The Human Factor: Training Works!

The accident history of the MU-2 makes the case that training works. So why aren't we requiring similar training for other aircraft?




(December 2011) Three and a half years ago I wrote about the Mitsubishi MU-2 controversy ("Dangerous Airplanes or Dangerous Pilots?" — May 2008). For much of its life, the MU-2 had the unenviable status of having one of the worst turboprop accident rates and fatal accident rates. People wrote articles questioning the sanity of anyone who would fly such a dangerous airplane, and a congressman even introduced legislation that would require the airplane to be recalled, an option that was seriously considered by the FAA. However, there was another side to this story. First, there were several other turboprops that had equally bad or even slightly worse accident rates, and yet no one was writing nasty articles about them or saying they should be recalled. Even more curious, the MU-2 was only dangerous in the United States. In Europe it had an excellent safety record, and even here in the United States there were many people who loved the MU-2 and had operated it safely for many years.

The only difference between operating an MU-2 in Europe and flying one in the United States was that European countries required a type rating to fly the MU-2.

In the United States, even though the MU-2 is a very high performance airplane with complicated systems that has to be flown like a jet, because it weighs less than 12,500 pounds there was no requirement for a pilot to get any training or pass a check ride before flying one. Any multi-engine-rated pilot could jump in an MU-2 without even getting a checkout and legally take off. For a pilot used to flying a Cessna 402 or a Navajo, an airplane that climbs at 2,000 fpm, cruises at 300 knots at FL 300 and descends at 4,000 fpm is going to be more than a challenge. The MU-2 also has some operating features and characteristics that require a response opposite of what a piston twin pilot would do. This wouldn’t have been so bad if it had been hard to get your hands on an MU-2, but because of its bad reputation, an MU-2 was about the least expensive turboprop, with airplanes available for the price of a high performance single-engine airplane.

When you add it all up — a very high performance airplane with unique operating characteristics available at dirt-cheap prices with no training or check ride required — you have a recipe for disaster. In reality, almost none of the MU-2 accidents were directly related to the airplane’s design. It was the same old stuff: not using a checklist, taking off downwind, taking off without the proper flap setting, taking off or continuing into weather the pilot was not rated for or prepared to handle, flying with known deficiencies in critical systems, fatigue, etc. It just was happening more frequently in the MU-2.

The solution to this problem was obvious to many people. For years Mitsubishi had pleaded with the FAA to increase the training requirements for the MU-2 and require a type rating. Even though the MU-2 was no longer in production, in 1994 Mitsubishi initiated free Pilot Review of Proficiency (PROP) seminars every two years at locations around the world in an attempt to increase the knowledge and proficiency of MU-2 owners and operators. The FAA did a study and found that there was a growing awareness of the need for professional training to safely operate the MU-2, with many operators providing their own in-house training.

Systems and simulator training was also available from SimCom. This voluntary training initiative was having a positive impact, with the MU-2 accident rate reduced by half over a five-year period and with most of the continuing accidents involving pilots who had not had simulator training.

However, the FAA also discovered that there were inconsistencies between the various training programs, with many unique and unauthorized procedures being taught. In 2005 it established a flight standards board to evaluate how difficult it was to learn how to fly an MU-2. The FAA took three pilots with no MU-2 experience, including one helicopter pilot who had very little fixed-wing experience, and put them through a training program conducted by Pat Cannon, vice president of Turbine Aircraft Services Inc. in Addison, Texas, filming every lesson. All three pilots came through the training with exemplary performance. The FAA couldn’t even discern any difference between the fixed-wing pilots and the helicopter pilot.

Having established that any competent pilot can be trained to fly the MU-2, the FAA developed a standardized checklist and training curriculum based on its study and input from people who had experience operating or training in the MU-2. In 2006 it issued a special federal aviation regulation (SFAR) that, while still not requiring a type rating for the MU-2, actually went beyond a type rating by requiring specific training based on total and recent experience with the MU-2 and requiring recurrent training each year. The key to the SFAR was the requirement that every MU-2 pilot, no matter how much experience he or she had in the MU-2, had to go through one of two training programs. This established a common baseline of MU-2 knowledge and skills, and introduced standardized procedures and a checklist that is required to be used on every flight. Here are the specifics:

1. Any pilot with less than 50 hours as PIC in the MU-2 in the last two years or less than 500 total hours’ PIC experience in the MU-2 had to go through the full initial training program.

2. A pilot with more than the required PIC time listed above could take a requalification training program that is more intense than the recurrent training but does not cover as much as the initial program.

Once a pilot has completed either initial or requalification training, that pilot is required to complete the approved recurrent training course each year. The SFAR also requires differences training for those pilots who operate more than one model of MU-2, as well as specifying minimum qualifications for a pilot who wishes to fly an MU-2 and for flight and simulator instructors and check pilots wishing to provide MU-2 training or check rides. Takeoffs and landings in other multiengine airplanes do not meet the currency requirement for the MU-2, and a biennial flight review conducted in another airplane is not valid for operating an MU-2. Finally, the SFAR requires a functioning autopilot for single-pilot operation of the MU-2. The entire 140-page approved training program is included in the SFAR as an appendix and is required to be followed during training and on operational flights.

To determine the effect on the accident rate now that the SFAR has been in effect for five years, I went to Bob Breiling (, who has been analyzing turbine aircraft accident statistics for almost 50 years. He compared the three turboprops that are no longer in production, the MU-2, the Merlin and the Turbo Commander. Looking at the all-year accident rate since the airplanes were introduced, the MU-2 had the highest at 3.78 per 100,000 flight hours, followed closely by the Merlin at 3.53 and the Turbo Commander at 3.45.

Breiling then analyzed the accident rates for the five-year period the SFAR has been in effect and found that the MU-2 now has by far the lowest accident rate, at 0.75 accidents per 100,000 flight hours. That is half the Merlin rate of 1.48. The Turbo Commander came in last at 1.89. In fact, the MU-2 came in at half of the average accident rate for all business turboprops, which was 1.49. There have been only four MU-2 accidents during this five-year period, and only one of those was fatal. Three of the five years there were no MU-2 accidents at all. While some owners worried that the SFAR would further reduce the value of the MU-2, it has actually had the opposite effect of reversing its decline in value as people have recognized what a safe and effective airplane it is.

The effectiveness of the MU-2 SFAR leads to the obvious question of why we are not requiring more training for other turboprops that also have high accident rates. In fact, why stop there? If we were worried about the MU-2, with an accident rate of 3.78 per 100,000 flight hours, why aren’t we worried about the 2009 commercial fixed-wing accident rate of 3.01, or the noncommercial fixed-wing accident rate of 6.60? It is a question that begs further discussion, so I will continue with this subject next month.