Madera, CA, USA
N400MC
CRAIG W SHERMAN SUPER GLASAIR TT
The commercial pilot, who was also the builder of the experimental airplane, departed with full fuel for the planned 2.5-hour cross-country flight. The airplane was equipped with two independent means of determining fuel quantity: a capacitance-based fuel quantity indication system that provided a direct measurement of the fuel quantity and a fuel totalizer that provided a calculated value using a pilot-entered initial fuel quantity and a sensed fuel flow rate. The pilot reported that both quantity indication systems were reliable and normally registered similar values. While in cruise, after he had been airborne for about 1.5 hours, the pilot noted that each system indicated a remaining fuel quantity of about 38 gallons, which was ample fuel for the completion of the trip. A few minutes later, the engine lost all power. The pilot noticed that the totalizer indicated about 37 gallons remaining but that the capacitance system indicated minimal or no fuel remaining. The pilot determined that he would be unable to reach the nearest airport and set up for a forced landing on a road. Due to motor vehicle traffic, the pilot landed on the side of the road, and the airplane sustained substantial damage due to impact with a bordering fence and vegetation. Postaccident examination of the airplane revealed that a B-nut in the fuel system had backed off. The loose B-nut was just downstream of a fuel pump, which allowed fuel to leak overboard at a location that could not be seen from the cockpit by the pilot. Because of the leak location, once the leak began, the totalizer-indicated fuel quantity became erroneous, but the capacitance-based system quantity remained accurate. The evidence indicated that once the leak began, the fuel was depleted rapidly. The pilot reported that when he was building the airplane, he installed the B-nut, torqued it to prevent loosening, and marked it with a torque stripe. The airplane had been flown about 56 hours since the B-nut was installed. Although maintenance activity conducted to repair an oil leak near the B-nut could have resulted in inadvertent and undetected loosening of the B-nut, the investigation was unable to determine a specific cause for the loosening of the B-nut.
HISTORY OF FLIGHTOn June 8, 2014, about 1630 Pacific daylight time, an experimental amateur-built Super Glasair TT, N400MC, was substantially damaged during an off-airport forced landing near Madera, California, following a complete loss of engine power while in cruise. The owner-pilot was not injured. The personal flight was conducted under the provisions of Title 14 Code of Federal Regulations Part 91. Visual meteorological conditions prevailed, and no Federal Aviation Administration (FAA) flight plan was filed for the flight. According to the pilot, he was enroute from McClellan-Palomar airport (CRQ), Carlsbad, California to Reno/Stead airport (RTS), Reno, Nevada. The pilot was cruising at an altitude of 10,500 feet, and was not in communication with air traffic control. The airplane was equipped with a single 60 gallon capacity fuel tank, and two independent means to determine fuel quantity; a capacitance-type quantity gauging system, and a totalizer system. Just prior to the power loss, the pilot noted that both systems indicated a remaining fuel quantity of 38 gallons, and the totalizer displayed a fuel burn rate of 15 gallons per hour, which indicated that there was sufficient fuel to complete the flight. Until the power loss, the flight was uneventful, and all systems appeared to be operating normally. About 1.5 hours into the planned 2.5 hour flight, in rapid succession, the pilot noticed a fuel smell, and the engine lost power, although the propeller continued to windmill. The pilot then noted that although the totalizer indicated that 37 gallons of fuel remained, the capacitance system indicated that the fuel tank was empty. The pilot initially turned towards Fresno Yosemite International airport (FAT), Fresno, California, but then opted to land off airport on a road instead. He observed a road that was long and straight, and appeared to be free of powerlines and other obstructions, and set up for landing on that road. However, after the pilot had extended the landing gear and was "committed" to that landing selection, he noticed that there was more traffic than he was comfortable with, and he therefore opted to land on the verge of the road. Because the verge was unpaved, the landing gear separated from the airplane after touchdown. The right wing struck vegetation and a wire fence that bounded the verge, and the airplane came to a stop just off the right side of the road. PERSONNEL INFORMATIONFAA records indicated that the 59 year old pilot held an airline transport pilot certificate. On his most recent FAA medical certificate application, the pilot reported that he had approximately 16,783 total hours of flight experience. The pilot separately reported that he had approximately 795 hours in the accident airplane make and model. His most recent flight review was completed in November 2013, and his most recent FAA first-class medical certificate was issued in April 2014. AIRCRAFT INFORMATIONThe FAA issued the airworthiness certificate for the airplane in March 2014. The airplane was a low-monowing kitplane, of primarily composite construction. The airplane was equipped with a Lycoming TIO-580 series engine. The pilot, who was also the builder, reported that he conducted the bulk of the maintenance on the airplane. The airplane was equipped with a fuel totalizer unit, which displayed various fuel-related parameters in the cockpit. The totalizer display received some information from a fuel flow sensor that was situated serially in the fuel line to the engine, between a fuel pump and the engine. The sensor provided data to the totalizer, which then calculated fuel flow rate and fuel-remaining information based on the sensor data. The totalizer "fuel remaining" value was a derived parameter, calculated by the device by integrating fuel flow over time, and subtracting that value from a pilot-entered initial fuel quantity. The system was completely independent of the more typical capacitance-style fuel quantity indication system, which provided a more direct measure of fuel quantity. METEOROLOGICAL INFORMATIONThe 1453 automated weather observation at Madera Municipal airport (MAE), Madera, California, located about 10 miles west-northwest of the accident site, included winds from 290 degrees at 7 knots, visibility 10 miles, clear skies, temperature 38 degrees C, dew point 7 degrees C, and an altimeter setting of 29.69 inches of mercury. AIRPORT INFORMATIONThe FAA issued the airworthiness certificate for the airplane in March 2014. The airplane was a low-monowing kitplane, of primarily composite construction. The airplane was equipped with a Lycoming TIO-580 series engine. The pilot, who was also the builder, reported that he conducted the bulk of the maintenance on the airplane. The airplane was equipped with a fuel totalizer unit, which displayed various fuel-related parameters in the cockpit. The totalizer display received some information from a fuel flow sensor that was situated serially in the fuel line to the engine, between a fuel pump and the engine. The sensor provided data to the totalizer, which then calculated fuel flow rate and fuel-remaining information based on the sensor data. The totalizer "fuel remaining" value was a derived parameter, calculated by the device by integrating fuel flow over time, and subtracting that value from a pilot-entered initial fuel quantity. The system was completely independent of the more typical capacitance-style fuel quantity indication system, which provided a more direct measure of fuel quantity. WRECKAGE AND IMPACT INFORMATIONThe airplane came to rest upright, with all three landing gear fracture-separated from it. The fuselage and cabin remained intact. The right wing, and the left horizontal stabilizer and elevator sustained substantial damage. Shortly after the accident, the pilot and responding FAA personnel decowled the engine. They observed that a "B-nut" on the fuel line between the fuel pump and the fuel flow sensor that provided information to the totalizer had come loose, which provided a leak path that allowed the fuel to be pumped overboard. The units were forward of the firewall and positioned relatively low; the leaking fuel drained out the bottom of the engine cowl in a location that was not visible from the cockpit. ADDITIONAL INFORMATIONLeak Effect on Calculated Fuel Quantity Because the leak was upstream of (before) the fuel flow sensor, the sensor-provided data to the totalizer did not account for the leaked fuel, which resulted in an erroneous totalizer calculation of remaining fuel quantity. Once the leak began, the totalizer-calculated fuel quantity over-represented the actual remaining fuel quantity, and that discrepancy increased with time as the fuel continued to leak out. The pilot reported that the nominal fuel pressure downstream of the pump was about 60 pounds per square inch. Although the investigation was unable to determine the leak flow rate, and therefore the time required to deplete the fuel tank, the pilot's observations indicated that the fuel loss occurred over a very short time period. B-Nuts "B-nut" is a common term for a nut, often used in conjunction with a sleeve, is the normal hardware and method used to secure hoses or flared-type fluid tubing to a fitting on a device or another tube. Typical fluid applications include pneumatic, fuel, oil, and hydraulic systems. The B-nuts are available in both steel and aluminum, and the material selection is generally a function of the application and location in the aircraft. B-nuts have no explicit provisions for safetying; they rely on proper installation torque for security, and torque striping as a ready means to detect any post-installation rotation/loosening. Installation torque values are specified in FAA or specific manufacturer's documentation, and are a function of both the B-nut material type and the tubing diameter. The torque values for steel B-nuts are typically at least twice those for like-sized aluminum B-nuts. The pilot reported that he used steel fittings for all fuel lines, and that during construction/assembly, he torqued each fitting and then torque-striped it immediately thereafter, as opposed to torqueing multiple fittings, and then subsequently torque-striping them. The B-nut that came loose was steel, and was marked with a torque stripe. According to the pilot, the airplane had accumulated a total of about 56 hours in service, over a period of about 4 months, since the B-nut was installed. That B-nut had not been intentionally disturbed since its installation. The pilot also noted that subsequent to the installation of that B-nut, the airplane had developed a slight oil leak near the location of the B-nut, which resulted in multiple maintenance activities in that vicinity to locate, and then correct, the oil leak. He suggested that that maintenance activity could have resulted in the inadvertent and undetected disturbance of the B-nut. COMMUNICATIONSThe pilot did not utilize any air traffic control services during the flight.
The loosening of a B-nut in the fuel system which resulted in a rapid, undetected fuel leak and subsequent fuel exhaustion.
Source: NTSB Aviation Accident Database
Aviation Accidents App
In-Depth Access to Aviation Accident Reports
