered tug. After the preflight, the pilot hopped in, cranked up the engines, taxied to the active runway and took off. All of this was uneventful until the pilot raised the gear handle. That's when a red light illuminated, informing the pilot that something was wrong with the gear. While the pilot was unable to determine the problem, a tower controller identified it soon enough: the tug was still attached to the nosewheel! Somehow -- rather unbelievably -- the pilot had been able to taxi and take off while dragging the tug with him. Some structural damage occurred to the aircraft as it touched down on landing, although considering the circumstances, many had been afraid it would result in far more serious consequences. "Pilot error" was determined to be the cause of this remarkable event, which was declared to be an "incident with serious potential."
Next, on June 17, 1994, a senior check airman was taxiing the unit's turbine-powered DC-3 from the ramp at Springerville, Ariz. The flight crew was busy performing the taxi checklist when the wing of the old Douglas smacked a fence post, causing substantial damage to the airfoil. Neither of the pilots was aware of the collision until notified by radio. The accident investigation report found that pilot error was the cause of the accident.
Finally, on Aug. 22, 1994, one of the unit's Twin Otters flew a routine cargo flight to an airstrip in the middle of Idaho and returned to its summer home base of McCall. Even though there were three different entries on the de Havilland's checklists advising crews to make certain the nosewheel tiller was centered, it was not. So when the Twin Otter touched down, the aircraft swerved violently swerved off the runway, coming to rest in a row of tall willows. The swerve was so severe, it bent the forward fuselage enough for the right propeller to slice into the fuselage skin. The internal accident investigation placed full blame for the accident on the pilot.
While all of these accidents and incidents had "pilot error" as their proximate cause, one must wonder if there were deeper factors involved within the organization. Perhaps it was coincidence that these six accidents involved the same three pilots or that four of the accidents directly involved check airmen in the unit? It's easy for accident investigators to fault an errant PIC, but a deeper thinking investigator would definitely take the time to explore whether some other issues within the organization were also links in the error chains.
After the nosewheel landing accident, the unit's managers were anxious to show that the accidents were solely the fault of a few error-prone pilots. Consequently, the chief pilot and two other check airmen -- pilots described by the manager as "the highest skilled" -- were placed in a DHC-6 simulator to prove their prowess. But when the simulator instructor failed the engine on takeoff, the chief pilot incorrectly tried to counter the loss of power and lift and the sim made an abrupt roll into the dead engine and crashed. The sim instructor put the simulator back at the takeoff point and had the chief pilot try the maneuver again, with the same results. Similarly, the other two pilots crashed on their first attempts to handle an engine failure.
As the simulator session continued, the instructor noticed the pilots ignored some important procedures and industry practices. At the end of the experiment, he wrote a letter to the unit's manager delineating a "summary of the weaknesses," which included:
*Lack of standard operating procedures; each pilot "did it their own way" and normal checklist needs revision.
*Lack of application of crew concept; no crew briefings, ineffective departure briefings, poor communications procedures, inconsistent checklist procedures.
*No application of human factors training (CRM); inability to create a synergistic cockpit environment for problem solving.
?228-137?Even a minor abnormality created a major loss of crew situational awareness.
The instructor further wrote, "The fact that flight currency with your crews and lack of SOPs are an issue should be a concern. This is especially true given your highly specialized type of operation." The chief pilot's confidential summary of the training admitted, "our cockpit resource management skills were poor and below industry standards. . . . Standard Operating Procedures were nonexistent. These procedures are an absolute must to operate as a safe and efficient crew."
In the end, the unit's manager admitted to his bosses that, "The results pretty clearly show that our pilots do not meet industry standards. We cannot only not fly single-pilot crew; we don't even meet standards with a two pilot crew."
Parenthetically, despite the unit's troubling safety record, the FAA never took action against it since it was a "public use" agency, and as such was then beyond the FAA's jurisdiction. Although adoption of Public Law 103-411 (49 U.S.C. 101) in 1994 in theory requires that portion of the "public aircraft" fleet used for carriage of passengers or cargo comply with FAA standards and be subject to FAA inspection, discussions with several managers of "public aircraft" units tell me that they still operate with very little oversight from the FAA.
With that as background, the question that surfaces is what factors cause an organization to be error-prone? In the book Managing the Risks of Organizational Accidents, Dr. James Reason, emeritus professor of psychology, University of Manchester, lists a number of organizational characteristics prominent in error-prone operations. Among these are a manifest lack of organizational safety culture, poor supervision and checking, group "norm" that accepts violations, misperception of hazards, a macho culture that encourages risk taking, perceived license to bend rules and ambiguous or apparently meaningless rules.
Dr. Alan Diehl, recently retired as the U.S. Air Force's technical advisor for human factors and the individual who pioneered the NTSB's Human Performance Investigation Procedures, spent several decades investigating both civil and military accidents. One of the earliest in his career involved a DHC-6 that impacted short of the runway at fog-enshrouded Rockland, Maine, airport in May 1979, killing 17 people. It was apparent that the crew had descended into terrain, but Diehl wanted to determine why they had done so.
Although the published minimums for the approach in Rockland were 440 feet, Diehl discovered pilots at this regional carrier had been exposed to a system of psychological reward and punishment that instilled the dangerous idea that pilots who really wanted to keep their jobs could safely descend below approach minimums, and that they did so regularly. Digging deeper, Diehl found a strong record of undesirable management pressures, inadequate pilot training and increased maintenance problems. These were brought about by strong management emphasis on profitability and minimal FAA surveillance activity. Such mismanagement is not confined to just civilian flying organizations; Diehl notes many similar conditions within military aviation units in his book Silent Knights: Blowing the Whistle on Military Accidents and Their Cover-Ups.
Inadequate oversight was among many of the specific issues that the STS Columbia Accident Investigation Board noted as contributing reasons for the tragic end to that space shuttle and its crew. U.S. Air Force Gen. Duane W Deal, then commander of the 21st Space Wing and a member of that accident board, issued a 10-page augmentation to the main report stressing the need for more and better trained inspectors to oversee contractor work. Inadequate oversight has been cited by the NTSB in many accidents in which FAA inspectors are based a significant distance from the aviation operations involved. In particular, the NTSB cited the lack of adequate oversight in many of the heli-logging industry's accidents and also in many of the EMS helicopter accidents. The NTSB also made a stinging report of the FAA's inadequate oversight of its own flight operations in the accident involving an FAA Beech 300 near Front Royal, Va., in 1993. (See "FAA as Mismanaged Operator" sidebar.)
There are other issues that can make an organization vulnerable to accidents. U.S. Air Force Pamphlet 127-1, the workbook for conducting accident investigations, directs investigators to examine a number of elements within the organization to determine their possible role in the causal chain. Among these are peer pressure and community factors, including expectations of conduct and vicarious learning. Supervisory issues also have a strong influence on an organization's propensity for error, and among the various issues that have played roles in the past are discipline enforcement, command tone, perception of double standards and command/ management pressure.
Dr. Robert Helmreich, director of the University of Texas CRM Human Factors Research Program and considered one of the world's foremost authorities on CRM, found that "an organization's culture demonstrates its attitudes and policies about human error, the openness of communications between management and flight crew, and the level of trust between flight crew and senior management. Organizational culture also influences norms regarding adherence to regulations and SOPs. Of great importance, the organizational culture determines the level of commitment to safety and the strength of a safety culture." (Proceedings of the First ICAO-IATA LOSA and TEM Conference, Dublin, Ireland, 5 to 7 November 2003)
The Dutch National Aerospace Laboratory studied reasons that flight crews do not adhere to procedures, and several of these are organizational issues. Among these are a lack of information provided by management, insufficient training and/or experience to perform procedures without errors, a company culture that stresses other elements as being more important than following SOPs, and inconsistent company philosophy, policies and procedures.
One important step in managing risk within a flight group, Jensen says, is to "communicate to all parts of the organization that safety is a deep concern and vitally important to the company. Safety concerns can and must be stated directly but must also be demonstrated in the way top people are seen to behave and in the policies they develop for the company."
According to Jensen, "Individual pilot judgment is influenced by the way management is seen to practice its own standard operating procedures." And one way to show managerial commitment is to create a safety officer who reports directly and only to the top officer within the company. Picking the right person is key, Jensen cautions, noting, "The safety officer must be someone who is seen as approachable by the pilots so that they can communicate their own weaknesses, errors and perceptions of hazards without fear of reprisal."
Jensen's research with small business aviation organizations found that, "The commercial pilot for a small charter operator, in particular, is faced with some of the most difficult decisions in aviation today. As in all commercial enterprises, the bottom line is to make money. Unfortunately, pilots flying for hire are often faced with the need to balance financial reward with safety on a day-to-day basis.
In these situations they may be faced with pressures by superiors, peers and/or passengers who are more in tune with the financial or convenience side of the balance than with the safety side." Accordingly, he recommends that owners of high-performance and turbine aircraft take an introductory course on the subjects of stress, crew rest and pilot judgment. He feels this would aid corporate pilots since intelligent operating policies endorsed and disseminated by an aviation-savvy management would help ease decision making, thereby relieving the crews of a great deal of stress and tension.
Perhaps. The managers of the accident-prone aviation operation cited in the beginning of this article attended over the years a plethora of CRM courses taught by various airline personnel and other experts. The effectiveness of these sessions can be discerned in the "Playing the Odds sidebar."
As revealed in the Columbia investigation, some traits of organizations are so ingrained culturally that mere changes on an organizational chart and weekend seminars cannot excise the deep-rooted characteristics that invite operational safety errors. The experts have spoken and the evidence is clear: Pilot error can result from organizational flaws allowed or promoted by inept or wrong-headed management. To help ensure the safety of crews, their passengers and aircraft, those concerned must focus attention on organizational culture and how it stymies or promotes a professional and safe flight operation.
