N914NKMelborn Joe B KOLB SPORT 600 · 2004

On October 9, 2016, about 1400 mountain daylight time, an experimental amateur-built Kolb Sport 600 airplane, N914NK, was substantially damaged during a forced landing in Roberts, Idaho, following a loss of engine power. The private pilot and his passenger received minor injuries. The airplane was owned and operated by the pilot under the provisions of Title 14 Code of Federal Regulations Part 91. Visual meteorological conditions prevailed and no flight plan was filed for the cross-country flight that departed a private airstrip at approximately 1400 mountain daylight time. The personal flight was destined for Joslin Field – Magic Valley Regional Airport (TWF), Twin Falls, Idaho.

The property owner reported that the pilot completed a high-speed taxi test and subsequently returned to the tie-down area at which time he told the property owner that he had encountered some vapor lock. He further communicated to the property owner that he felt comfortable with the engine's ability to develop power after the previous taxi test. The pilot had indicated to the property owner that he had experienced vapor lock during previous flights.

According to the pilot, he departed runway 24 after a successful pre-flight inspection and engine run-up. As the airplane reached approximately 100 ft above ground level (agl), the pilot perceived that the engine was losing power. He then lowered the nose and initiated a left turn teardrop maneuver to land on runway 06. The pilot engaged the airplane's auxiliary fuel pump in an effort to maintain engine power, but the propeller stopped rotating a few seconds later. Towards the end of the turn, the pilot leveled out early to land parallel to the runway, as he determined that the airplane would not reach the dirt landing strip. The airplane touched down about 200 ft south of the runway in sage brush, and subsequently impacted terrain before it nosed over and came to rest inverted, which resulted in substantial damage to the rudder. The pilot later recalled that he had discussed the possibility of vapor lock with the property owner but could not remember a previous encounter with this condition.

According to the owner of the airport property, the runway is a dirt strip about 2,200 ft long by 40 ft wide in a 24/06 runway configuration. A fly-in was held on the day of the accident with about 17 airplanes in attendance. At the conclusion of the event, the property owner walked out to the tie-down area with the accident pilot to see his airplane; the pilot and his wife then boarded the airplane. The airplane then lifted off the surface of runway 24 about 1,000 ft down the runway. During climbout the property owner observed the airplane veer to the right, and approximately 100 ft agl the airplane began a left turn, but did not appear to be accelerating. The pilot later told the property owner that the engine lost power during the turn and that he determined he would not be able to make it to runway 06. The airplane touched down on uneven terrain and flipped over inverted after it impacted a 4-foot-deep hole that contained debris. With assistance from his guests, the property owner turned the airplane over. Almost immediately, a blue colored liquid with an odor and appearance similar to 100 low lead (100 LL) aviation grade gasoline started to leak from the airplane.

The pilot held a private pilot certificate with ratings for single-engine land and single-engine sea. He did not possess a valid medical certificate at the time of the accident. According to the pilot, he had accumulated 10 total flight hours in the 30 days that preceded the accident, 500 flight hours in the airplane make and model, and 1,864 total flight hours in all aircraft.

Construction of the airplane was completed in 2004 by the airplane's previous owner and subsequently registered to the accident pilot on May 11, 2010. The airplane was powered by a Rotax 914, turbo-charged, direct drive, air/liquid cooled, 115 hp engine. The airplane's most recent conditional inspection was completed on February 2, 2016, at which time the airplane and engine had accumulated 521.3 total flight hours. The airplane's tachometer displayed 577 hours at the time of the accident.

Postaccident examination by a Federal Aviation Administration inspector revealed that the right wing fuel tank was approximately half full, and the left fuel tank was low. According to the pilot, the recovery team drained approximately 6 gallons of fuel from the right tank, and 1 gallon from the left before the airplane was moved. He stated that he purchased 14.1 gallons of 100 LL aviation grade gasoline at Jerome County Airport, Jerome Idaho, before he departed for Roberts. The pilot normally uses automobile gasoline.

The airplane was equipped with two composite wing fuel tanks, each with a capacity of 12 gallons (11.25 gallons usable). The dihedral angle of the wings is 0° with a 1° angle of incidence. Fuel quantity was gauged by fuel senders located in each wing connected to a fuel quantity indicator in the cockpit. Each of the three tanks are inter-vented and each wing tank was vented to the outside. The airplane was not equipped with a fuel selector valve. Fuel was fed by gravity down two fuel lines in the cabin to a 0.2 gallon aluminum header tank located behind the seat back. The fuel flowed to a fuel strainer, through a fuel shutoff valve, and then to a main boost pump and an auxiliary boost pump before it arrived at the engine. Both pumps were actuated by separated switches at the instrument panel; the main boost pump was located on the pilot's side and the auxiliary boost pump on the passenger's side. The design allowed fuel to be fed from both tanks simultaneously, and there was no provision for the pilot to make a fuel tank selection. The airplane was equipped with a fuel return line that routed unused fuel to the wing fuel tanks to prevent fuel vapor lock.

An examination of the fuel lines did not reveal any obstructions or preaccident breaches.

The airplane was equipped with a Garmin Area 500 GPS unit. The unit recorded multiple flights prior to the accident flight; however, the accident flight was not recorded. The unit data showed a flight from KJER to U56, which is about 10 nm east of Roberts and a total distance of 127 miles over 1.2 hours of flight time.

The 1353 recorded weather observation at Idaho Falls Regional Airport, Idaho Falls, Idaho, about 14 nm southeast of the accident site, included wind from 210o true at 15 knots with gusts to 22 knots, visibility 10 statute miles, clear skies, temperature 21o C, dew point 01o C, and an altimeter setting of 30.07 inches of mercury.

Construction of the airplane was completed in 2004 by the airplane's previous owner and subsequently registered to the accident pilot on May 11, 2010. The airplane was powered by a Rotax 914, turbo-charged, direct drive, air/liquid cooled, 115 hp engine. The airplane's most recent conditional inspection was completed on February 2, 2016, at which time the airplane and engine had accumulated 521.3 total flight hours. The airplane's tachometer displayed 577 hours at the time of the accident.

Postaccident examination by a Federal Aviation Administration inspector revealed that the right wing fuel tank was approximately half full, and the left fuel tank was low. According to the pilot, the recovery team drained approximately 6 gallons of fuel from the right tank, and 1 gallon from the left before the airplane was moved. He stated that he purchased 14.1 gallons of 100 LL aviation grade gasoline at Jerome County Airport, Jerome Idaho, before he departed for Roberts. The pilot normally uses automobile gasoline.

The airplane was equipped with two composite wing fuel tanks, each with a capacity of 12 gallons (11.25 gallons usable). The dihedral angle of the wings is 0° with a 1° angle of incidence. Fuel quantity was gauged by fuel senders located in each wing connected to a fuel quantity indicator in the cockpit. Each of the three tanks are inter-vented and each wing tank was vented to the outside. The airplane was not equipped with a fuel selector valve. Fuel was fed by gravity down two fuel lines in the cabin to a 0.2 gallon aluminum header tank located behind the seat back. The fuel flowed to a fuel strainer, through a fuel shutoff valve, and then to a main boost pump and an auxiliary boost pump before it arrived at the engine. Both pumps were actuated by separated switches at the instrument panel; the main boost pump was located on the pilot's side and the auxiliary boost pump on the passenger's side. The design allowed fuel to be fed from both tanks simultaneously, and there was no provision for the pilot to make a fuel tank selection. The airplane was equipped with a fuel return line that routed unused fuel to the wing fuel tanks to prevent fuel vapor lock.

An examination of the fuel lines did not reveal any obstructions or preaccident breaches.

The airplane was equipped with a Garmin Area 500 GPS unit. The unit recorded multiple flights prior to the accident flight; however, the accident flight was not recorded. The unit data showed a flight from KJER to U56, which is about 10 nm east of Roberts and a total distance of 127 miles over 1.2 hours of flight time.

Engine and Fuel System Examinations

The engine was examined at a recovery facility in Idaho by a representative of the engine manufacturer under the supervision of the NTSB Investigator-in-Charge. The inspection included an examination of the carburetors and two engine test runs.

According to the airplane's maintenance logbooks, both carburetors were removed on May 12, 2016, and sent to California Power Systems for overhaul. The logbook entry that captured the removal of the carburetors was recorded at a total tachometer time of 537.1 total flight hours, 38 flight hours before the accident flight. The entry from the carburetor overhaul facility stated that a "Rotax carb overhaul kit [was] installed as per Rotax Heavy Maintenance Manual." The entry further stated that the "Jet needle was moved to a leaner setting per customer request." During a follow-up conversation, the accident pilot reported that he requested the adjustment because he was operating the airplane at high field elevations. The component examination revealed that the needle jets to both carburetors were in the number 1 positions, the leanest setting. Both pistons exhibited no visual scoring, which indicated that the carburetors were properly balanced. Each carburetor main jet was free of obstructions.

A maintenance manual excerpt furnished by the engine manufacturer showed that carburetor needle clips should remain in the number 2 position.

According to a representative of the carburetor overhaul facility, the pilot's service facility purchased an overhaul kit that did not come with carburetor floats. When the components arrived at the overhaul facility, the units were disassembled and the symmetrical floats were flipped, which they stated was a standard industry practice. However, the engine manufacturer requires the installation of new floats instead of reinstalling them upside down. Both sets of carburetor floats were removed and weighed. The floats were placed in glasses filled with 100 LL aviation grade gasoline and allowed to soak overnight for a period of 12 hours, which is in accordance with the engine manufacturer's test procedure. At the end of the test period, the right carburetor floats' weight increased to 8.9 grams and the left carburetor floats' weight increased to 10.1 grams. It is presumed that any remaining fuel within the floats would have evaporated between the accident and the postaccident engine examination, a span of approximately 2 months. According to the Rotax maintenance manual, the floats are required to be weighed every 200 hours. Additionally, the guidance states that the carburetors must be removed and inspected every 200 hours.

In an effort to test engine operation, a fuel tank was plumbed directly into the header tank through the left wing fuel line at the wing root. Water was fed into the radiator to prevent the engine from overheating, as the radiator was damaged during the accident. Two complete engine tests were subsequently performed: the first was accomplished with the main fuel pump on, and the second was performed with both the main and auxiliary fuel pumps on. The throttle was advanced to full power approximately 5 times during these tests to capture the performance of the engine with only the main fuel pump engaged, and then with both the main and auxiliary fuel pumps on. The following engine parameters were documented during each test at idle power, maximum rated power, and maximum continuous power: fuel pressure, oil temperature, oil pressure, exhaust gas temperature, cylinder head temperature, and manifold pressure.

During the first test, the engine lost power for approximately 1 second at 2,500 rpm and 24 in. Hg manifold pressure when the throttle was advanced to full power. The results of the first complete engine test were within the ranges prescribed by the engine manufacturer.

Prior to the second complete engine test the throttle was advanced twice to the full forward position with only the main fuel pump on, and once with both the main and auxiliary fuel pumps on. During both tests, the engine lost power for about 1 second at 2,500 rpm and 24 in. Hg. manifold pressure. A 4th and 5th test were administered with the main fuel pump on and both the main and auxiliary fuel pumps on, respectively. During these final tests, the engine advanced to full power smoothly without any interruptions.

The results of the second complete engine test were within the ranges prescribed by the engine manufacturer.

Vapor Lock

The pilot could not recall a prior encounter with vapor lock before the day of the accident.

A representative of the engine manufacturer reported that the accident airplane type was susceptible to vapor lock, particularly in the presence of high ambient temperatures, which may increase the temperature under the cowling and lead to fuel system heat soak. Additionally, the representative stated that low fuel in either wing may allow air to be ingested, which may lead to vapor lock and fuel starvation.

According to the FAA Pilot's Handbook of Aeronautical Knowledge, vapor lock is defined as "a problem that…occurs when the liquid fuel changes state from liquid to gas while still in the fuel delivery system. This disrupts the operation of the fuel pump, which causes a loss of feed pressure to the carburetor or fuel injection system, and results in transient loss of power or complete stalling. Restarting the engine from this state may be difficult. The fuel can vaporize due to being heated by the engine, by the local climate or due to a lower boiling point at high altitude."

The FAA Airplane Flying Handbook defines vapor lock as "a condition in which air enters the fuel system and it may be difficult, or impossible, to restart the engine. Vapor lock may occur as a result of running a fuel tank completely dry, allowing air to enter the fuel system."

On the day of the accident the pilot was attempting to depart with the engine still hot from the previous flight.