This past October, the NTSB issued its final report on the July 7, 2013, crash of a de Havilland DHC-3 single-engine Otter just after takeoff from the Soldotna, Alaska, airport. The Otter is a high-wing, tailwheel, single-engine airplane designed for heavy hauling. It was in production from 1951 to 1968, and has been popular among bush pilots in Alaska and elsewhere. The U.S. Army used it to haul personnel and supplies around Vietnam during the war. It originally was powered by a radial engine, but turboprop conversions were developed for increased payload and performance.
The mission during which the accident occurred was typical of those for which the Otter is ideally suited: moving hefty amounts of people and supplies short to moderate distances without the necessity for luxurious airfields at both ends. The commercial pilot and nine passengers on board the Part 135 on-demand charter flight were killed. The NTSB's investigation led to the conclusion that the turboprop taildragger had been overloaded, and the center of gravity (CG) was beyond the aft limit. As a result, when the airplane lifted off, it was in a tail-heavy mode, which worsened as it gained altitude and left ground effect. Within seconds, the airplane stalled and descended uncontrolled to the ground. The Safety Board's report included a Òwe told you soÓ moment by pointing out that in both 1990 and 2014, the NTSB urged the FAA to expand the requirements of Part 135.63(c) regarding weight and balance documents to cover not just multi-engine aircraft, but single-engine aircraft as well. Single-engine Part 135 operators still are exempt from the requirements to prepare a load manifest that includes items such as the number of passengers, total weight of the loaded aircraft, the maximum allowable takeoff weight, and center of gravity location. A copy of the load manifest is supposed to be carried on multi-engine aircraft, and the records need to be kept for at least 30 days.
In its Probable Cause statement for the Soldotna accident, the NTSB says that the agency's inaction contributed to the deaths of those 10 people in Alaska. To quote the NTSB, the probable cause of the accident was "the operator's failure to determine the actual cargo weight, leading to the loading and operation of the airplane outside of the weight and center of gravity limits contained in the airplane flight manual, which resulted in an aerodynamic stall. Contributing to the accident was the Federal Aviation Administration's failure to require weight and balance documentation for each flight in 14 Code of Federal Regulations Part 135 single-engine operations."
Figuring out weight and balance should be second nature for a pilot, especially if you're used to cramming as much into an aircraft as you safely can before taking off for an exotic hunting or fishing location in the Alaska wilderness. As students, we're supposed to have been taught how to do a loading problem by doing the math from scratch, in addition to knowing how to use any graphs or tables provided by the manufacturer. We're supposed to be able to demonstrate our skill at adding weights, figuring moments, and locating the limits of CG envelopes when taking FAA written exams and in the ground portion of biennial flight reviews. If we still can't do a loading problem on our own, the current availability of handheld calculators for aircraft loading, along with apps for doing weight and balance calculations on a smartphone or other device, make it easy. There's no longer an excuse for even the most arithmetic-challenged among us being unable to correctly determine whether our airplane is going to be within weight and balance limits for any given flight. I'm not going to provide a lesson in how to do it here. If you need a refresher, take a look at the FAA's Aircraft Weight and Balance Handbook FAA-H-8083-1A. It can be downloaded from https://www.faa.gov/regulations_policies/handbooks_manuals/aircraft/media/FAA-H-8083-1A.pdf
As an aside, although there's no specific requirement for those of us pilots operating under Part 91 to prepare weight and balance calculations before each flight, the FAA has a "gotcha" in Part 91.9. It requires the pilot in command to comply with the aircraft's operating limitations as defined in an aircraft's required flight manual. Typically, an Airplane Flight Manual (AFM) will have charts and/or graphs to help define the loading envelope. Since the AFM requires the airplane to be operated within its weight and balance envelope, you can only be sure that you're complying with the AFM's operating limitations by determining the weight and CG location for your flight. However, pilots frequently just jump in and go, confident that they're well below the gross weight limit and well inside the CG range.
The on-demand charter flight was to take the passengers from Soldotna to Bear Mountain Lodge, about 90 miles to the southwest. The airplane was one of six operated by the company, Rediske Air. The company was based in Nikiski, Alaska, and had a satellite base in Soldotna. The accident pilot was the company owner and its Director of Operations. The company employed four full-time pilots, one part-time pilot, and a part-time check airman. The accident pilot was 42 years old, had logged 7,765 flight hours, held a commercial certificate with single-engine, multi-engine and instrument ratings, was approved as a check airman for his company, and had received Part 135 competency and line checks less than two months before the accident.
The accident pilot began the day in Nikiski, where he had been scheduled to fly the DHC-3 on a different trip. When that trip was cancelled, he decided to use the Otter to transport the nine passengers from Soldotna to the Bear Mountain Lodge in one airplane instead of dividing them among two smaller aircraft as planned. The lodge operator had brought groceries and other supplies to the airport, and helped the pilot load them onto the DHC-3. The lodge operator estimated that the cargo weighed about 300 pounds. In reality, after the airplane had been fully loaded at Soldotna, the weight of the cargo and baggage in the rear came to more than 800 pounds. Investigators calculated that once the passengers were onboard with their carry-on items, the airplane exceeded its maximum gross weight of 8,000 pounds by at least 21 pounds. True, that's only a tiny amount of extra weight to be pulled around by the powerful turboprop engine. However, investigators calculated that the CG was dangerously aft at 157.7 inches, perhaps even reaching 161.0 inches, well beyond the 152.2 inches aft limit.
The airplane was manufactured in 1958. In 2010, before Rediske became the operator, it was overhauled and converted into a turboprop. Other STC modifications included installation of a four-bladed prop, modification with a STOL kit, increasing the aft cargo area limit to 750 pounds from the original 650 pounds, adding an extended range fuel system, and upgrading avionics.
The DHC-3 was equipped with three fuel tanks: forward, center and aft. In Nikiski, a helicopter pilot saw the accident pilot adding fuel to at least the forward and center tanks. When the helicopter pilot went to use the fuel pump, he noticed that it read 56 gallons. He said no one had used the pump between the time the accident pilot finished and he began. Varying amount of fuel carried in the tanks can help adjust aircraft balance, but the manufacturer requires that at least 20 gallons be in each tank to prevent air from entering the fuel system. The DHC-3's AFM contains weight and balance information specific to the fuel tanks, to help pilots ensure loading within the CG envelope. Investigators theorized that the pilot intended to top-up the tanks.
The positioning flight from Nikiski to Soldotna was uneventful. The NTSB's accident report suggested that this provided evidence that there were no mechanical problems with the airplane. Witnesses at the Soldotna Airport observed nothing out of the ordinary as the passengers and their baggage were loaded. They saw the airplane taxi from the ramp, but there were no witnesses to the accident. The DHC-3 taxied to runway 25 for departure. Runway 25 is 5,000 feet long by 130 feet wide and paved with asphalt. It has a 0.4% upslope, and there are 50-feet trees 2,265 feet from the departure end. The airplane struck the ground about 2,320 feet from the departure end of the runway, and about 154 feet to the right of the centerline. The nose and the right wing were down at impact. There was an extensive post crash fire.
The airplane didn't have a cockpit voice recorder, flight data recorder or image recorder. Investigators had precious little to go on, until they discovered an iPhone 5 belonging to the passenger who was seated at the fourth window on the left side of the airplane. When they retrieved its contents, they discovered that the passenger had recorded video of the taxi, takeoff roll, liftoff and climb. Because the iPhone was handheld during the video recording, the images were affected by the shakiness of the passenger as well as the vibrations and jostling of the airplane. To steady the images, the NTSB's Office of Research and Engineering applied aircraft window measurements and data from runway site surveys conducted by the FBI to create alignment points on each video frame showing the image of the window. This allowed them to precisely position the window as shown in each video frame, eliminating the random shakiness and creating measurable movement of the ground and objects seen through the airplane's window.
The useful part of the video ran just over 22 seconds. For the first 12 seconds, the airplane accelerated smoothly and the pitch angle decreased slightly as the tail lifted off. After remaining steady for about four seconds, the pitch angle began to increase. The pitch angle increased at a constant rate of just under three degrees per second, reaching a maximum of 30 degrees. The airplane's ground speed at liftoff was estimated as 68 mph. Within about 8 1/2 seconds, the airspeed decreased to about 44 mph. By about 11 seconds after takeoff, the airplane's speed and angle of attack were well within the range for an aerodynamic stall. The ground references disappeared from the video as the airplane entered a sharp roll to the right.
While the NTSB's investigators developed a reasonable set of facts and a logical scenario for this accident, the report did not address two items which make me wonder whether the pilot suspected he might have to deal with an aft CG. First, the video shows that the airplane's flaps were set in the landing position, fully extended, throughout the takeoff roll and climb. The NTSB didn't discuss possible reasons for the flaps not being at the proper takeoff setting. Was this pilot error, or could the pilot have mistakenly believed flap extension would somehow help overcome an aft CG? Second, the NTSB didn't explore possible reasons for the elevator trim being found about one and a half units forward of the recommended takeoff position. Was this a pre-takeoff setting error, or a conscious effort to compensate for a possible tail heavy condition? Consider that the pilot was no novice in Alaska flying, and had enough aviation experience to avoid taking aircraft weight and balance lightly. Of course, there's no way for the NTSB investigators to know whether the pilot really was aware of the dangerous loading condition. But, now that you and I have the benefit of this investigation, we can be sure we're always aware of why we need to remain within the weight and balance envelopes for every flight.
Peter Katz is editor and publisher of NTSB Reporter, an independent monthly update on aircraft accident investigations and other news concerning the National Transportation Safety Board. To subscribe, visit www.ntsbreporter.us or write to: NTSB Reporter, Subscription Dept., P.O. Box 831, White Plains, NY 10602-0831.