The annual rates of flight operations events, or predefined exceedances of selected parameters, continued to decrease in the most recent analysis of corporate flight operational quality assurance (C-FOQA) business jet data for the years 2006–2011 by Austin Digital C‑FOQA Centerline.1 Also notable was a reduction in the rate of unstable approach events in 2011, to the lowest level in the six-year period.
The trend was particularly encouraging because of growth of participation in the program, with these flights and events making the data even more statistically relevant to this industry sector. While event rates continue to decline overall, there is still concern that go-arounds following unstable approaches are not being conducted, and there is persistent evidence of high-energy approaches, both of which are strongly associated with runway excursion events as shown in the 2009 briefing of the Runway Safety Initiative coordinated by Flight Safety Foundation.
Exceedances represent cases when an event’s parameter data are considered less than optimal for safe operation. For example, during approach, C-FOQA analysis flagged an exceedance if the flight data recorder showed that the aircraft was above or below the glideslope, or was out of alignment with the localizer, in each case by a stipulated deviation. It also tagged an event, for another example, if the ground-proximity warning system (GPWS) produced a “sink rate” or “pull up” warning at certain altitudes and rates of descent. Events were categorized under the headings of unstable approach, flight operations, risk monitoring, aircraft limitations and aircraft systems.
During 2010 and 2011, flight operations event rates have been under 10 per 100 flights, with an improving trend since 2007 (Figure 1). The length of the error bars has also been decreasing.2
For risk reduction in flight operations — as in other event categories — it is important not only to know the rate of events, but what kind of events they were. This offers a clue to the relative prominence among the event types that might be accident precursors. In 2011, the highest rate — nearly double the next highest — involved GPWS glideslope alerts below 3,000 ft radio altitude (Figure 2). Analysis discovered that most of the glideslope excursions revealed a tendency to “duck under” during very short final. This was supported by the average threshold crossing height of 36.1 ft, somewhat lower than the target height of 50 ft.
Other relatively common events included bank angle exceedance, altitude excursion, a traffic-alert and collision avoidance system resolution advisory lasting more than two seconds, and exceedance of expected deceleration during rollout.
Annual unstable approach event rates in 2011 decreased notably from those of 2010 and were the lowest in the 2006–2011 data set (Figure 3).3 These events, unlike flight operations events, showed no discernible trend during the five previous years of the C-FOQA program. For all program years aggregated, the highest rate of unstable approaches occurred in the third quarter — July, August and September.
Considering types of unstable approach events, the highest rate in 2011 was for flying above the glideslope, only slightly less frequent than events involving being fast on the approach (Figure 4). Of those flights when the approach was flown above the glideslope, nearly half were between 0.00 and 0.25 dots high. About 0.3 percent of the flights were between 2.00 and 2.25 dots high, past the caution limit, and the greatest deviation — about 0.1 percent of the flights — was 3.50 dots high or more. Of greater concern is that four of the top five unstable approach event causes indicated high-energy approaches, strongly associated with runway excursions.
Preliminary analysis of unstable approach events correlated with both time of day for the event and length of the runway have been introduced. The aggregate data are statistically irrelevant at this point; however, individual operators now have the ability to look more closely into their own operations, where event rates may be more statistically relevant.
Risk monitoring events concerned alerts or cautions for threats such as fuel exhaustion, controlled flight into terrain (CFIT), stall, landing overruns, hard landing and tail strike. The highest rates in 2011 were for CFIT risk and risk of a landing overrun (Figure 5).
C-FOQA Centerline says that it is now working to provide pilots with more detailed information concerning the risk of landing overrun events by offering analysis of landing performance for the first time. That includes monitoring threshold crossing height, airspeed at threshold, float distance, tailwind at threshold and runway remaining when slowed to 80 kt. These events will be combined to provide analysis of the newly drafted stabilized landing concept.
Event rates in 2011 that exceeded the aircraft’s recommended operating limits tended to be low; those events primarily consisted of calibrated airspeed beyond the aircraft model’s flap speed limit, based on the aircraft’s reference flight manual. That occurred slightly more than 0.7 times per 100 flights. C-FOQA Centerline says its data suggest that flap overspeed events generally decline markedly after the first two years in the program when flight departments take measures to reduce them.
All the other measured operating limits exceedances occurred less than 0.2 times per 100 flights.
Aircraft system events for 2011 were negligible in number except for selecting or maintaining reverse thrust while decelerating at relatively slow speed.
Participation in the C-FOQA program has grown steadily since it was initiated in 2006 (Figure 6, p. 50). In 2011, more than 10,000 flights contributed data, for a total of more than 30,000 flights since the program’s origin. Twenty-five operators participated in 2011, with the data representing 73 aircraft of 16 types or variants.
Pilot Fatigue Barometer
The European Cockpit Association, which represents national pilot associations of 37 European states, has summarized the results of surveys conducted by some of its members in a report titled “Pilot Fatigue Barometer.”4 The surveys were carried out between 2010 and 2012 in Austria, Denmark, France, Germany, the Netherlands, Norway, Sweden and the United Kingdom. Some 6,000 pilots responded to queries about how fatigue affected their flying performance.
Mentioning several well-known accidents in which pilot fatigue was cited as a causal factor, the report also suggests that fatigue often goes unreported in accidents and incidents — first, because pilots are reluctant to admit flying “under the influence” of fatigue out of concern it could provoke punitive action by an employer or even criminal prosecution; second, if the pilots happen to be killed in an accident, fatigue leaves no material evidence.
The main potential consequences of fatigue during flight duty include degradation of thought processes, perception and reaction time; periods of unintended sleep; and momentary “micro-sleep.” Percentages of pilots who reported having experienced fatigue in the cockpit ranged from 93 percent of those in the Denmark survey to 45 percent who responded to the U.K. survey. Pilots who said they had dozed off or had a spell of micro-sleep ranged from 54 percent of respondents from Sweden to 10 percent of those from France. The report did not speculate on the reasons for the national differences among responses.
“More than three out of five pilots in Sweden (71 percent), Norway (79 percent) and Denmark (80–90 percent) acknowledge [having] made mistakes due to fatigue, while in Germany it was four out of five pilots,” the report says.
Responses indicated that 92 percent of German pilots reported that they had felt “too tired” or “unfit” for duty on the flight deck at least once in the previous three years. In the Austrian pilot association, 85 percent of respondents reported that they had been too fatigued for flight duty but nevertheless had reported for their assignments. Two-thirds of those said they had flown under that condition more than once. Swedish and Danish pilots reported similar percentages.
“According to the surveys among pilots, night flights or a series of night flights are major contributors to fatigue,” the report says. “For example, in France, almost 70 percent of the pilots identify night flights as a cause of fatigue. Nearly half of the respondents in Germany agree that night flights are one of the major causes of pilot fatigue. …
“The study among British pilots shows that fatigue prevalence is associated with the number of sectors, flying and duty hours or [the commander’s decision making] frequency.”
Other identified causes included a series of morning departures; insufficient rest between duty periods; being recalled from standby status; and inadequate rest accommodations.
Nevertheless, the report says, only 20 to 30 percent of the pilots polled reported that they had acknowledged feeling unfit for duty. “Such under-reporting of fatigue has been confirmed by an independent survey of 50 U.K. aviation medical examiners in April 2011,” the report says. “The vast majority (70 percent) of the aviation medical examiners believe that pilots are reluctant to report fatigue within their company.”
About a third of the pilots who chose not to file fatigue reports gave as their reason that they were too tired at the end of an exhausting workday.
- The C-FOQA User’s Group — comprising all program participants — is led by a steering committee plus Austin Digital, Flight Safety Foundation and other external parties.
- Austin Digital explains the meaning of error bars: “When displaying event rates (e.g., events per 100 flights), it is appropriate to compute proportion confidence intervals — error bars — along with the raw event rate. These bars indicate a range, based on the number of flights sampled, within which the true rate likely falls with high confidence. In general, the larger the sample of data (i.e., more flights), the smaller the error bar will be and the more confident you can be that the resulting rate is a statistically significant value.”
- Unstable approach criteria are aligned with the elements published by Flight Safety Foundation, which include nine requirements, all of which must be satisfied. The criteria also specify that flights must be stabilized by 1,000 ft above airport elevation in instrument meteorological conditions and by 500 ft above airport elevation in visual meteorological conditions. FSF ALAR Tool Kit, Briefing Note 7.1.