WPR17LA190

The pilot's incorrect positioning of the fuel tank selector valve during a tank switch, which resulted in fuel starvation and a loss of engine power. Contributing to the accident was the worn condition of the fuel selector valve handle position detents and markings.

The pilot was conducting a personal cross-country flight. The fuel selector was set to the left main fuel tank for takeoff and climbout. After reaching cruise altitude, he switched to the auxiliary tanks and later to the right main fuel tank. While in cruise flight, he also activated the two fuel pumps to transfer fuel from the tip tanks to the main tanks. When the airplane was about 4 miles from and 2,000 ft above the intended destination airport, the pilot switched to the left main fuel tank; the fuel-injected engine lost power, but the propeller continued to windmill. The pilot selected the landing gear down and attempted to restart the engine to no avail. The pilot determined that he would not make the runway and selected an open area as his landing target. He switched to the right main fuel tank, but the propeller continued only to windmill. The pilot then switched back to the left main tank and again could not start the engine. The airplane landed hard on flat desert terrain, which resulted in the nose landing gear collapsing and the fuselage buckling. The airplane slid upright to a stop. On-site examination of the fuel tanks revealed that they were not breached; the left main fuel tank contained about 23 gallons of fuel, the right main tank contained about 20 gallons, and the auxiliary tanks and the tip tanks did not contain any fuel. Examination of the fuel system did not reveal any obvious reason for the engine power loss. Although the left main tank fuel pickup screen was found separated from its line, the line appeared clear and functional. The ports and chambers of the fuel selector valve were clear, and the valve was functional. The fuel selector valve handle position detents were worn so that proper selection of a fuel tank by tactile method alone was difficult or impossible, and the placard was worn to the point where the position marker for the left tank was absent. The engine was removed for examination and a test run, and there was no evidence of a mechanical anomaly that would have precluded normal operation. Based on the pilot's reported sequence of events, the airplane examination, and the engine test run, it is likely that the pilot did not correctly position the fuel selector valve handle when he switched to the left tank for landing, which resulted in fuel flow interruption and power loss due to fuel starvation. The worn condition of the fuel selector valve made it susceptible to being mis-set, which would impede or terminate fuel flow to the engine. In addition, a fuel-injected engine can be more difficult and/or take longer to start after it is deprived of fuel. Due to the airplane's proximity to the ground, the pilot had limited time to troubleshoot or restart the engine.

The 0656 automated weather observation from Needles Airport (EED), Needles California, located about 18 miles northwest of HII, included calm wind, visibility 10 miles, few clouds at 10,000 ft, temperature 35oC, dew point 8oC, and an altimeter setting of 29.85 inches of mercury. The pilot held a private pilot certificate with an airplane single-engine land rating. The pilot reported that he had about 400 hours total flight experience, including about 145 hours in the accident airplane make and model. The pilot reported that he had successfully completed Bonanza Pilot Proficiency Program (BPPP) training through the American Bonanza Society. His most recent Federal Aviation Administration (FAA) third-class medical certificate was issued in July 2016, and his most recent flight review was completed in February 2017. The airplane came to rest upright on open, flat desert terrain about 1 mile south-southwest of the HII runway 32 threshold. None of the fuel tanks had been breached.. According to the recovery personnel, just after the accident, the left main tank contained 23 gallons of fuel, the right main tank contained 20 gallons of fuel, and the other four tanks were empty. Each tank was equipped with a filler neck and cap assembly, and all caps were found installed. Airframe and Engine Examination The airplane was not examined by NTSB personnel until after it had been recovered. The wings had been unbolted and removed for transport, and except for impact-related damage, the airplane was otherwise intact. When examined, the tachometer registered about 1,674 hours. The left main tank bladder material was slightly brittle, consistent with aging. The header tank flapper valve door moved freely, and appeared to be fully functional. The header tank bladder was cut open, revealing two layers of foam, each half the height of the tank. The top layer of foam was intact, but the lower layer of foam was deformed and wrapped/folded over on itself in the region of the finger screen, in a direction consistent with the rotation direction of the screen during installation. The foam had taken a set, consistent with it having been deformed for some undetermined but substantial duration. The finger screen was detached (broken solder joints) from the inlet/pickup line, and was trapped in the deformed foam. Although the finger screen no longer performed its function, the fuel inlet line did not appear to be obstructed. The right main tank bladder material appeared newer and more flexible than that of the left main bladder. The header tank was opened. Unlike the left tank, it was not foam-filled, and the pickup screen was intact and attached. The flapper valve door moved freely, and appeared to be fully functional. The fuel selector placard was worn to the point where the position marker for the left tank was absent. The fuel selector valve detents were less than positive/firm, consistent with wear, and there was moderate resistance in handle rotation. The handle was able to be rotated continuously in either direction. The valve port dimensions were such that a handle rotation arc of about 25° (about 1/16 of a full 360° rotation) would transition any port between fully open and fully closed. This arc equates to about 3/8 to 1/2 inch of travel of the selector handle end. According to FAA publication AC43-13-1B paragraph 8-34d, an "improper fuel valve position setting can seriously reduce engine power by restricting the available fuel flow. Check [the mechanism] for wear and excessive clearance which prevent the valve from positioning accurately." The engine-driven fuel pump (EDP) connection of the fuel line from the selector valve to the EDP was observed to have utilized Teflon tape on the male fitting that threaded into the EDP; the use of Teflon tape for this connection is prohibited, per FAA AC43-13-1B, paragraph 8-38f. Functional tests using low pressure shop air revealed that all fuel lines were properly routed and clear, and that the fuel selector functioned properly in all four positions. Subsequent disassembly of the fuel selector valve revealed that its design limited handle rotation to about 180o, through all 4 positions, but that the internal mechanism limiting handle rotation had failed, which permitted the handle to be rotated continuously in either direction. The failure did not appear to affect valve functionality. The electric fuel boost pump was located in the left wing wheel well. The pump was observed to be installed securely, with secure fittings and no evidence of any damage or fuel leakage. The pump was removed from the wing and tested via the application of 12vdc. The motor operated, and fuel was ejected from the outlet side of the pump. Visual examination of the engine indicated that it appeared to be intact. All cylinders were undamaged, and securely installed on their cylinder bays. The top spark plugs were removed and visually inspected; they displayed normal operating and wear signatures when compared to published guidance by Champion. Borescope examination of the upper chambers of the cylinders revealed that the cylinder bores, valve heads, and piston faces displayed normal operating and combustion signatures. The crankshaft was rotated manually, and continuity was established between the crankshaft, camshaft, connecting rods, and the associated components driven by the accessory gears on the aft face of the engine. During crankshaft rotation all cylinders displayed thumb compression and suction During manual crankshaft rotation, both magneto impulse couplings operated normally, and both magnetos produced sparks. The EDP remained attached to its installation point and was undamaged, and all fuel lines remained secured to the EDP. The throttle and metering assembly remained attached to its installation point and was undamaged. The throttle and mixture control arms remained secured to their shafts, and the control cable rod ends remained secured to the control arms. All fuel lines were secure. The fuel manifold valve remained attached to its installation point and was undamaged. There were no signs of fuel leaks or any other anomalies noted. The induction system displayed minor impact damage to the induction Y-pipe. The exhaust system remained attached to the engine and displayed minor impact damage. There were no signs of induction or exhaust leaks noted. The two blade, variable pitch propeller remained attached to the crankshaft. Both blades remained attached to the propeller hub and displayed impact damage. One blade displayed aft bending deformation and scratches in multiple directions on the propeller face, with chordwise scratches at the propeller tip. The other blade displayed minor aft bending deformation and chordwise scratches at the propeller blade tip. Engine Test Run Based on the examination results, the engine was removed from the airframe and shipped, without the propeller, to Continental Motors in Mobile for a test run attempt. The rear two mount legs were fractured. Some blue/green staining was noted on the engine oil sump. The magneto-to-engine timing for both magnetos was found to be set 1o earlier than the manufacturer's specified value. The engine was placed on a stand, and some baffling and accessories were removed for the test run The engine was prepared for operation by installing the appropriate thermocouples, pressure lines, and test pads for monitoring purposes. The engine started normally on the first attempt without hesitation or stumbling in observed rpm. The rpm was advanced in steps for warm-up in preparation for full power operation. The engine throttle was advanced to each of four different rpm values (1,200, 1,600, 2,450 and full) and held at that rpm for five minutes to stabilize. At each setting, the engine performed normally without any hesitation, stumbling, or interruption in power. The throttle was then rapidly advanced from idle to full throttle five times, and agai