I recently received an interesting message from Kevin Recker, who is a senior engineering manager for General Dynamics in Scottsdale, Arizona. The group he leads has built space flight hardware for the Viking missions, the Apollo Program, the International Space Station and the Mars Rovers. The equipment it builds has to be right and can't fail despite extreme operating conditions. Because of this, risk management is very important to the group, and it has many tools that it uses to analyze and then minimize risk. In fact, I have provided error prevention training to many of the employees at General Dynamics, and Kevin has attended my training.
Kevin said he has been trying to apply some of the group's risk mitigation techniques to his flying, and he was wondering if it would be possible to "lay out very plainly for pilots, in numbers, how much they can reduce their risk by taking certain actions." For example, Kevin said that, because of the limited options available in the unlikely event of engine issues over Arizona terrain, he has decided not to fly there at night. He also always files a flight plan for all flights longer than 20 minutes, and visually checks the fuel level before each flight.
The simple answer to Kevin's question is that it is not possible to come up with a number that would in any way accurately indicate the reduction in risk from steps such as deciding not to fly at night. The problem is that we have no detailed data about the amount of time flown at night, on instruments, etc. We have only total flight hours for categories like single engine and multiengine, and even those are not necessarily accurate. However, we can use accident data from the Air Safety Foundation's 2009 Nall Report to make some general statements about how pilots can reduce their risk. We will be examining only the pilot-related accidents, which accounted for 72 percent of all the general aviation accidents in 2008.
The most important steps a pilot can take to reduce his risk of an accident are to learn how to land well, especially in gusty crosswinds, and to maintain that skill through regular practice. One-third of all general aviation accidents in 2008 involved a bad landing, making it by far the most likely cause of an accident. Because all flights that don't end in an accident involve a landing, in this case we can make a rough guess that a pilot who can confidently and professionally land an airplane in all reasonable wind conditions reduces his chance of an accident by about 33 percent. This would seem to make landing practice a good investment!
The second highest number of pilot-related accidents involves takeoff and initial climb, coming in at 11 percent of GA accidents. However, while very few landing accidents were fatal, accidents during takeoff and initial climb also represented 11 percent of the fatal accidents. This is probably due to the fact that a quarter of these accidents involved departure stalls. All flights start with a takeoff, so we can estimate that a pilot who can successfully take off and establish an initial climb will reduce his chance of an accident by approximately 11 percent.
Coming in third place, at about 6 percent of GA accidents, is fuel management. There is actually some good news here. Accidents related to fuel management have been steadily declining during the past 10 years, from 159 in 1999 to only 73 in 2008. This is likely due to the influx of glass-cockpit airplanes that give the pilot a very accurate indication of how much fuel is on board and even provide range rings on moving map displays and automatic reminders to switch tanks. However, even the most advanced cockpit can't compensate for inadequate flight planning, failure to confirm how much fuel is on board before taking off, incorrect fuel system operation and failure to land for fuel when faced with unanticipated conditions. Since all of the general aviation flights analyzed by the Nall Report involved the use of fuel, we can estimate that a pilot who has a detailed understanding of his fuel system, carefully plans the fuel requirements for each flight, verifies the required fuel is on board and diverts for fuel when necessary will reduce his chance of an accident by around 6 percent.
Following the nebulous "Other" category, the three remaining causes of pilot-related accidents are so close (between 4 and 5.3 percent of all GA accidents) that the order is not important. However, just because these three causes combined represent only about 15 percent of all GA accidents does not mean they are not worthy of our attention. The sad fact is that these accidents account for almost half of GA fatalities.
• Half of the accidents that occurred during the descent and approach phases of flight were fatal. This included all 12 of the accidents resulting from improperly executed instrument approaches, and 12 of the 19 accidents that resulted in a stall or spin.
• Close to 70 percent of the maneuvering accidents were fatal. While some of these occurred during maneuvering in the pattern, most involved buzzing, low-altitude flight at night and aerobatics by untrained pilots, often in airplanes not approved for aerobatics.
• Fully 70 percent of all the weather-related accidents were fatal, including almost all of the VFR into IMC accidents, and all of the accidents attributed to poor execution of instrument procedures.
The Air Safety Foundation has discovered another tidbit hidden in the statistics. Researcher David Kinney reported that 75 percent of all general aviation accidents that occurred in fog between 1998 and 2007 were fatal. Even more interesting is the fact that 203 of those fatal accidents involved pilots who did not file a flight plan, while only 49 involved pilots who had filed a VFR flight plan. Pilots on IFR flight plans were in the middle, with 106 fatal accidents.
There is no way to know the specific connection between VFR flight plans and fatal accidents in fog. Perhaps the safer, more conservative pilots who follow the rules and avoid fog are more likely to file a VFR flight plan. But it is also possible that the very act of filing a VFR flight plan and the research required to do so lead a pilot to be more aware of the current and forecast weather conditions, and less likely to continue into adverse conditions. I would have to say that Kevin's decision to always file a VFR flight plan for any flights over 20 minutes is not only a good idea, but may also reduce his chance of a fatal accident.
In the final analysis, while there are many steps a pilot can take to enhance safety, ultimately it all boils down to the basics of flying. The first step in reducing the odds of being involved in an accident is a rigorous and brutally honest self-assessment:
How confident are you about your takeoff and landing skills? Do you find yourself dreading taking off or landing in anything other than perfect conditions? Do you have to make excuses to your passengers after every flight? Do you feel like you "cheated death again" after every landing? If the answer is yes, or if you even have to think about what your answer would be, you need to find a good instructor and work on your takeoffs and landings until a smooth, controlled departure and arrival become second nature even in a gusty crosswind. This would also be a good time to work on flying at minimum controllable airspeed.
Are you "a little weak" on your fuel system operation? Do you sometimes fail to visually verify how much fuel is on board before each flight? Do you take off without carefully assessing how much fuel will be necessary, or whether you will have at least an hour of reserve fuel? Are you tempted to keep going even when it becomes evident you will not have the planned fuel reserves at your destination? If the answer is yes, then it is time to get the POH out and perhaps get some help from someone who knows your fuel system inside and out. It would also be a good idea to write a contract with yourself agreeing to improve your flight planning, reserve fuel and divert policies.
For instrument-rated pilots, do you struggle through an instrument approach? Does the controller ask where you are going or if you would like vectors back to the final approach course? Or has it been so long since you flew in actual instrument conditions that you couldn't even answer those questions? If the answer is yes, then regardless of your legal currency, it is time to seek out an instructor and work on your instrument approaches and use of cockpit resources until you are right on the money every time.
If you don't have an instrument rating, have you gradually become complacent about flying in conditions below the legal minimums for VFR? Or has the initial excitement of flying worn off, so that you are now seeking to renew that thrill by flying low, buzzing or doing impromptu aerobatics? If you answer affirmatively to any of these questions, it is time for a serious assessment of the risks you are taking and the likely result of that risk-taking. You might consider asking a friend to be an accountability partner. After every flight, give your friend a call and discuss how the flight went, including the good, the bad and, most important of all, the unnecessary risks you took on that flight. The fact that you will need to tell your friend about anything stupid you did may be enough to hold you back when temptation strikes again.