The Sept. 6, 2016, issue of Flight School Business noted that while the recent reduction in the overall rate of non-commercial fixed-wing accidents has been broad-based, with no dramatic improvements in any specific area, the decrease in the corresponding rate of instructional accidents has not. Fixed-wing flight training saw the greatest improvement where most of the accidents are, which is to say on landing attempts. While those accidents are almost always survivable, aircraft that spend less time in for repair and more on the flight line has to please school operators and their underwriters.
Helicopter training also has registered significant improvements during the same period, though the smaller numbers involved amplify the relative magnitude of year-to-year variations. Grouping several years’ data together helps smooth out those swings, and in this case the annual counts suggest some natural breakpoints:
Between 2002 and 2005, there were an average of 45 crashes on training flights each year. Three, on average, were fatal—only about half the lethality rate of fixed-wing instructional accidents. At first glance the next four years appear almost unchanged, averaging 43 per year, but 2007 was an outlier: It saw 61, one-third more than any other year in the period.
If we exclude that, the average for 2006, 2008, and 2009 drops to 37 with fatal injuries in three. Since 2010, though, we’ve averaged only 28 per year—a 38 percent decline from that initial period, with only eight fatal accidents in the past six years (an average of 1.3). The worst year of the six—2011, when there were 32—would have been the second best of the preceding decade.
And to answer the obvious question, it’s not because we’re flying less. According to the FAA activity survey, helicopters logged 20 percent more instructional hours in 2014 than they had in 2007, and this went up another 10 percent in 2015. If that holds, the instructional accident rate in rotorcraft will have dropped a staggering 71 percent from 11.8 per 100,000 hours in its worst recent year to 3.4 in the most recent, which also qualifies as its best.
So, what’s been getting better? Funny you should ask. One thing appears to be the aircraft themselves. The share of accidents precipitated by some kind of mechanical failure dropped from 16 percent in the first four years to 10 percent in the most recent six. Kissing the ground hello and good-bye also appears to be getting a little easier; clumsy airmanship on takeoffs and landings accounted for 16 percent in 2002-2005 but 12 percent in 2010-2015. And remember, those are diminishing slices of a dwindling pie. Mechanical problems caused an average of seven accidents per year from 2002 through 2005, six in 2006-2009, and three per year since 2010. The numbers for takeoff and landing accidents are nearly identical.
Perhaps inevitably, errors during low-altitude maneuvering remain the most common problem; their share of all instructional accidents has remained steady at about 40 percent. And as every CFI-helicopter knows, the vast majority of these—90 percent in 2002-2005 and in 2010-2015 alike—are practice autorotations that don’t work out as planned. (They dropped to 77 percent in 2006-2009 due to an odd but probably random spike in losses of control while attempting pedal turns.) Even so, reductions in autorotation accidents contributed to the overall decline. Annual averages fell from 16 in the first four years to 14 in the next four—then to 9.5 in the final six. Losses of control while hovering in ground effect have also fallen off from an average of four per year to just one.
So helicopter training, too, has concentrated its improvements where the greatest problems are. But not everything’s getting better. The number of dynamic rollovers, while small, has crept up from three in four years to nine in six. And while 2002-2005 saw just one accident ascribed to improper procedures while practicing simulated failures of governors or hydraulic systems, there were three between 2006 and 2009. Six more have occurred since.
Still, these are minor setbacks when gauged against the overall progress that’s been made. Further improvement is certainly possible—but a 42-percent annual reduction in autorotation accidents qualifies as a pretty good start.