United States of America

On July 10, 2021, about 1254 mountain standard time, a Beech C-90, turbo prop airplane, N3688P, was destroyed when it was involved in an accident near Wikieup, Arizona. The pilot and Air Tactical Group supervisor were fatally injured. The airplane was operated as a public use firefighting aircraft in support of the Bureau of Land Management conducting aerial reconnaissance and supervision. The airplane was on station for about 45 minutes over the area of the Cedar Basin fire. The ADS-B data showed the airplane had accomplished multiple orbits over the area of the fire about 2,500 ft above ground level (agl). The last ADS-B data point showed the airplane’s airspeed as 151 knots, its altitude about 2,300 ft agl, and in a descent, about 805 ft east southeast of the accident site. No distress call from the airplane was overheard on the radio. According to a witness, the airplane was observed in a steep dive towards the ground. The airplane impacted the side of a ridgeline in mountainous desert terrain. The main wreckage was mostly consumed by a post-crash fire. Debris was scattered over an area of several acres. Another witness observed the left wing falling to the ground after the aircraft had impacted the terrain. The left wing had separated outboard of the nacelle and was located about 0.79 miles northeast of the main wreckage and did not sustain thermal damage.

The pilot reported that he performed the “before starting engine” and “starting engine” checklists and everything was normal before taking off in the twin-engine airplane. He performed an engine runup and then started his takeoff roll. The pilot reported that about halfway down the runway the airplane was not accelerating as fast as it should. He attempted to rotate the airplane; however, “the airplane mushed off the runway.” The airplane settled back onto the runway, then exited the departure end of the runway, where it sustained substantial damage to the wings and fuselage. The airplane engine monitor data indicated the airplane’s engines were operating consistent with each other at takeoff power at the time of the accident. Density altitude at the time of the accident was 7,170 ft and according to performance charts, there was adequate runway for takeoff. The reason for the loss of performance could not be determined.

Transair flight 810, a Title 14 Code of Federal Regulations Part 121 cargo flight, experienced a partial loss of power involving the right engine shortly after takeoff and a water ditching in the
Pacific Ocean about 11.5 minutes later. This analysis summarizes the accident and evaluates (1) the right engine partial loss of power, (2) the captain's communications with air traffic control (ATC) and the first officer's left and right engine thrust reductions, (3) the first officer's misidentification of the affected engine and the captain's failure to verify the information, (4) checklist performance, and (5) survival factors. Maintenance was not a factor in this accident. The flight data recorder (FDR) showed that, when the initial thrust was set for takeoff, the engine pressure ratios (EPR) for the left and right engines were 2.00 and 1.97, respectively. Shortly after rotation, the cockpit voice recorder (CVR) recorded a “thud” and the sound of a low-frequency vibration. The captain (the pilot monitoring at the time) and the first officer (the pilot flying) reported that they heard a “whoosh” and a “pop,” respectively, at that time. As the airplane climbed through an altitude of about 390 ft while at an airspeed of 155 knots, the right EPR decreased to 1.43 during a 2-second period. The airplane then yawed to the right; the first officer countered the yaw with appropriate left rudder pedal inputs. The CVR showed that the captain and the first officer correctly determined that the No. 2 (right) engine had lost thrust within 5 seconds of hearing the thud sound. After moving the flaps to the UP position, the captain reduced thrust to maximum continuous thrust, causing the left EPR to decrease from 1.96 to 1.91 while the airplane was in a climb. (The right EPR remained at 1.43). The captain reported that he did not move the thrust levers again until after he became the pilot flying. The first officer stated that, after the airplane leveled off at an altitude of about 2,000 ft, he reduced thrust on both engines. FDR data showed that thrust was incrementally reduced to near flight idle (1.05 EPR on the left engine and then 1.09 EPR on the right engine) and that airspeed decreased from about 250 to 210 knots. (A decrease in airspeed to 210 knots was consistent with the operator’s simulator guide procedures for a single-engine failure after the takeoff decision speed [V1]. The simulator guide, which supplemented information in the company’s flight crew training manual, contained the most recent operator guidance for single-engine failure training at the time of the accident.) The captain was unaware of the first officer’s thrust changes because he was busy contacting the controller about the emergency. The captain told the controller, “we’ve lost an engine,” but he had declared the emergency to the controller twice before this point, as discussed later in this analysis. The captain instructed the first officer to maintain a target speed of 220 knots (which the captain thought would be “easy on the running engine”), a target altitude of 2,000 ft, and a target heading of 240°. (About 52 seconds earlier, the controller had issued the 240° heading instruction to another airplane on the same radio frequency.) About 3 minutes 14 seconds after the right engine loss of thrust occurred, the captain assumed control of the airplane; at that time, the airplane’s airspeed was 224 knots and heading was 242°, but the airplane’s altitude had decreased from about 2,100 ft (the maximum altitude that the airplane reached during the flight) to 1,690 ft. The captain increased the airplane’s pitch to 9°; the airplane’s altitude then increased to 1,878 ft, but the airspeed decreased to 196 knots. The captain subsequently stated, “let’s see what is the problem...which one...what's going on with the gauges,” and “who has the E-G-T [exhaust gas temperature]?” The first officer stated that the left engine was “gone” and “so we have number two” (the right engine), thus misidentifying the affected engine. The captain accepted the first officer’s assessment and did not take action to verify the information. Afterward, the EPR level on the right engine began to increase in response to the captain advancing the right thrust lever so that the airplane could maintain airspeed and altitude. Right EPR increased and decreased several times during the rest of the flight (coinciding with crew comments regarding the EGT on the right engine and low airspeed) while the left EPR remained near flight idle. The first officer asked the captain if they “should head back toward the airport” before the airplane traveled “too far away,” and the captain responded that the airplane would stay within 15 miles of the airport. During a postaccident interview, the captain stated that, because there was no fire and an engine “was running,” he intended to have the airplane climb to 2,000 ft and stay within 15 miles of the airport to avoid traffic and have time to address the engine issue. The captain also stated that he had been criticized by the company chief pilot for returning to the airport without completing the required abnormal checklist for a previous in-flight emergency. Although the captain’s decision resulted in the accident airplane flying farther away from the airport and farther over the ocean at night, the captain’s decision was reasonable for a single-engine failure event. The captain directed the first officer to begin the Engine Failure or Shutdown checklist and stated that he would continue handling the radios. The first officer began to read aloud the conditions for executing the Engine Failure or Shutdown checklist but then stopped to tell the captain that the right EGT was at the “red line” and that thrust should be reduced on the right engine. The captain then decided that the airplane should return to the airport and contacted the controller to request vectors. The flight crew continued to express concern about the right engine. The first officer stated, “just have to watch this though…the number two.” The captain asked the first officer to check the EGT for the right engine, and the first officer responded that it was “beyond max.” Afterward, the captain told the first officer to continue with the Engine Failure or Shutdown checklist and finish as much as possible. The first officer resumed reading aloud the conditions for performing the checklist but then stopped to state, “we have to fly the airplane though,” because the airplane was continuing to lose altitude and airspeed. The captain replied “okay.” As a result, the flight crew did not perform key steps of the checklist, including identifying, confirming, and shutting down the affected (right) engine. The first officer told the captain that the airplane was losing altitude; at that time, the airplane’s altitude was 592 ft, and its airspeed was 160 knots. The captain agreed to select flaps 1 (which the first officer had previously suggested likely because the airplane was slowing). The CVR then recorded the first enhanced ground proximity warning system (EGPWS) annunciation (500 ft above ground level); various EGPWS callouts and alerts continued to be annunciated through the remainder of the flight. The captain then told the controller that “we’ve lost number one [left] engine…there’s a chance we’re gonna lose the other engine too it’s running very hot….we’re pretty low on the speed it doesn't look good out here.” Also, the captain mentioned that the controller should notify the US Coast Guard (USCG) because he was anticipating a water ditching in the Pacific Ocean. Because of the high temperature readings on the right engine, the flight crew thought, at this point in the flight, that a dual-engine failure was imminent. During a postaccident interview, the captain stated that his priority at that time was figuring out how the airplane could stay in the air and return safely to the airport. The captain also stated that he attempted to resolve the airplane’s deteriorating energy state by advancing the right engine thrust lever. However, with the left engine remaining near flight idle, the right engine was not producing sufficient thrust to enable the airplane to maintain altitude or climb. The captain’s communication with the controller continued, and the first officer stated, “fly the airplane please.” The controller asked if the airport was in sight, and the captain then asked the first officer whether he could see the airport. The first officer responded “pull up we’re low” to the captain and “negative” to the controller; the captain was likely unable to respond to the controller because he was trying to control the airplane. The captain asked the first officer about the EGT for the right engine; the first officer replied “hot…way over.” The captain then asked about, and the controller responded by providing, the location of the closest airport. Afterward, the CVR recorded a sound similar to the stick shaker, which continued intermittently through the rest of the flight. The CVR then recorded sounds consistent with water impact. The airplane came down into the Pacific Ocean about two miles offshore and sank. Both crew members were rescued, one was slightly injured and a second was seriously injured. The wreckage was later recovered for investigation purposes.

The pilot was conducting a cross-country flight when, about 8 miles north of his intended destination, he reduced engine power, pitched for level flight, and waited for indicated airspeed to drop below 174 kts to add 20° of flaps. As soon as the drag was introduced, the airplane began to “buck back and forward,” and the two engines were “throttling up and down on their own.” He noted that the right engine seemed to be “sputtering and popping” more than the left engine, so he decided to raise the flaps and to shut down and feather the right engine. He declared an emergency to air traffic control. The pilot then noticed that the left engine was “slowly spooling down” and the airplane was not able to maintain airspeed and altitude. The pilot performed a forced landing to a flat, muddy wheat field about 4 nautical miles from the airport. The airplane sustained substantial damage to the fuselage and to both wings. A Federal Aviation Administration inspector traveled to the accident site to examine the airplane. Flight control and engine control continuity were confirmed. The master switch was turned on and the fuel gauges showed a zero indication. There was no evidence of fuel at the accident site or in the airplane. During the recovery of the airplane from the field, no fuel was found in the three intact fuel tanks, nor in any of the engine fuel lines. The pilot later stated that he ran the airplane out of fuel during the accident flight. The pilot reported that, during the preflight checks and twice during the accident flight, he activated the low fuel warning light, and no anomalies were noted. Postaccident testing of the low fuel warning light in an exemplar Piper Aerostar 602P revealed no anomalies.

The instrument-rated pilot of the business jet airplane, pilot-rated passenger, and five passengers departed on a cross-country flight and entered the clouds while performing a climbing right turn. The airplane then began to descend, and air traffic control (ATC) asked the pilot to confirm altitude and heading. The pilot did not respond. After a second query from ATC, the pilot acknowledged the instructions. The airplane entered a climbing right turn followed by a left turn. After ATC made several attempts to contact the pilot, the airplane entered a rapid descending left turn and impacted a shallow reservoir at a high rate of speed. Postaccident examination of the recovered wreckage and both engines revealed no evidence of any preimpact mechanical malfunctions or failures that would have precluded normal operation. Flight track data revealed that after takeoff, the airplane entered the clouds and made a series of heading changes, along with several climbs and descents, before it entered a steep, descending left turn. This type of maneuvering was consistent with the onset of a type of spatial disorientation known as somatogravic illusion. According to a National Transportation Safety Board performance study, accelerations associated with the airplane’s increasing airspeed were likely perceived by the pilot as the airplane pitching up although it was in a continuous descent. This occurred because the pilot was experiencing spatial disorientation and he likely did not effectively use his instrumentation during takeoff and climb. As a result of the pilot experiencing spatial disorientation, he likely experienced a high workload managing the flight profile, which would have had a further adverse effect on his performance. As such, the airplane entered a high acceleration, unusual attitude, descending left turn from which the pilot was not able to recover. The pilot and the pilot-rated passenger did not report any medication use or medical conditions to the Federal Aviation Administration on their recent and only medical certification examinations. Postaccident specimens were insufficient to evaluate the presence of any natural disease during autopsy. However, given the circumstances of this accident, it is unlikely that the pilot’s or pilot-rated passenger’s medical condition were factors in this accident. Low levels of ethanol were detected in the pilot’s muscle tissue and the pilot-rated passenger’s muscle and kidney tissue; n-butanol was also detected in the pilot’s muscle tissue. Given the length of time to recover the airplane occupants from the water and the circumstances of this accident, it is reasonable that some or all of the identified ethanol in the pilot and the pilot-rated passenger were from sources other than ingestion. Thus, the identified ethanol in the pilot and the pilot-rated passenger did not contribute to this accident.

The airplane departed Myrtle Beach International Airport (MYR), Myrtle Beach, South Carolina, at 1812, with the intended destination of Grand Strand Airport (CRE), North Myrtle Beach, South Carolina. According to automatic dependent surveillance-broadcast and air traffic control (ATC) communications information, the pilot established contact with ATC and reported that he was ready for departure from runway 18. He was instructed to fly runway heading, climb to 1,700 ft mean sea level (msl), and was cleared for takeoff. Once airborne, the controller instructed the pilot to turn left; however, the pilot stated that he needed to return to runway 18. The controller instructed the pilot to enter a right closed traffic pattern at 1,500 ft msl. As the airplane continued to turn to the downwind leg of the traffic pattern, it reached an altitude of about 1,000 ft mean sea level (msl). While on the downwind leg of the traffic pattern, the airplane descended to 450 ft msl, climbed to 700 ft msl, and then again descended to 475 ft msl before radar contact was lost. About 1 minute after the pilot requested to return to the runway, the controller asked if any assistance was required, to which the pilot replied, “yes, we’re in trouble.” There were no further radio communications from the pilot. The airplane crashed in a field and was destroyed by impact forces and a post crash fire. The pilot, sole on board, was killed.

A Cirrus SR22 and a Swearingen AS226TC were approaching to land on parallel runways and being controlled by different controllers on different control tower frequencies. The pilot of the Swearingen was established on an extended final approach for the left runway, while the pilot of the Cirrus was flying a right traffic pattern for the right runway. Data from an on-board recording device showed that the Cirrus’ airspeed on the base leg of the approach was more than 50 kts above the manufacturer’s recommended speed of 90 to 95 kts. As the Cirrus made the right turn from the base leg to the final approach, its flight path carried it through the extended centerline for the assigned runway (right), and into the extended centerline for the left runway where the collision occurred. At the time of the collision, the Cirrus had completed about ½ of the 90° turn from base to final and its trajectory would have taken it even further left of the final approach course for the left runway. The pilot of the Swearingen landed uneventfully; the pilot of the Cirrus deployed the airframe parachute system, and the airplane came to rest upright about 3 nautical miles from the airport. Both airplanes sustained substantial damage to their fuselage. During the approach sequence the controller working the Swearingen did not issue a traffic advisory to the pilot regarding the location of the Cirrus and the potential conflict. The issuance of traffic information during simultaneous parallel runway operations was required by Federal Aviation Administration Order JO 7110.65Y, which details air traffic control procedures and phraseology for use by persons providing air traffic control services. The controller working the Cirrus did issue a traffic advisory to the Cirrus pilot regarding the Swearingen on the parallel approach. Based on the available information, the pilot of the Cirrus utilized a much higher than recommended approach speed which increased the airplane’s radius of turn. The pilot then misjudged the airplane’s flight path, which resulted in the airplane flying through the assigned final approach course and into the path of the parallel runway. The controller did not issue a traffic advisory to the pilot of Swearingen regarding the location of the Cirrus. The two airplanes were on different tower frequencies and had the controller issued an advisory, the pilot of the Swearingen may have been able to identify the conflict and maneuver his airplane to avoid the collision.

The pilot, who was the owner of the airplane, and the pilot-rated passenger, whose maintenance facility had recently completed work on the airplane, departed on the second of two local flights on the day of the accident as requested by the pilot, since he had not flown the airplane recently. Flight track and engine monitor data indicated that, about 15 minutes after takeoff, fuel flow and engine exhaust gas temperature (EGT) values were consistent with a total loss of left engine power at an altitude about 2,500 ft. Engine power was fully restored about 4 minutes later. Between the time of the power loss and subsequent restoration, the airplane directly overflew an airport and was in the vicinity of a larger airport. It is likely that the left engine was intentionally shut down to practice one engine inoperative (OEI) procedures. Had the loss of power been unanticipated, the pilot would likely have initiated a landing at one of these airports in accordance with the airplane’s published emergency procedure, which was to land as soon as possible if engine power could not be restored; however, data indicated that engine power was restored, and the flight continued back to the departure airport. About 7.5 minutes later, about 6 nautical miles from the departure airport, engine data indicated a total loss of right engine power, followed almost immediately by a total loss of left engine power, at an altitude about 3,500 ft. A battery voltage perturbation consistent with starter engagement was recorded about 1 minute later, followed by a slight increase in left engine fuel flow; however, the data did not indicate that left engine power was fully restored during the remainder of the flight. The airplane continued in the direction of the departure airport as it descended and ultimately impacted a tree and terrain and came to rest upright. A witness saw the airplane flying toward her with the landing gear extended and stated that it appeared as though neither of the two propellers was turning. A doorbell security camera near the accident site captured the airplane as it passed overhead at low altitude. Sound spectrum analysis of the footage indicated that one engine was likely operating about 1,600 rpm while the other was operating at less than 1,000 rpm. The right propeller was found feathered at the accident site. An examination and test run of the right engine revealed no anomalies that would have precluded normal operation. The left propeller blades exhibited bending, twisting, and chordwise polishing consistent with the engine producing some power at the time of impact. Examination of the left engine and engine-driven fuel pump did not reveal any anomalies. Based on the available information, it is likely that the pilots were conducting practice OEI procedures and intentionally shut down the right engine. The loss of left engine power immediately after was likely the result of the pilot’s failure to properly identify and verify the “failed” engine before securing it, which resulted in an inadvertent shutdown of the left engine. Although partial left engine power was restored before the accident (as indicated by fuel flow values, damage to the left propeller, and sound spectrum analysis of security camera video), the left engine power available was inadequate to maintain altitude for reasons that could not be determined, and it is likely that the pilot was performing a forced landing when the accident occurred. It is also likely that the pilot’s decision to conduct intentional OEI flight at low altitude resulted in reduced time and altitude available for troubleshooting and restoration of engine power following the inadvertent shutdown of the left engine. The 67-year-old pilot was a Canadian national and had never applied for a Federal Aviation Administration medical certificate. According to the Transportation Safety Board of Canada, the pilot was issued a category 1 license with knowledge of a previous condition and knowledge of currently taking Xarelto (rivaroxabam). No acute or historical cardiovascular event was found on autopsy. Toxicology testing detected the sedating antihistamine cetirizine just below therapeutic levels in the pilot’s blood. A very low concentration of the narcotic pain medication codeine was detected in the pilot’s blood and urine; codeine’s metabolite morphine was also detected in his urine. The mood stabilizing medication lamotrigine was detected but not quantified in the pilot’s blood and urine. Thus, the pilot was taking some impairing medications and likely had a psychiatric condition that could impact decision-making and performance; however, given the circumstances of the accident, including the presence of the pilot-rated passenger to operate the airplane, the effects from the pilot’s use of cetirizine, codeine, and lamotrigine were not likely factors in this accident.

The pilot in command (PIC) and second-in-command (SIC) completed an uneventful positioning flight to pick up passengers and then continued to the destination airport. Cockpit voice recorder (CVR) information revealed that, while en route, the PIC expressed a desire to complete the flight as quickly as possible and arrive at the destination before another airplane that was also enroute to the destination airport, presumably to please the passengers. The PIC compared the flight with an automobile race, and the airplane’s overspeed warning annunciated multiple times during the descent. The flight crew elected to conduct a straight-in visual approach to land. Throughout the final approach, the airplane was high and fast, as evidenced by the SIC’s airspeed callouts. When the SIC asked whether s-turns should be made, and the PIC responded that such turns were not necessary. An electronic voice recorded by the CVR repeatedly provided “sink rate” and “pull up” warnings while the airplane was on final approach, providing indications to the crewmembers that the approach was unstable, but they continued the landing. The airplane touched down about 1,000 ft down the 4,200-ft-long runway. The PIC described that the airplane’s wheel brakes, thrust reversers, and ground air brakes did not function after touchdown, but witness and video evidence showed that the thrust reversers deployed shortly after touchdown. In addition, tire skid marks indicated that wheel braking occurred throughout the ground roll and increased heavily during the final 1,500 ft of the runway when the antiskid system activated. The ground air brakes did not deploy. The airplane overran the runway and came to rest about 400 ft past the departure end of the runway in marshy terrain. The fuselage and wings sustained substantial damage. The switch that controlled the automatic deployment of the ground air brake system was found in a position that should have allowed for their automatic deployment upon landing. There was no evidence to indicate a preaccident mechanical malfunction or failure with the hydraulic system, wheel brakes, thrust reversers, and weight-on-wheel switches, or electrical issues with either air brake switches. The airplane’s ground air brake deployment system logic required that both throttle levers be below 18° (throttle lever angle) in order to activate. The accident airplane’s throttle lever position microswitches were tested after the accident. The left throttle microswitch tested normal, but the right throttle microswitch produced an abnormal electrical current/resistance during initial testing. When the throttle was touched and then further manipulated by hand, the electrical resistance tested normal. The investigation was unable to determine whether the intermittent right throttle microswitch resistance prevented the ground air brakes from deploying because the testing was inconclusive. Landing performance calculations showed that, without ground air brakes, the landing ground roll exceeded the runway that was available from the airplane’s touchdown point about 1,000 ft down the runway. Mobile phone video evidence revealed that a quartering tailwind of about 10 to 15 knots persisted during the landing, which exceeded the manufacturer’s tailwind landing limitation of 10 knots for the airplane, and thus would have further increased the actual ground roll distance beyond that calculated. Throughout the final approach, the flight crew received several indications that the approach was unstable. The flight crew was aware that the airplane was approaching the runway high, fast, and at an abnormal sink rate. Both pilots had an opportunity to call for a go-around, which would have been the appropriate action. However, it is likely that the external pressures that the PIC and SIC accepted to complete the flight as quickly as possible influenced their decision-making in continuing the approach.

The pilot was flying a non precision approach in instrument meteorological conditions at night. While flying the procedure turn for the approach, the airplane’s speed decayed toward the stall speed before the airplane accelerated back to the standard approach speed. During the descent from the final approach fix, the airplane’s descent stopped for about 30 seconds and then the airplane descended at a rate of about 1,300 ft per minute. The airplane decelerated and continued to descend until the airspeed was about 85 knots (about 7 knots above the calculated stall speed for flaps 20°) and the altitude was 500 ft mean sea level. The last recorded data point showed the airplane about 460 ft mean sea level and 750 ft from the accident site. The airplane impacted a private residence, and a postcrash fire ensued and destroyed the airplane. Impact signatures were consistent with a low-energy impact. Examination of the airframe and engines did not detect any preimpact anomalies that would have precluded normal operations. Signatures on the engines and propellers were consistent with both engines providing power at impact. A review of the pilot’s toxicological information found that the level of eszopiclone in his specimens was subtherapeutic and thus not likely a factor in the accident. The circumstances of the accident are consistent with an inadvertent aerodynamic stall from which the pilot was unable to recover.