Piper PA-61 Aerostar (Ted Smith 601)

While approaching Durango Airport on a flight from Celaya, the twin engine aircraft entered an uncontrolled descent and crashed in an open field located near the village of Ceballos, about 16 km northwest of the Durango Intl Airport. The burned wreckage was found near the Peña del Aguila Dam. Both occupants were killed.

The student pilot, who was not qualified to operate the airplane, reportedly flew from California to Pennsylvania on a commercial flight the morning of the accident to pick up and fly the accident airplane, which he purchased, to California. The student pilot departed Pennsylvania at 0719 and made several en route stops before arriving at the departure airport at 1949. A witness stated that the airplane’s right engine quit during taxi at the last en route stop and the pilot said he "cut it a little close on fuel." Another witness said that the pilot was “really tired” and planned to fly over the mountains for his return flight. The student pilot was not in communications with air traffic control while en route from the departure airport and did not receive an instrument flight rules clearance to operate the flight in class A airspace as required by Federal Aviation Regulations. The airplane was not equipped with automatic dependent surveillance-broadcast as required for flight in class A airspace. Radar track data indicate the airplane last departed from Fort Collins, Colorado, and maneuvered while climbing to 16,000 ft. The airplane proceeded west/southwest for a little over 40 miles before climbing to about 22,000 ft. The airplane then made several large heading changes and altitude changes between 20,000 ft and 23,000 ft before entering a tight looping turn to the left and losing altitude rapidly before track data was lost. All components of the airplane were distributed along the wreckage path in a manner consistent with a low-angle, high-speed impact with terrain. The airplane was destroyed. A green cylindrical tank consistent in color with an oxygen tank was separated from the airframe and was found along the wreckage path. Portions of pneumatic lines were attached to the tank and exhibited impact damage and separations from impact. Due to accident-related damage, the amount of the tank’s contents prior to the accident are unknown, and the functionality, if any, of the oxygen system is unknown. The reason for the airplane’s impact with terrain could not be determined based on available evidence.

The pilot was conducting an instrument landing system (ILS) approach in instrument meteorological conditions at the conclusion of a cross-country flight. The airplane had been cleared to land, but the tower controller canceled the landing clearance because the airplane appeared not to be established on the localizer as it approached the locator outer marker. The approach controller asked the pilot if he was having an issue with the airplane’s navigation indicator, and the pilot replied, “yup.” Rather than accept the controller’s suggestion to use approach surveillance radar (ASR) approach instead of the ILS approach, the pilot chose to fly the ILS approach again. The pilot was vectored again for the ILS approach, and the controller issued an approach clearance after he confirmed that the pilot was receiving localizer indications on the airplane’s navigation equipment. The airplane joined the localizer and proceeded toward the runway while descending. The pilot was instructed to contact the tower controller; shortly afterward, the airplane entered a left descending turn away from the localizer centerline. At that time, the airplane was about 3 nautical miles from the locator outer marker. The pilot then told the tower controller, “we’ve got a prob.” The tower controller told the pilot to climb and maintain 3,000 ft msl and to turn left to a heading of 180°. The pilot did not respond. During the final 5 seconds of recorded track data, the airplane’s descent rate increased rapidly from 1,500 to about 5,450 ft per minute. The airplane impacted terrain about 1 nm left of the localizer centerline in a left-wing-down and slightly nose down attitude at a groundspeed of about 90 knots. A postimpact fire ensued. Although the pilot was instrument rated, his recent instrument flight experience could not be determined with the available evidence for this investigation. Most of the fuselage, cockpit, and instrument panel was destroyed during the postimpact fire, but examination of the remaining wreckage revealed no anomalies. Acoustic analysis of audio sampled from doorbell security videos was consistent with the airplane's propellers rotating at a speed of 2,500 rpm before a sudden reduction in propeller speed to about 1,200 rpm about 2 seconds before impact. The airplane’s flightpath was consistent with the airplane’s avionics receiving a valid localizer signal during both instrument approaches. However, about 5 months before the accident, the pilot told the airplane’s current maintainer that the horizontal situation indicator (HSI) displayed erroneous heading indications. The maintainer reported that a replacement HSI was purchased and shipped directly to the pilot to be installed in the airplane; however, the available evidence for the investigation did not show whether the malfunctioning HSI was replaced before the flight. The HSI installed in the airplane at the time of the accident sustained significant thermal and fire damage, which prevented testing. During both ILS approaches, the pilot was cleared to maintain 3,000 ft mean sea level (msl) until the airplane was established on the localizer. During the second ILS approach, the airplane descended immediately, even though the airplane was below the lower limit of the glideslope. Although a descent to the glideslope intercept altitude (2,100 ft msl) would have been acceptable after joining the localizer, such a descent was not consistent with how the pilot flew the previous ILS approach, during which he maintained the assigned altitude of 3,000 ft msl until the airplane intercepted the glideslope. If the HSI provided erroneous heading information during the flight, it could have increased the pilot’s workload during the instrument approach and contributed to a breakdown in his instrument scan and his ability to recognize the airplane’s deviation left of course and descent below the glideslope; however, it is unknown if the pilot had replaced the HSI.

The mechanic who maintained the airplane reported that, on the morning of the accident, the right engine would not start due to water contamination in the fuel system. The commercial pilot and mechanic purged the fuel tanks, flushed the fuel system, and cleaned the left engine fuel injector nozzles. After the maintenance work, they completed engine ground runs for each engine with no anomalies noted. Subsequently, the pilot ordered new fuel from the local fixed-based operator to complete a maintenance test flight. The pilot stated that he completed a preflight inspection, followed by engine run-ups for each engine with no anomalies noted and then departed with one passenger onboard. Immediately after takeoff, the right engine stopped producing full power, and the airplane would not maintain altitude. No remaining runway was left to land, so the pilot conducted a forced landing to a field about 1 mile from the runway; the airplane landed hard and came to rest upright. Postaccident examination revealed no water contamination in the engines. Examination of the airplane revealed numerous instances of improper and inadequate maintenance of the engines and fuel system. The fuel system contained corrosion debris, and minimal fuel was found in the lines to the fuel servo. Although maintenance was conducted on the airplane on the morning of the accident, the right engine fuel injectors nozzles were not removed during the maintenance procedures; therefore, it is likely that the fuel flow volume was not measured. It is likely that the corrosion debris in the fuel system resulted when the water was recently purged from the fuel system. The contaminants were likely knocked loose during the subsequent engine runs and attempted takeoff, which subsequently blocked the fuel lines and starved the right engine of available fuel.

Before departing on the flight, the private pilot, who did not hold a current medical certificate, fueled the multiengine airplane and was seen shortly thereafter attempting to repair a fuel leak of unknown origin. The pilot did not hold a mechanic certificate and review of the maintenance logbooks revealed that the most recent annual inspection was completed 2 years before the accident. After performing undetermined maintenance to the airplane, the pilot reported to a witness that he had fixed the fuel leak. The pilot then taxied to the runway for takeoff. Witnesses reported that a large fuel stain was present on the ramp where the airplane had been parked; however, the amount of fuel that leaked from the airplane could not be determined. The pilot aborted the first takeoff shortly after becoming airborne. Although he did not state why he aborted the takeoff, he told the tower controller that he did not need assistance; shortly thereafter, he requested and was cleared for a second takeoff. During the initial climb, the pilot declared an emergency and was cleared to land on any runway. Witnesses reported that the airplane was between 400 ft and 800 ft above the ground in a left bank and appeared to be turning back to land on an intersecting runway. They thought the airplane was going to make it back to the runway, but the airplane's bank angle increased past 90° and the nose suddenly dropped; the airplane subsequently impacted terrain. One of the pilots likened the maneuver to a stall/spin, Vmc roll, or snap roll. Examination of the flight controls and engines did not reveal any anomalies that would have prevented normal operation. The position of the fuel valves was consistent with the fuel being shut off to the left engine. The fuel valves, with the exception of the left main valve, functioned when power was applied. The left main valve was intact, but the motor was found to operate intermittently. The amount of fuel found in the left engine injection servo was less than that in the right engine; however, the cylinder head temperatures and exhaust gas temperatures were consistent between both engines for the duration of the flight, and whether or to what extent the left engine may have experienced a loss of power could not be determined. The available evidence was insufficient to determine why the pilot declared an emergency and elected to return to the airport; however, the airplane's increased left bank and nose-down attitude just before impact is consistent with a loss of control.

The commercial pilot stated that, during the takeoff roll, the airplane swerved to the right, and he corrected to the left and aborted the takeoff; however, the airplane departed the left side of the runway and collided with an embankment. At the time of the accident, a quartering tailwind was present. The pilot had no previous experience in the accident airplane make and model or in any other multiengine airplane equipped with engines capable of producing 300 horsepower. During a postaccident conversation with a mechanic, the pilot stated that the airplane "got away from him" during the attempted takeoff. Because a postaccident examination of the airplane did not reveal any evidence of a preimpact mechanical malfunction or failure of the airplane's flight controls or nosewheel steering system that would have precluded normal operation and the pilot did not have any previous experience operating this make and model of airplane, it is likely that the pilot lost directional control during takeoff with a quartering tailwind.

The pilot reported that the purpose of the flight was to reposition the airplane to another airport for refuel. During preflight, he reported that the airplane's two fuel gauges read "low," but the supplemental electronic fuel totalizer displayed 55 total gallons. He further reported that it is not feasible to visual check the fuel quantity, because the fueling ports are located near the wingtips and the fuel quantity cannot be measured with any "external measuring device." According to the pilot, his planned flight was 20 minutes and the fuel quantity, as indicated by the fuel totalizer, was sufficient. The pilot reported that about 12 nautical miles from the destination airport, both engines began to "surge" and subsequently lost power. During the forced landing, the pilot deviated to land in grass between a highway, the airplane touched down hard, and the landing gear collapsed. The fuselage and both wings sustained substantial damage. The pilot reported no preaccident mechanical malfunctions or failures with the airplane that would have precluded normal operation. The pilot reported in the National Transportation Safety Board Pilot/ Operator Aircraft Accident Report that there was a "disparity" between the actual fuel quantity and the fuel quantity set in the electronic fuel totalizer. He further reported that a few days before the accident, he set the total fuel totalizer quantity to full after refueling, but in hindsight, he did not believe the fuel tanks were actually full because the wings may not have been level during the fueling. The "Preflight" chapter within the operating manual for the fuel totalizer in part states: "Digiflo-L is a fuel flow measuring system and NOT a quantity-sensing device. A visual inspection and positive determination of the usable fuel in the fuel tanks is a necessity. Therefore, it is imperative that the determined available usable fuel be manually entered into the system."

Following an uneventful flight from Bitburg, the pilot was cleared for an approach to Karlsruhe-Baden Baden Airport Runway 21. On final approach, while completing a sharp turn to the left to join the runway, the twin engine airplane stalled and crashed in an open field, bursting into flames. The pilot was seriously injured and the aircraft was destroyed by a post crash fire. The wreckage was found about 500 metres from the runway threshold and 350 metres to the left of the runway extended centerline.

Witnesses reported observing the airplane flying slowly toward the airport at a low altitude. The left engine was at a low rpm; "sputtering," "knocking," or making a "banging" noise; and trailing black smoke. One witness said that, as the airplane passed over his location, he saw the tail "kick" horizontally to the right and the airplane bank slightly left. The airplane subsequently collided with trees and impacted a field 1/2 mile north of the airport. Disassembly of the right engine revealed no anomalies, and signatures on the right propeller blades were consistent with power and rotation on impact. The left propeller was found feathered. Disassembly of the left engine revealed that the spark plugs were black and heavily carbonized, consistent with a rich fuel-air mixture; the exhaust tubing also exhibited dark sooting. The rubber boot that connected the intercooler to the fuel injector servo was found dislodged and partially sucked in toward the servo. The clamp used to secure the hose was loose but remained around the servo, the safety wire on the clamp was in place, and the clamp was not impact damaged or bent. The condition of the boot and the clamp were consistent with improper installation. The time since the last overhaul of the left engine was about 1,050 hours. The last 100-hour inspection occurred 3 months before the accident, and the airplane had been flown only 0.8 hour since then. It could not be determined when the rubber boot was improperly installed. Although the left engine had failed, the pilot should have been able to fly the airplane and maintain altitude on the operable right engine, particularly since he had appropriately feathered the left engine.