NTSB Factual Report

On February 23, 2011, at 1456 eastern standard time, a Cessna 172G, N3740L, was substantially damaged during a forced landing in Clarksville, Florida. The certificated commercial pilot and the passenger were not injured. Visual meteorological conditions prevailed. No flight plan had been filed for the flight, between Tri-County Airport (1J0), Bonifay, Florida, and Sandy Creek Airpark (75FL), Panama City, Florida. The aerial observation flight was conducted under the provisions of 14 Code of Federal Regulations Part 91.

According to the pilot, he first took off in the borrowed airplane from 75FL with half-filled fuel tanks, and about 1/2 hour later, landed at Bountstown Airport (F95), Bountstown, Florida. There, he completely filled both tanks, with the left tank taking perhaps slightly more fuel. He then took off again on a power line inspection flight, landing at 1J0 for a rest break. He subsequently took off once more to continue the power line inspection, and about 2 hours, 45 minutes of flight time after refueling, the engine lost power. The pilot switched the fuel tank selector valve from "Both" to the left tank, then to the right tank, with no improvement noted. He then applied carburetor heat, also with no improvement. After that, he pumped the primer knob, and was able to observe a couple of short bursts of power, but the airplane descended into trees, coming to rest nose-down in a near-vertical attitude. The pilot and the passenger subsequently released their seat belts and dropped to the ground without injury.

When asked about the fuel gauges, the pilot stated that during the accident flight, the fuel gauges had read differently, with the left tank gauge, when checked, indicating about 1/4 full while the right tank gauge was indicating off the full mark, perhaps around 3/4 full.

The pilot also stated that he didn't think much of it, since he had the fuel selector on Both, the gauges normally showed greater burn from the left tank, he had seen fuel sending units go bad, and he knew by timing that he had more than enough fuel onboard to complete the mission.

The pilot further stated that he did not run the airplane out of fuel. When he returned to the accident site with a Federal Aviation Administration (FAA) inspector the next day, there was a strong odor of fuel and there was fuel leaking from the right wing fuel cap vent.

When asked if the owner had mentioned any maintenance of the airplane prior to the accident, the pilot stated that the owner informed him that an O-ring in the fuel selector had been replaced.

Weather, recorded at an airport 20 nautical miles to the north, at 1453, included clear skies, calm winds, temperature 26 degrees C, dew point 8 degrees C, and an altimeter setting of 30.12 inches Hg. A carburetor icing probability chart did not predict any probability of carburetor icing.

On August 5, 2011, the engine and fuel system were examined at a recovery facility under NTSB oversight. According to the NTSB investigator's written observations, the airplane's wings and empennage had been removed for its transport from the accident site. The propeller and spinner remained attached to the engine, and the engine remained attached to the firewall with no observed damage to the engine mounts.

The right wing root fuel line opening was clear. The left wing root fuel line opening contained a small amount of organic material that appeared to be small, dried flower petals.

The gascolator was opened to check for contamination. When the bowl was dropped, a small steel ball, approximately 3/16 inch in diameter, fell out. Otherwise, the bowl was clean. No damage was noted to the ball. The fuel strainer brass screen was clean except for a very small amount of fibrous, lint-like material on the outside of the screen.

When the engine top spark plugs were removed, they appeared normal in color and wear when compared to a Champion inspection chart. They were then reinstalled and the engine was prepared for a test run on the airframe. The airplane battery was dead and a start cart was used to energize the starter. A portable plastic fuel tank was plumbed to the left wing root fuel line.

With external battery power applied directly to the starter, the engine was started and run at low power for about 50 seconds before being shut down due to no oil pressure indication. Subsequent turnover of the engine revealed positive oil pressure, consistent with an oil pressure indication problem.

The engine was then run a second time for about 3 minutes. The engine was run at low power initially; then power was increased to about 2,000 rpm. Magneto checks were normal.

The external fuel tank was subsequently plumbed to the right wing root fuel line with the fuel selector in Right position. The engine was started and ran at low power for about 40 seconds before it lost power and quit. Two additional attempts to start the engine with the fuel selector in the Both position were also unsuccessful.

A final attempt to run the engine was made, but with the fuel selector valve bypassed. The external fuel tank was plumbed directly to the gascolator inlet. The engine started immediately and ran smoothly for about 1 minute, up to 2,400 rpm.

The fuel tank finger screens were removed and did not reveal evidence of blockage. There was no fuel tank contamination observed using a flashlight to inspect the interior of the tanks.

The fuel selector valve was disassembled. There was some debris inside the valve, and a small steel ball was missing from the right-tank side of the selector. There were no internal failures evident that would have resulted in the ball liberating itself from the valve in flight.

According to a representative from the airplane manufacturer, as supplemented by the illustrated parts diagram for the fuel selector valve assembly, the fuel selector shaft included a cam at its lower end, which, when rotated, would normally press against one or two small balls, depending if the valve selected was for the right tank, the left tank or both tanks. For a selected fuel tank source, each small ball would push a spring-loaded larger ball away from an orifice, which would then allow fuel to flow from that source.

The smaller balls rested against the cam; if the cam was displaced upward, the smaller balls could fall into the main fuel selector chamber, then be passed through the single fuel feed line leading to the gascolator.

According to the airplane's owner, about 10 flight hours before the accident flight, he and a friend had fixed a fuel leak in the fuel selector by tightening the four retaining screws for the fuel valve cover/gasket. The owner also noted that the cover was under significant pressure against the tightening.

When asked if he had noticed any difference in fuel burn between the left and right tanks after the fix, the owner stated that he hadn't flown the airplane much, and that the accident pilot could better answer the question since he had been flying the airplane more often. Neither the owner nor the friend was an FAA-certificated mechanic.

During a follow-up interview with the accident pilot, he stated that he had smelled fuel in the airplane before the owner had advised him of the fix, and did not smell fuel after the fix. He also stated that he had flown the airplane several times after the fix, for about 8 hours, and had filled the fuel tanks before the day of the accident, but couldn't recall if one fuel tank took more fuel than the other.

The pilot further noted that during each preflight inspection, he would always check to see if the fuel selector would work by checking all positions, including Both and Off. Subsequent to the fix, he did not notice anything abnormal when doing so.

According to the airplane's maintenance records, new O-ring seals and a cover gasket were last installed in the fuel selector on July 16, 1992.

NTSB Probable Cause

A loss of engine power due to fuel starvation as a result of the owner's improper maintenance of the fuel selector valve, which resulted in a lack of fuel supply from the right fuel tank.