NTSB Factual Report

The pilot stated that he performed a preflight inspection of the airplane which included a visual inspection of both fuel tanks. The level of fuel in each tank was ½ to ¾ inch below the fuel filler cap sealing surface. He did not report any discrepancies related to the engine start or engine run-up before takeoff. After departure from Danbury Municipal Airport (DXR), Danbury, Connecticut, the pilot slightly leaned the fuel to air ratio and climbed to 9,000 feet mean sea level (msl), where the pilot leaned the fuel to air ratio to the peak turbine inlet temperature (TIT) reading, then continued leaning lean of peak 30 degrees. The flight remained at 9,000 feet msl for a period of time then climbed to 12,000 feet msl where the pilot again leaned the fuel to air ratio lean of peak TIT. He verified that the cylinder head temperature (CHT) and exhaust gas temperature (EGT) readings were in the green arc, and monitored the fuel flow and other engine and aircraft gauges noting that the fuel appeared to be consumed evenly between the tanks.

The pilot further stated that when the flight was approximately 10 minutes from the destination airport (Grand Strand Airport (CRE), North Myrtle Beach, South Carolina), while flying at 5,000 feet, the engine surged and experienced a loss of power but did not quit. He increased the mixture control to full rich and turned the auxiliary fuel pump on, and informed the controller who provided a vector to the nearest airport. He continued trouble shooting but the engine eventually quit but the propeller continued rotating. Unable to land at the vectored airport, the pilot landed gear-up on grass short of the intended runway.

Inspection of the airplane before recovery by a Federal Aviation Administration (FAA) airworthiness inspector revealed a total of 3 to 4 ounces of fuel were drained from both fuel tanks which were not damaged. No fuel staining aft of either fuel filler cap or aft of either fuel tank sump drain valve were noted. After applying the aircraft’s battery, both fuel quantity indicators indicated empty.

While the pilot calculated that he had 50.0 gallons usable fuel for the accident flight, based on the fuel level noted in the tanks during his preflight inspection (1/2 to ¾ inch below the top of the filler neck), and the fact that he did not observe holes in the filler neck, each tank was less than full (25.0 gallons usable), but greater than 22 gallons usable.

The pilot has owned the airplane since 1998, and in 2003, it was modified in accordance with separate Supplemental Type Certificate (STC) by installation of a turbocharger and an engine-monitoring instrument with fuel flow indicating capability. Review of the maintenance records revealed the engine was overhauled on August 23, 2007, and installed into the airplane on October 4, 2007.

The pilot questioned whether the fuel flow transducer, or engine-monitoring instrument could have contributed to experiencing fuel exhaustion. Safety Board review of documents from the instrument manufacturer and postaccident examination of the instrument panel revealed two placards were not installed. One placard indicated not to use the instrument to determine the fuel level in the fuel tanks. The second placard indicated in summary that since the airplane was originally equipped with a primary fuel flow indicator, it should be used for reference and not the engine-monitoring instrument with fuel flow capability. The second placard conflicted with a flight manual supplement (FMS) associated with the turbocharger installation. The FMS indicated that the airplane’s primary fuel flow indicator may not accurately measure fuel flow after installation of the turbocharger.

Data from the installed engine-monitoring instrument was provided to Safety Board personnel for review. Data associated with all but 14 seconds of the accident flight (4.26 hours from first to last logged data rows) revealed the instrument indicated 44.5 gallons fuel used, and the calculated average fuel flow was 10.3 gallons-per-hour (gph). The pilot also questioned whether the engine overhaul could have contributed to experiencing fuel exhaustion. Data from two complete flights before the engine overhaul revealed the calculated average fuel flow were 10.1 and 10.2 gph, while data from two complete flights after the engine overhaul revealed the calculated average fuel flow were 9.6 and 9.7 gph.

The pilot explained that he had made the same trip (DXR to CRE) safely numerous times in the previous 8 years and at the completion there was 7 to 10 gallons of fuel remaining. He also reported he had flown into CRE from other destinations. Safety Board review of invoices from Ramp 66 located CRE, whose records go back to 2002, revealed only 4 invoices associated with the airplane. Three of the four invoices documented fuel purchases of 42.5, 45.0, and 31.8 gallons, respectively. The airplane’s total usable fuel capacity is 50.0 gallons when each tank is completely filled.

The pilot also questioned whether fuel pressure or fuel flow changes as a result of the turbocharger installation had any affect. Information provided by the STC holder revealed that the fuel flow at full throttle and rich mixture increases from the test specification maximum of 125.2 pounds-per-hour (pph) to 132 pph, an increase of 6.8 pph. Literature associated with the fuel flow transducer revealed the gph range is 0.6 to 60, or approximately 3.6 to 360 pph, and the working pressure is 200 pounds per square inch (psi).

NTSB Probable Cause

The pilot's failure to divert to an alternate airport resulting in the total loss of engine power due to fuel exhaustion. Contributing to the accident were the pilot's incorrect determination of the fuel load prior to the flight, the lack of placards near the engine-monitoring instrument, and the pilot's reliance of the on-board engine-monitoring instrument instead of the fuel quantity gauges.