LAX99LA318

About 25 minutes into the flight over mountainous terrain, the pilot noticed that the left fuel cap was missing, and that fuel was escaping from the left tank filler opening. The pilot said he selected the right tank, but then returned the selector to the both position. He descended to make a precautionary landing at an airport ahead of the airplane and the engine quit during the en route descent. He landed in a field with the landing gear retracted and impacted a fence. An FAA inspector examined the airplane during recovery operations. As it was being lifted, he observed about 1 quart of fuel drain from the left wing tank. When the selector was positioned to both about 1 gallon drained from the belly in a continuous stream. According to the pilot and a mechanic who worked on the airplane before the flight, while at the departure airport the pilot had several electrical system problems that required several days to repair. During this time as electrical components were being replaced, engine starts became problematic. The mechanic pressurized the fuel tank with air through the left tank fuel filler opening and the engine started without difficulty. The mechanic suspected that the electric fuel pump was the cause of the problem and advised the pilot to have the pump examined. The pilot decided to continue with his flight plans instead. The pilot stated that he did not do his own preflight check of the airplane; he assumed that the fuel cap had been replaced. The mechanic reported that the pilot replaced the cowling and restarted the engine. Three witnesses watched the airplane taxi out and depart. No fuel caps were found on the ramp or the runway.

On September 23, 1999, at 1900 hours Pacific daylight time, a amateur-built Ciernia Glasair III, N153JC, experienced a loss of engine power while in cruise flight and force landed in a field in Snelling, California. The experimental airplane, owned and operated by the pilot, was substantially damaged. The commercial pilot was not injured. The personal cross-country flight, conducted under the provisions of 14 CFR Part 91, had originated from the Minden, Nevada, airport at 1835, and was en route to the San Luis Obispo, California, airport. Visual meteorological conditions prevailed and no flight plan was filed.

The pilot reported that he took off with 35.7 gallons of fuel and had the fuel selector set to "both." Upon climbing to cruise altitude over the Sierra Nevada Mountains, he noticed that the left fuel cap was missing. He reported that he could see some fuel escaping but it didn't appear to be flowing out very rapidly and he assumed that it was only "fuel vapor." He explained that initially he had been flying into the sun and because of that he could not see that the fuel cap was missing.

The pilot reported that he elected to initiate a descent and precautionary landing to Merced, where he planned to further assess the situation. He switched the fuel selector to the right tank, to see if that would affect the fuel being drawn from the left tank, and then switched it back to "both." He stated that he experienced a sudden engine failure approximately 15 miles north-northeast of Merced while at 1,100 feet agl. He turned the boost pump on but did not get any response from the engine. He stated that the aircraft had an "engine-out sink rate of 2,500 feet per minute" so he did not attempt an engine restart.

The pilot set up an approach to a gravel road. He opted to leave the landing gear retracted due to the rough terrain. As he descended, he noted that the road became a dead-end a few hundred yards ahead and was bordered by fence posts on each side. He landed in a field and slid approximately 50 yards before impacting a barbed-wire fence. The airplane ground looped before coming to a stop.

After the accident, the pilot reported that he had not understood what was happening in the fuel system at the time of the accident. He had not realized he was losing as much fuel as he was, and had assumed that he would have plenty of fuel to land at Merced. Additionally, he stated that, "I did not understand that the engine pump would draw air from the left tank before using the fuel in the right tank. Had I known this, I would have switched to the right tank setting . . . "

Further, the pilot reported that before he departed Minden, he had a vapor lock problem during engine start and had a mechanic pressurize the tanks through the left tank fuel cap opening. He stated that he assumed the fuel cap had been replaced on the left wing before takeoff; he did not do a preflight check of the airplane.

The day after the accident, a Federal Aviation Administration (FAA) inspector from the Fresno Flight Standards District Office responded to the accident site and interviewed the pilot. A copy of the interview is appended to this file. The pilot stated that there were two fuel gauges in the airplane, but he informed the inspector that he didn't trust the gauge that indicated the fuel quantity remaining; he relied instead on the gauge that he had programmed to indicate total fuel used and fuel remaining. When asked if the fuel gauge he trusted had any sensors that could accurately tell how much fuel was in the tanks, the pilot replied "no." The pilot reported that during the emergency descent, he had no indication of low fuel on the digital readout fuel indicator, and he did not look at the secondary fuel gauge, which should show fuel remaining. The inspector reported that he checked the fuel gauge that the pilot said he didn't trust, and the gauge indicated fuel remaining in both tanks.

The inspector further reported that as the aircraft was being hoisted up during the recovery process, he noted that approximately 1 quart of fuel came out of the left wing tank. The inspector changed the fuel selector position in the cockpit from "off" to "both" and observed that fuel came out from the belly of the aircraft in a continuous stream. He estimated that it was at least 1 gallon. He stated that he didn't note any fuel staining on the left wing of the aircraft around the fuel cap location.

In a written statement, the mechanic who worked on the airplane at Minden reported the sequence of events as he recalled them. He stated that on September 15, the pilot asked him to look at a problem with the airplane's alternator. The mechanic reported that he determined that the alternator needed to be replaced. He stated that the pilot took the alternator to a local parts store, tested the unit, and secured another. The mechanic installed the replacement alternator and performed a satisfactory run-up. On September 17, the pilot called the mechanic after realizing the airplane would not start. The mechanic reported that he removed a spark plug and found that it was dry, so he disconnected the fuel line. He stated that he turned the fuel pump on, but no fuel came out of the line. He applied air pressure to the right fuel tank with the fuel pump running and fuel flowed out of the line. The mechanic reported that he reinstalled the fuel line and performed a satisfactory run-up and leak check. He stated that the pilot restarted the engine and taxied for takeoff, but returned a few minutes later with an electrical problem. The mechanic determined that the alternator was again faulty.

The mechanic further reported that the pilot secured another alternator; bought a new starter, lower electronic spark plug leads, and two aircraft batteries. All components were installed the afternoon of the accident. The pilot again was not able to start the engine. The mechanic stated that he pressurized the left fuel cell with air and the engine fired. He stated that he suspected that the electric fuel pump was the cause of the problem since there was no change in noise or audio pitch until the air pressure was added. He advised the pilot to have the fuel pump further examined. The mechanic reported that since the fuel pump was difficult to reach, the pilot made the decision to continue with his flight plans.

The mechanic reported that the pilot replaced the cowling and restarted the engine with no noted problems. Three witnesses watched the airplane taxi out and depart. The mechanic further stated that no fuel caps were found on the ramp or the runway. He reported that the fuel caps were the flush type with half-moon, pull-up openers.