Comair Flight 5191: Case Study in Fatigue Aviation Safety: Fatigue

Comair Flight 5191: A Case Study in Fatigue

Comair Flight 5191: A Case Study in Fatigue

Comair Flight 5191 came to a disastrous end in 2006 when the flight crew attempted to take off from a runway much shorter than required for the aircraft they were piloting, resulting in the deaths of 49 of the 50 people on board (Pruchnicki, Wu, and Belenky, 2011). The Comair Captain, First Officer, and air traffic controller failed to perform the necessary checks to ensure the plane was lined up on the proper runway before takeoff. The National Transportation Safety Board (NTSB) investigated the accident and could not definitively determine the cause. Years later at a sleep conference, the NTSB chairman Deborah Hersman mentioned the Comair Flight 5191 tragedy and noted that establishing fatigue as a significant contributing factor is often so difficult that it is only offered as a last resort.

A survey of reported fatigue levels was published the same year that the Comair tragedy occurred and based on the replies of 162 short-haul pilots, close to 75% were suffering from severe fatigue (Jackson and Earl, 2006). Given this data, it should come as no surprise that there have been several incidents recently when flight crews fell asleep during the flight and missed the runway, sometimes by more than 100 miles (Caldwell, 2012). To better understand the human factors that can lead to dangerous levels of fatigue, a forensic analysis of the Comair Flight 5191 tragedy in terms of sleep, circadian rhythm, and workload irregularities will be reviewed.

Defining Fatigue

Pruchnicki and colleagues (2011) believe fatigue is the product of at least three factors: (1) sleep history, (2) circadian phase, and (3) workload. Sleep history is relevant because poor sleep quality will cause sleep debt to accumulate. Unless crew members eventually engage in sufficient recovery sleep, performance will suffer further as the magnitude of sleep debt grows. The circadian rhythm is important for regulating a number of important biological processes, including the sleep and wake cycles. If a flight crew member were to pilot an aircraft at a time they are normally asleep, their performance would be negatively impacted due to reduced levels of cognitive arousal. Workload levels are relevant because an absence of downtime can aggravate the performance-degrading effects of sleep debt and circadian rhythm disruption. With market pressures forcing reductions in operational costs, flight crews are experiencing shorter turnaround times at airports. This practice increases the amount of duty time accumulated per shift and thus the overall workload.

Case Study in Fatigue: Comair Flight 5191

Pruchnicki and colleagues (2011) conducted an analysis of the fatigue-related factors that may have contributed to the Comair Flight 5191 tragedy. Their goal was to test whether the Sleep, Activity, Fatigue, and Task Effectiveness Model/Fatigue Avoidance Scheduling Tool (SAFTE/FAST) would have been capable of predicting impaired performance due to fatigue. After reviewing the previous four days of sleep history for the Comair Captain, First Officer, and the air traffic controller, the only data deemed reliable enough for use with the SAFTE/FAST model was the sleep and workload schedule for the air traffic controller. The information concerning the captain and first officer contained gaps that precluded its use with the model. Estimates of flight crew fatigue were therefore based on what occurred in the cabin 30 minutes prior to the accident. It was assumed that all three personnel were synchronized in terms of circadian phase, because all three lived within the Eastern Time Zone.

Air Traffic Controller

The bedtimes and wake times of the air traffic controller from Wednesday, August 23 to Sunday, August 27 (day of accident) were 1:45/9:15 AM (in bed 7.5 hrs), 10:30 PM/7:15 AM (in bed 8.75 hrs), 10:30 PM/5:40 AM (in bed 7 hrs), and a nap between 3:30 and 5:30 PM on Saturday during an off duty period (Pruchnicki, Wu, and Belenky, 2011). The shifts worked were 5:00 PM -- 1:00 AM (8 hrs), 3:00 PM -- 9:00 PM (6 hrs), 8:15 AM -- 4:15 PM (8 hrs), and 6:30 AM -- 2:30 PM (8 hrs) and 11:30 PM until the accident at 6:06 AM (7.5 hrs) the following Sunday morning. Based on this information, the air traffic controller was able to sleep on a consistent schedule until the night before the accident. This schedule represents a typical 2-2-1 rotational shift beginning with two afternoon shifts, followed by two daytime shifts and one nighttime shift.

The shift schedule would therefore have interfered with the ability of the air traffic controller to maintain a healthy circadian rhythm....

By Saturday, the air traffic controller tried to get enough sleep during the off duty period, but the sleep/wake schedule maintained over the previous 3 days would have resulted in the body temperature increasing arousal levels to the point that sleep was difficult. This would explain an inability to get more than 3 hours of poor quality sleep before starting the shift when the accident happened. The accident also happened at 6:06 AM Sunday morning, a time when the air traffic controller would normally have been asleep. Based on this information, the air traffic controller at the time of the accident was suffering from acute sleep deprivation, experiencing significant circadian rhythm disruption, and working a second shift without benefit of restorative sleep.
According to the voice recording immediately before the accident, the air traffic controller made several mistakes (Pruchnicki, Wu, and Belenky, 2011). First, the crew was not notified that they would be passing a closed runway as they taxied towards the correct runway. Second, the air traffic controller cleared Comair Flight 5191 for takeoff as it was waiting next to the closed runway. The air traffic controller was conducting routine tasks as the plane attempted to take off from a runway that was too short, then took a full 45 seconds to notify emergency personnel. The air traffic controller then began to fill out incident paperwork instead of directing emergency personnel to the crash site.

Flight Crew

An analysis of the flight crew's behaviors immediately prior to the accident is also consistent with a state of fatigue (Pruchnicki, Wu, and Belenky, 2011). Both the Captain and First Officer were required to get up early to prepare the flight for takeoff, which probably truncated the previous night's sleep period. The flight crew then began to power up the wrong plane until the ground crew notified them of their mistake. Once on the correct plane, the captain and first officer made a number of procedural omissions. Checklists were not performed correctly and briefings were incomplete. The most important mistake was conducting an abbreviated taxi briefing, which was performed atypically by the First Officer. The information missed in the briefing was the location of the closed runway they would need to pass before arriving at the correct runway. As a result, they positioned the plane next to the first runway they came to. The First Officer yawned twice and on five occasions discussed inappropriate topics during takeoff preparations just minutes before the accident, the latter violating sterile cockpit procedures. The Captain also violated sterile procedure once. The workload right before the accident would have been higher, since the flight crew went to the wrong plane first and arrived quickly at the first runway they noticed. The stress of rushing preflight preparations would have aggravated the effects of sleep deprivation and disrupted circadian rhythms. Based on this analysis, Pruchnicki and colleagues (2011) concluded that the evidence is consistent with Comair's flight crew suffering from significant fatigue at the time of the accident.

During the period immediately prior to the Comair Flight 5191 accident, two other flights had taken off successfully (Pruchnicki, Wu, and Belenky, 2011). When interviewed, both crews admitted experiencing some confusion about which runway they were supposed to use. Given the survey finding that 75% of short-haul commercial pilots report suffering from severe fatigue (Jackson and Earl, 2006), the suggestion by Pruchnicki and colleagues (2011) that the flight crews of these successful takeoffs were also suffering from fatigue is reasonable given the early hour. Since the same air traffic controller was involved in all three takeoffs, the confusion that the successful flight crews experienced probably caused enough concern that they persisted in locating the correct runway. By contrast, the taxi briefing by the First Officer of Comair Flight 5191 did not generate any noticeable concern based on the cockpit recording.

To summarize, the air traffic controller responsible for Comair flight 5191 was probably suffering from acute sleep deprivation and circadian rhythm disruption at the time of the accident. Pruchnicki and colleagues (2011) concluded that perseveration, which is the repetition of cognitive and motor activities whether or not appropriate, played a significant role in the Comair tragedy. As Comair Flight 5191 was beginning to taxi down the runway, the controller became engaged in routine tasks unrelated to the Comair flight. After the accident, the air traffic controller did not notify emergency personnel for 45 seconds and then began filling out accident-related paperwork rather than helping to direct emergency personnel to the crash site.…