NTSB Factual Report

HISTORY OF FLIGHT

On July 1, 2005, about 1855 Pacific daylight time, a Beech E-55 twin-engine airplane, N686T, experienced a loss of power from both engines and ditched in ocean waters near Lopez Island, Washington, about 8 miles southeast of the flight's planned destination of Friday Harbor, Washington. The airplane, which was registered to a private individual and operated by the commercial pilot, was destroyed by impact damage and immersion in salt water. Of the five people aboard, two passengers sustained fatal injuries, the pilot and one passenger sustained serious injuries, and one passenger sustained minor injuries. Visual meteorological conditions prevailed and no flight plan was filed for the personal cross country flight conducted under 14 CFR Part 91. The flight departed from Redmond, Oregon, about 1713.

In a written statement submitted to the NTSB investigator-in-charge (IIC), the pilot reported that on the afternoon of July 1, 2005, he and one of his passengers (who was a private pilot) pulled the airplane out of its hangar at Roberts Field in Redmond and conducted a pre-flight inspection. He explained that "the way to visually inspect the fuel level on this plane is to check the float gauges on the wing, which for the past eleven years of ownership, [have] been pin-point accurate. The left engine had 45 gallons (3 hours) and the right engine had 55 gallons (3:45 hours), and both engines had 11 qts. of oil. This was a sufficient amount of fuel for an hour and fifty minute flight." The pilot reported that the flight from Redmond to Friday Harbor proceeded uneventfully until they were "a few miles north of Seattle airspace." As he was waiting for Whidbey Island Naval Air Station Approach Control to respond to his request for an instrument flight rules (IFR) clearance to descend through an overcast cloud layer, the right engine started "surging." He asked Whidbey Approach to expedite his request, and they cleared him to descend through the cloud layer and gave him a vector to Friday Harbor. The pilot turned the auxiliary fuel pumps on low, and "the engine began to act somewhat normally." After descending through the clouds and with the airport in sight, he cancelled his IFR flight plan and changed to the common traffic advisory frequency (CTAF) for Friday Harbor. "Soon after," the right engine began surging again. "After several minutes," the left engine started surging. His attempts to remedy the situation were unsuccessful. The pilot stated that "the engines never shut down, but there wasn't sufficient thrust to keep us in the air, and the plane started to lose altitude." He realized they were descending below the elevation of the cliff top on Lopez Island and turned away from the cliff. "Within seconds of turning away from the cliff," the airplane impacted the water and "almost immediately began sinking."

According to information provided by air traffic control personnel, after departing from Redmond, the pilot contacted the Seattle Air Route Traffic Control Center (ARTCC) and requested and received visual flight rules (VFR) flight following. As the flight proceeded towards Friday Harbor, Seattle ARTCC handed the flight off to Seattle Terminal Radar Approach Control (TRACON). The pilot reported no difficulties with the airplane during the time the flight was in contact with Seattle ARTCC and Seattle TRACON. At 1839, Seattle TRACON handed the flight off to Whidbey Island Naval Air Station Approach Control. The pilot made a request to Whidbey Approach for an IFR clearance to Friday Harbor. Whidbey Approach issued an IFR clearance and subsequently pointed out the position of the Friday Harbor Airport to the pilot. At 1849, the pilot acknowledged that he had the airport in sight and cancelled the IFR clearance. Whidbey Approach then cleared the pilot to switch to the CTAF for the airport. During the time the pilot was communicating with Whidbey Approach, he reported having an engine problem, but when queried by the controller prior to the frequency change to CTAF, he reported the flight did not require any assistance. Whidbey approach did not receive any distress calls from the airplane; however, the pilot of another airplane reported to the controller that he had witnessed the airplane impact the water near the south end of Lopez Island.

According to information provided by the San Juan County Sheriff's Office, about 1855, emergency operators in Friday Harbor received multiple reports of the airplane "crash landing" in the waters just south of Lopez Island in an area known as Davis Bay. When emergency responders arrived on the scene, the airplane was completely submerged, and they found four people in the water, one of whom was unresponsive and was pronounced deceased by emergency medical personnel. One of the survivors told the responders that one passenger had not exited the airplane before it sank. About 2100, San Juan County Sheriff's Office divers entered the water and located the airplane approximately 150 feet from shore, at a depth of about 54 feet, latitude 48:26.690 north, and longitude 122:53.414 west.

PERSONNEL INFORMATION

The pilot, who was seated in the left front seat, held a commercial pilot certificate with airplane single engine land and instrument ratings. Additionally, he was rated as a private pilot in multi-engine land airplanes. He held a first class medical certificate dated September 11, 2003, with the limitation, must wear corrective lenses. The pilot reported that he had accumulated 290 hours total flight time with 4 hours flown in the past 90 days. Additionally, he reported that he had accumulated 150 hours in the Beech E-55 airplane.

The pilot rated passenger, who was seated in the right front seat, held a private pilot certificate with single and multi-engine land airplane ratings and the limitation, night flying prohibited. He held a third class medical certificate dated September 15, 2003, with the limitation, must have available glasses for near vision. On the application for his most recent medical certificate, the passenger reported that he had accumulated 450 hours total flight time with 25 hours flown in the past six months.

AIRCRAFT INFORMATION

The airplane's total fuel capacity was 172 gallons of which 166 gallons were usable. Each wing contained three fuel cells, an outboard leading edge cell, an inboard leading edge cell, and a wing box section fuel cell. The cells were interconnected to provide 83 gallons usable fuel in each wing. One flush-type filler cap was located in the outboard end of each outboard wing leading edge fuel cell. The Pilot's Operating Handbook for the airplane stated that "fuel quantity is measured by float type transmitter units which transmit the common level indication to a single indicator [on the instrument panel] for each respective wing system." The Handbook also stated that "a visual fuel level sight gage in each wing leading edge, outboard of the engine nacelle, can be used for partial filling or off-loading of fuel. This gage is to be used only when it reads within the calibrated areas." (Examination of the airplane's fuel sight gauges revealed the calibrated area to be between 40 and 60 gallons.)

Work orders documenting the most recent maintenance on the airplane were obtained from the facilities that performed the work. According to records obtained from one maintenance facility, the most recent annual inspection was completed on December 3, 2004, at an hour meter reading of 2,200.0 hours and a total airframe time of 3,753.5 hours. As of that date, the left engine, a Continental IO-520-C(7), S/N 816824-R, and the right engine, a Continental IO-520-C(7), S/N 816825-R, had both accumulated 431.8 hours since major overhaul. During examination of the wreckage, it was noted that the hour meter read 2,218.2 hours, indicating that when the accident occurred, the airplane had been flown 18.2 hours since the annual inspection.

According to records obtained from another maintenance facility, on March 15, 2005, fuel leaks in the airplane's left and right hand wings were repaired by replacing the right hand outboard and left hand inboard leading edge fuel cells. Further records concerning this work were found in the airplane. The records found in the airplane included a stick-on label stating the following: "14 March 2005 N686T HM:2202.7 Replaced RH OTBD leading edge and LH INBD leading edge fuel cells with new cells P/N 2F1-6-40514-12 S/N: 04-04372 and P/N 2F1-6-40866-13 S/N: 04-03443. Serviced with fuel - no leaks after 36 hours."

During a telephone interview conducted by the NTSB IIC, the owner of the airplane stated that on June 28, 2005, he had the airplane topped off with fuel at Roberts Field in Redmond and then flew from Redmond to Friday Harbor, flight time 1.6 hours. On June 29, 2005, he flew the airplane from Friday Harbor to Anacortes, Washington, and back, flight time 0.3 hours. On June 30, 2005, he again flew the airplane from Friday Harbor to Anacortes and back, flight time 0.3 hours. On July 1, 2005, he flew the airplane from Friday Harbor to Redmond, flight time 1.6 hours. The owner reported flying the airplane a total of 3.8 hours following the refueling on June 28 at Redmond. He stated that he did not check the fuel after landing at Redmond. He also stated that the airplane burned 15 gallons of fuel per hour per engine.

The owner said there were no discrepancies with the airplane on any of his flights from June 28 to July 1. He indicated that there had been no discrepancies with the airplane since the annual inspection other than the fuel leaks in March 2005. In March 2005, in preparation for a flight, he had the airplane topped off, and fuel poured out the bottom of the wings. In order to troubleshoot this problem, a maintenance facility removed the inboard and outboard fuel cells from each wing. The facility installed a new inboard tank in the left wing and a new outboard tank in the right wing. The other tanks were reinstalled. Following this repair, some of the fuel sumps did not drain. There are 4 sumps in each wing, and only 2 or 3 drained. Also, the inboard left sump was hanging down a couple of inches, which was different from its former position. Additionally, when he looked at the wing fuel sight gauges, they "didn't appear to be in the right position." The sight gauges were "slanted" and were "not at the 12 o'clock position." He asked a mechanic at another maintenance facility, the facility that had performed the airplane's annual inspection, to look at the airplane. The mechanic called him back in a few days and said that he had repositioned the sight gauges. The mechanic said he had put the gauges back to 12 o'clock. The mechanic also fixed the sump that was hanging down and the quick drains that were not working. The owner test flew the airplane following this work and reported that there were no discrepancies on the test flight.

When questioned specifically about the fuel gauges in the instrument panel, the owner stated that during the time he had owned the airplane, they had never worked properly. Several maintenance facilities had worked on them without resolving the problem. The fuel gauges would read full when the airplane was topped off, but after that, as fuel was burned they were never consistent. The right gauge showed empty even when the right tank was half full. However, the sight gauges in the wings worked "flawlessly." He and other family members who routinely flew the airplane (including the accident pilot) used the sight gauges to determine the amount of fuel aboard the airplane for partial fuel loads.

During a telephone interview conducted by the NTSB IIC, a mechanic employed by the facility that performed the fuel cell work in March 2005 reported that he replaced the right outboard tank and another mechanic replaced the left inboard tank. After the airplane was returned to the owner's hangar, the owner informed him that he was concerned that one of the wing fuel sight gauges was not reading correctly. The mechanic stated that he thought it was the right sight gauge that the owner was concerned about. In response to the owner's concern, he double checked to see that the gauge was installed properly. He did this by removing the access panels for the sight gauges on both wings so that he could see the marks that indicated the gauges were in the same position as when they were removed. He explained that when the sight gauges were removed to replace the tanks, he and the other mechanic marked the sight gauges "R" and "L" and made hash marks on the plates. When they reinstalled the sight gauges, they used the hash marks to put them back in their original positions.

During a telephone interview conducted by the NTSB IIC, a mechanic employed by the facility that performed the annual inspection on the airplane reported that he was asked by the owner to look at the airplane following the fuel cell work. He recalled that the owner was "not happy" with the work and complained specifically about the length of the inboard right fuel drain. He looked at the airplane and verified that the right fuel drain was "okay." When asked if he did any work on the fuel drains, fuel cells or fuel sight gauges, the mechanic answered "no."

The airplane's maintenance logbooks were not located during the investigation. According to the owner of the airplane, the maintenance logbooks were not in the airplane when he flew it on July 1, 2005. The owner reported that he specifically looked for the logbooks as he wanted to put the entry on the fuel cell work in the aircraft logbook. Requests for the logbooks were made by the NTSB IIC to the owner and to the two facilities that had performed work on the airplane in the past year; however, the maintenance logbooks were not provided to the NTSB IIC.

METEOROLOGICAL INFORMATION

At 1853, the reported weather conditions at Friday Harbor Airport, located approximately 8 miles northwest of the accident site, were wind from 220 degrees at 7 knots, visibility 10 statute miles, sky overcast at 4,400 feet above ground level (agl), temperature 15 degrees C, dew point 11 degrees C, and altimeter setting 30.05 inches Hg.

WRECKAGE AND IMPACT INFORMATION

The wreckage was recovered on July 16, 2005, and transported by barge to Skyline Marina in Anacortes, Washington, where it was examined by the NTSB IIC and a representative from the FAA. The airplane was recovered intact with the exception of the separated nose cone and portions of the left engine cowling. The cabin and baggage doors were open, and the windshield and left forward window were broken out. The bottom fuselage skin was crushed upwards and torn open in multiple locations. The left engine was broken from its two upper mounts and was hanging from the wing by its two lower mounts. The right engine was in place and remained attached to its mounts. The bottoms of both wing leading edges, between the fuselage and the engine nacelles, were buckled upward. The corresponding tops of the wing leading edges were not torn open or ballooned outward. The landing gear and the flaps were retracted. Both fuel selectors were found in the "ON" position, and both electric fuel boost pump switches were found in the "HI" position. The hour meter reading was 2,218.2 hours.

The wreckage was examined at the facilities of Av-Tech Services, LLC in Kent, Washington, on July 21, 2005, by the NTSB IIC and representatives of Raytheon Aircraft Company, Teledyne Continental Motors and the FAA. Examination of the left and right engines did not reveal any abnormalities that would have prevented normal operation and production of rated horsepower. The wings had been removed from the fuselage by recovery personnel. The fuel tanks were empty. The fuel sight gauges in the outboard end of each inboard leading edge fuel cell were examined. The gauges were marked to indicate quantities between 40 and 60 gallons with the area of the gauge face from less than 40 gallons to more than 60 gallons cross hatched. The right wing fuel sight gauge indicated 55 gallons, and the left wing fuel sight gauge indicated in the cross hatched area. When the fuel sight gauge

NTSB Probable Cause

The pilot's failure to refuel the airplane which resulted in a dual loss of engine power during normal descent due to fuel exhaustion. A contributing factor was the incorrect installation of the left and right wing fuel sight gage/float assemblies by unknown persons, resulting in a false indication on the wing fuel sight gages.