ERA12LA164

The total loss of engine power due to fuel starvation for reasons that could not be determined because postaccident testing of the engine did not reveal any malfunctions or failures that would preclude normal operation. Contributing to the accident was the pilot’s improper in-flight fuel management following the loss of engine power.

The pilot stated that, during the preflight inspection, he did not visually inspect the fuel tanks or use an available dipstick to determine the fuel quantity. Rather, he relied on the fuel quantity gauges, which indicated the left wing tank had slightly more than 1/2 capacity or about 25 gallons, and the right wing tank had slightly more than 1/3 capacity or about 15 gallons. During the takeoff roll, with the fuel selector positioned to the right tank, the pilot noted the fuel flow reading was satisfactory. After takeoff, the pilot turned left to a west-southwesterly heading and applied rudder trim to center the ball of the turn coordinator, reporting that the flight was uncoordinated for about 10 seconds during the left turn. According to the airplane’s GPS and engine data monitor, while continuing in the west-southwesterly direction and climbing, the pilot reduced the fuel flow from about 41 gallons-per-hour (GPH) to 36 GPH. About 1 minute 18 seconds after takeoff, the fuel flow decreased to 16 GPH and was the same value at the next recorded data point 6 seconds later. The fuel flow then decreased to 0 GPH, increased to 31 GPH, and again decreased to 0 GPH. When the airplane was about 1,600 feet mean sea level, the engine lost power. The pilot turned both auxiliary fuel pump switches on for 2 seconds in an attempt to restore engine power but was not successful. He stated that he did not move the fuel selector. The pilot did not report performing any other actions to restore engine power. He declared an emergency and initiated a right descending turn toward the departure airport while the recorded fuel flow was about 1 GPH. After completion of the right descending turn while flying in a southeasterly direction, the fuel flow increased though engine power was not restored. The pilot turned to the right to maneuver the airplane for a forced landing and during that time the fuel flow again decreased, consistent with fuel starvation. While maneuvering, the left wing collided with a tree followed by the right wingtip contacting the ground. The pilot rolled the airplane to a wings-level attitude and it impacted the ground, resulting in damage to the right 1-gallon reservoir tank. The airplane slid about 110 yards before coming to rest upright. Witnesses and fire department personnel noted fuel leaking due to a cracked fuel line from the right reservoir tank. Any fuel remaining in the right main tank would leak due to this breach in the airplane’s gravity-fed fuel supply system. The amount of fuel leakage could not be determined. Twenty-nine gallons of fuel were drained from the left tank. According to Federal Aviation Administration documentation, the airplane’s engine was installed about 2 months before the accident, in accordance with the applicable supplemental type certificate (STC). During postaccident testing, the engine operated normally when configured to simulate the configuration of the airplane. Although the 41 GPH fuel flow recorded by the engine data monitor before the engine lost power is about 4 GPH greater than the maximum specified in the flight manual supplement pertaining to the STC, flight and ground-based testing of different airplanes equipped with the same engine model with the same engine limitations indicated that excessive fuel flow did not result in the loss of engine power. Therefore, it is unlikely that the excessive fuel flow recorded on the accident flight contributed to the loss of engine power. Review of maintenance records for the accident airplane indicated that it did not have a service kit that was announced via an airplane manufacturer service information letter, nor was it required to. The service kit made available the installation of fuel lines from each reservoir tank to each respective wing tank for excess fuel/vapor return. Because rapid fluctuating fuel flow did not occur during the accident flight, vapor lock is not considered to be a factor in the accident. Based on the available fuel flow data, the engine lost power most likely due to fuel starvation. The reason for the fuel starvation could not be determined because postaccident testing of the engine did not reveal any malfunctions or failures that would preclude normal operation. The pilot operating handbook (POH) cautions that fuel starvation can result if uncoordinated flight occurs for more than 1 minute with a fuel tank that is 1/4 full or less. Although the reported conditions of the accident flight (uncoordinated turn for about 10 seconds with a fuel tank about 1/3 full) do not correspond exactly to the POH guidance, the exact amount of fuel in the right tank could not be determined and it is possible that it was less than 1/3 full. The pilot’s initiation of the flight with the fuel selector positioned to a tank with a lesser quantity of fuel, as indicated by the fuel quantity gauge, and his failure to change the fuel selector following the loss of engine power contradicted procedures in the POH and Federal Aviation Administration-approved airplane flight manual. If the pilot had repositioned the fuel selector to the left tank following the power loss, he would likely have been able to restore engine power.

The airplane was equipped with an Aspen electronic flight display (EFD) 1000, a Garmin GPS Map 495 GPS receiver, and a JP Instruments (JPI) EDM 830; however the Aspen EFD does not record any data. The data from the EDM was provided by the pilot-in-command to the Federal Aviation Administration (FAA) and National Transportation Safety Board, while the Garmin GPS was sent to the NTSB Vehicle Recorder Division for readout. According to the NTSB Factual Report concerning the readout of the GPS, the accident flight from takeoff to accident was recorded as a track log. A plot of the track log tabular data was overlayed onto a Google earth Plot, and the GPS Factual Report with raw data is included in the NTSB Public Docket for this case. The JPI EDM recorded data from the accident flight begins after engine start with the first time stamp index point indicating the engine was operating at 986 rpm, the fuel flow was 3.4 GPH, and the fuel used was 44.3 gallons. The last time stamp index point indicates the rpm was 1, the fuel flow was 0 GPH, and the fuel used was 46.5 gallons. A total of 129 index points were recorded with an index point being recorded approximately every 6 seconds. The recorded data included fuel flow, rpm, manifold pressure, and calculated fuel used along with other engine parameters. Review of the data indicates that during takeoff power application, the fuel flow increased above 37 GPH, and remained above that value for 6 data points. During that time the manifold pressure was between 31.4 and 32.0. The highest recorded value of 41.1 GPH was noted at 2 of the data points, and then the fuel flow at the next data point 6 seconds later was 37.4 GPH. The fuel flow then decreased to about 36 GPH, and remained at that value for about 15 data points or approximately 1 minute 30 seconds. During this time the manifold pressure was between 30.7 and 31.4. The fuel flow decreased to about 16 GPH, with a corresponding decrease in engine rpm, then decreased to 0 GPH, with an engine rpm at that data point of 1,351 rpm. The fuel flow then increased to 31.0 GPH (the highest value after the decrease to 0 GPH), though the engine rpm at that data point was 1,149, and the manifold pressure was 29.1. The fuel flow again decreased to 0, with corresponding decrease of engine rpm, with the last recorded engine rpm of 1. No rapid fluctuation of fuel flow was noted. At the request of the NTSB, ground based testing was performed using N4789K, a 1979 Cessna P210N, S/N P2100320, owned and operated by the STC holder. The ground based testing was requested in order to determine the affect of excessive fuel flow on engine performance, and also to determine the affect different oil temperatures have on fuel flow, rpm, and manifold pressure. The testing was performed by the STC holder on November 11, 2012. The airplane utilized for the testing was equipped with the same engine and propeller as the accident airplane. Although the turbocharger installed on N4789K was different from the turbocharger installed on the accident airplane, the STC holder reported that they use the same operating engine parameters for the test airplane as the accident make and model airplane. With respect to the fuel supply system, the airplane used for the testing was similar to the accident airplane including the dimension of the fuel supply lines in the forward and aft door posts, and also pertaining to the excess fuel/vapor return lines to the reservoir tanks. The testing of N4789K was performed at the Pickaway County Memorial Airport (CYO), Circleville, Ohio, which has a surveyed field elevation of 685 feet, and was recorded by two cameras mounted in the cockpit. The engine was started about 1557 eastern standard time, and while at the run-up area the automated terminal information service (ATIS) from Ross County Airport (RZT), Chillicothe, Ohio, at 1559, indicated the temperature was 20 degrees Celsius (same as the ATIS for the accident flight), the dew point was minus 3 degrees Celsius, and the altimeter setting was 30.17 inches of Mercury. The RZT Airport is located about 5 nautical miles south-southwest of the CYO Airport. Although only ground based testing was requested using N4789K, the STC holder actually initiated takeoff and applied takeoff power when the oil temperature was indicating 129 degrees (within 4 degrees of the temperature during takeoff of the accident flight). With full power applied, the camera recorded exceedances for fuel flow (40.4 GPH), rpm (2,725), and manifold pressure (33.2 inches of Mercury), although no degraded engine performance was noted. The STC holder continued the takeoff remaining in the traffic pattern and climbed to 2,500 feet mean sea level (msl) during the downwind leg. The pilot returned for landing and landed uneventfully with the oil temperature indicating 161 degrees. He immediately returned to the approach end of the runway and initiated takeoff applying full throttle with the oil temperature indicating 166 degrees. With full power applied an exceedance of engine rpm was noted, while the maximum fuel flow was recorded to be 37.4 GPH and the manifold pressure was recorded to be 31.1 inches of Mercury; no degraded engine performance was noted. The pilot returned for an uneventful landing. Further ground based engine run testing was performed with FAA oversight of a 1979 Cessna T210N, N4888C, which was also modified by installation of the same make and model engine installed in the accident airplane. The testing was approved by the airplane owner, and was executed by the STC holder. The STC holder reported that they use the same operating engine parameters for the test airplane and accident make and model airplane. With respect to the fuel supply and excess fuel/vapor return systems, the airplane used for the testing was similar to the accident airplane with the exception of the dimension of the fuel supply lines in the forward and aft door posts. The dimension of the aluminum lines in the test airplane forward door post was 3/8 inch while the aluminum line in the aft door post was ½ inch. The testing was performed on April 3, 2013, also at Pickaway County Airport (CYO), Circleville, Ohio. The airplane was equipped with a JPI EDM, and two cameras were installed in the cockpit to record the readings of the JPI EDM. The weather at RZT Airport about the time of the testing indicated the temperature and dew point were 40 and minus 13 degrees Fahrenheit, respectively, and the altimeter setting was 30.40 inches of Mercury. Prior to the ground based test, the STC holder adjusted the manifold pressure and fuel flow to levels exceeding normal operating parameters to demonstrate the consequences or lack of consequences of excessive fuel flow and manifold pressure. The testing revealed that with the oil temperature at about 110 degrees Fahrenheit and all cylinders above 200 degrees F, full throttle application resulted in 34.9 inches manifold pressure and fuel flow reached a high of about 47 GPH. Observers outside said there was no smoke or indications the fuel flow was too high. The FAA inspector who witnessed the ground based testing reported that there were no indications to indicate engine failure due to excessive fuel flow. Civil Aeronautics Regulation (CAR) Part 3, titled “Airplane Airworthiness; Normal, Utility, and Acrobatic Categories” contains design criteria for certification of the airplane and its systems. Section 3.434, which pertains to the fuel flow rate of gravity fuel systems, indicates that the fuel flow rate for gravity systems (main and reserve supply) shall be 150 percent of the actual takeoff fuel consumption of the engine. The airplane’s type certificate data sheet indicates that the airplane’s takeoff horsepower rating is 310 limited to 5 minutes, and the maximum fuel limit is 186 PPH. Using the 150 percent criteria from section 3.434 of CAR 3, the airplane’s fuel supply system was designed to deliver