IAD03LA064

The pilot reported that he visually checked the fuel on the morning of the accident and noted that the left and right fuel tanks had about 55 gallons of fuel in each. He then flew about 1.5 hours and landed at an airport. When he departed from that airport to fly back to the origination point, he did not visually check the fuel prior to the flight. During the return flight he noticed that the fuel quantity gauge for the left fuel tank was "dropping at an excessive rate." As a precaution, he diverted to a nearby airport. At approximately 4,000 feet, both engines sputtered and subsequently lost power within a few seconds of each other. Realizing he could not make the airport, the pilot performed an off field landing into trees. At the time of the power loss, he remembered the right tank was indicating about 20 gallons, and the left tank was indicating about 10 gallons. No fire, defoliation of vegetation, or a significant amount of liquid fuel was noted at the accident site; however, a small amount of residual fuel, and an odor of fuel was noted. Due to the compromised fuel system that resulted from impact forces, the presence of fuel on board at the time of the accident could not be positively ascertained. Examination of the wreckage did not reveal any pre-impact mechanical malfunction with the engines, fuel strainers, or fuel selector valves. The right fuel selector valve was found between the OFF and X-FEED positions. The left fuel selector valve was found half way between the OFF and LEFT position; however, impact damage could have altered the position of the valves. Near simultaneous power loss can occur as a result of fuel exhaustion of both wing tanks while the fuel selector valves are in the normal position (each tank feeding each engine), or fuel exhaustion of one of the two fuel tanks while the fuel selector valves are on cross feed to both engines from the empty tank.

On July 11, 2003, about 1950 central daylight time, a Piper PA-34-220T, N2463M, registered to TK2 Bentwood Ltd., of Austin Texas, was substantially damaged when it impacted trees during a forced landing near Burnet, Texas. The forced landing was preceded by a total loss of engine power during a precautionary landing. The certificated commercial pilot and his two passengers were not injured. Visual meteorological conditions prevailed, and no flight plan had been filed. The personal flight departed from Addison Airport, near Dallas, and was en route to Georgetown. The flight was conducted under 14 Code of Federal Regulations Part 91.

In a telephone interview with the Safety Board, and in a written statement provided about 1 week after the accident, the pilot stated that he visually checked the fuel in the airplane prior to departing for a flight from Burnet to Addison on the morning of the accident. He noted that there was about 55 gallons of fuel in each wing tank. He further stated that the airplane had been "topped off" with fuel several days prior to the accident. He reported that the flight to Addison was uneventful and lasted about 1.5 hours. After a brief stay at Addison, he departed to fly back to Burnet.

The pilot stated that upon departure from Addison, the indicated fuel onboard was 35 gallons per side. A power setting of 75 percent was used in the climb to 6,500 feet, and 65 percent for cruise. The pilot recalled that the indicated fuel flow to both engines was the same at approximately 12 gallons per hour. In the vicinity of Waco, the pilot noticed that the left fuel gauge was showing a greater decline than the right. He noticed that the indicated fuel quantity on the gauge for the left fuel tank was "dropping at an excessive rate." At the time, the right gauge was showing slightly less than 30 gallons, and the left gauge was down to approximately 15 gallons.

The pilot stated that as a result of the fuel loss trend, he reduced power to 55 percent with an indicated fuel flow of slightly less than 10 gallons per hour. The pilot then elected to make a precautionary landing at Georgetown to try to determine whether the fuel on the left side had really declined that much or whether there was a problem with the gauge. At the time the decision was made to divert, the left gauge was showing approximately 10 gallons, and the right was indicating 25 gallons. According to the pilot, the airplane was about 12 nautical miles from the Georgetown airport at that time. The pilot stated that he considered "going to cross-feed" with the fuel selector controls, but determined that "there didn't seem to be a need."

The pilot reported that as he approached Georgetown, engine power was reduced, and a decent and setup for a straight in approach was initiated. Approaching pattern altitude, the airplane was leveled to slow to gear speed. The landing gear was extended and the throttles were advanced to stabilize the approach. The pilot reported that at that point, power was lost in both engines essentially simultaneously (within 10 seconds of each other) at an altitude of about 4,000 feet. The pilot stated that his reactions were to reach down to ensure the fuel selectors were in the proper position, and bring the mixtures to rich. The pilot reported that he put the landing gear back up and turned left toward a field for an off-airport landing. He stated that he then "stalled it into some trees" to minimize the impact. At the time of the power loss, he remembered that the right tank was indicating about 20 gallons, and the left tank was indicating about 10 gallons.

The pilot stated that he attempted to restart the engines, but they would not start. He also stated he had not noticed any problems with the fuel system during the nearly 50 hours that he had flown the airplane previously.

According to an inspector with the Federal Aviation Administration (FAA), both wings were partially sheared off from the fuselage. No evidence of fuel staining or defoliated vegetation was found at the accident site; however, an odor of aviation gasoline, and a residual amount of fuel in the airplane, was detected.

In a letter sent to the Safety Board 2 months after the accident, the pilot stated:

"The first and only indication of a potential problem was the divergence of the indication on the left fuel gauge from the right which prompted the precautionary diversion into Georgetown. At the time of the decision to divert into Georgetown, both gauges should have been indicating approximately 2.5 hours of fuel. The left gauge was showing about 1.0 hour, the right 2.5 to 3/0 hours.

"Engine stoppage was evidently a result of fuel exhaustion. For simultaneous exhaustion to occur, both engines would have to be feeding off the same system. At the time of stoppage, I confirmed that both selectors were in the "on" position. This leads to the probability that both engines were some how feeding from the left tank. That would explain the divergence of the left fuel gauge. At the time of stoppage, the left gauge was still showing some fuel but I have no knowledge of the accuracy of the gauge in that range of its operation. There was certainly 25 to 30 gals of fuel on the right side that was not getting to either engine despite proper positioning of the fuel selectors. All of this would suggest some type of malfunction of the fuel delivery system.

"To my knowledge, the fuel selector valves had been operated only once since the aircraft had been acquired in March of this year. Approximately a week before the accident flight I had demonstrated the operation of the fuel system to one of the owners. While the aircraft was on the ground, engines not running, the selectors were moved from the "on" through the "off " to the "crossfeed" back to the "on" position. The aircraft was not flown again until the July 11 flight.

"In the two years prior to acquisition by TK2, the airplane had flown little. On occasion I would find minor amounts of sediment when sumping the tanks. During one preflight, the crossfeed drains were found to be partially clogged. All of this suggests the possibility that the crossfeed valve some how became clogged or fouled, perhaps when they were operated on the ground. In any event, the result appears to be both engines pulling from the same tank."

According to technical information from Piper regarding the PA-34-220T "Seneca III" airplane, fuel is stored in fuel tanks located in each wing. The tanks in each wing are interconnected to function as a single tank. All tanks on each side are filled through a single filler in the outboard tank, and as fuel is consumed from the inboard tank, it is replenished by fuel from outboard.

Only 2 1/2 gallons of fuel in each wing are unusable, giving the Seneca III a total of 93 useable gallons with standard fuel tanks, or 123 useable gallons with the optional fuel tanks installed.

Before each flight, fuel must be drained from low points in the fuel system to ensure that any accumulation of moisture or sediment is removed from the system and to check for proper fuel. Fuel drains are provided for each fuel filter (2), each fuel tank (4), and each crossfeed line (2).

The fuel filter drains are located on the outboard underside of each engine nacelle; two fuel tank drains are located on the underside of each wing; fuel crossfeed drains are located at the lowest point in the fuel system, on the underside of the fuselage, just inboard of the trailing edge of the right wing flap.

Fuel management controls are located on the console between the front seats. There is a control lever for each of the engines, and each is placarded ON - OFF - X FEED. During normal operation, the levers are in the ON position, and each engine draws fuel from the tanks on the same side as the engine. The two fuel systems are interconnected by crossfeed lines. When the X FEED p