Probable Cause #23: Unnatural Selection

  • E-Mail this Article
  • View Printable Article
  • Text size:

    • A
    • A
    • A

This article originally appeared in IFR Refresher, Nov. 2005.

Probable Cause

The fuel systems on some piston-engine airplanes can be quite complicated and require a great deal of the pilot's attention during flight. Pilots must maintain an awareness of which fuel tanks the engine or engines are feeding from, as well as an idea of how much fuel is left in each tank. Unfortunately, it appears that the pilot of a Cessna 340 failed to do that as he was approaching Denver's Centennial Airport (KAPA) in March of 2002.

The pilot was approximately 3 nm from the runway threshold on the ILS when he reported an engine failure and shortly thereafter crashed. The impact killed the pilot and three passengers. Although the NTSB calculated that the aircraft should have had between 50 and 60 gallons of fuel on board, they determined that the left engine failed due to fuel starvation as a result of the pilot's improper use of the fuel system.


A Busy Day

Cessna 340

The pilot's day began at approximately 8 a.m. when he contacted the Denver Flight Service Station and filed three IFR flight plans that would take him 96 nm from KAPA to Aspen, Colo., then 42 nm to Gunnison, Colo., and finally 118 nm back to Denver.

Prior to leaving KAPA, the 50-gallon tip tanks, which are the main tanks on the C-340, were topped off and 20 gallons were pumped into each auxiliary tank. Based on this fuel order, the NTSB calculated that the plane departed with approximately 140 gallons on board.

The pilot flew to Aspen and picked up three passengers, who he then flew to Gunnison where the passengers owned a ranch. At approximately 3:45 p.m. they departed Gunnison for the return leg to KAPA. The pilot requested 17,000 feet as a cruising altitude and flight planned 42 minutes for the leg.

The airplane began its descent at 4:16 p.m. and was told by the Denver Approach controller seven minutes later that he could expect light to moderate icing below 8,000 feet. The controller said he would keep the aircraft above that altitude as long as possible, which the pilot acknowledged.

The pilot continued descending per the controller's instructions and at 4:27 p.m. the controller told the C-340 pilot, "N1DM is seven miles from CASSE [the final approach fix], turn left heading 020, maintain 8,500 until established on the localizer, cleared for the ILS 35R approach [at] Centennial."

The pilot acknowledged the instructions, then the controller instructed the pilot to "maintain at or above 8,000 until established on the localizer. Contact Centennial Tower on 118.9." That was also acknowledged and nothing seemed to be amiss at that point in the flight. At 4:28 p.m. contact was established between the pilot and Centennial Tower and the controller cleared the aircraft to land.

The first hint of trouble came when the C-340 was about 3 nm from the runway. Radar data indicates that the plane began a left turn as it was descending through 7,200 feet (The airport elevation is 5,883 feet). Shortly thereafter, the pilot radioed that he had lost an engine, followed by an "Oh, [expletive]!" That was the last transmission from the aircraft. Radar data shows that the plane had turned 180 degrees at a rate of six degrees per second before radar contact was lost.

The aircraft crashed approximately 3.5 miles south of the approach end of the runway on the left side of the localizer.

Impact

Crash Site

One witness was driving southbound on Interstate 25 when he saw the airplane flying southbound on his left side. He said the airplane was 150 to 200 feet above the ground at that time and that there was no "tipping or rolling" of the wings. The airplane then attempted to clear a hill that was in its path, now flying at about 50 feet above the ground. "We then saw the plane tip 90 degrees to the left (right wing straight up)," the witness told the NTSB. The plane then descended into the ground behind a hill.

Another witness was also driving southbound on Interstate 25. He told investigators that the airplane was flying "very low. It appeared to gain a slight amount of altitude before banking sharply to the left and nose diving into the ground just over the crest of the hill." He exited the highway and ran to the crash site. He said he saw no movement in the aircraft and was afraid to approach it too closely because of "massive fuel fumes" coming from the wreck.

Since the pilot had reported an engine failure just before the crash, investigators placed their focus on the engines and the aircraft itself to determine what might have caused the event, but no mechanical anomalies were found. It was determined that the landing gear was in the extended position, the flaps were extended approximately 10 degrees, neither propeller was feathered and that, based on damage to the propeller assemblies, the right engine was producing more power than the left.

The investigation also focused heavily on the fuel system. Investigators found that the left fuel selector handle was set to the right main tank position, but the actual fuel valve was found in the off position. The fuel selectors and fuel valves are connected and actuated by a 9.6-foot-long cable. According to the NTSB report, "any tensional force applied to this cable, such as during an accident sequence, will pull fuel valves to their off position and the fuel selectors to the cross feed (other wing main tip tank) position." Investigators extended their focus by carefully studying the fuel system's O-ring seals. Based on their wear and damage, the NTSB determined that the left fuel selector had been set to either the left auxiliary fuel tank or the off position before the accident occurred.

This is in contrast to the procedure in the Owner's Manual, which states that the fuel selector handles should be set to the main tanks during takeoffs, landings and emergency operations.

No Novice At The Helm

The pilot appeared to be well qualified for the flight. According to his personal logbook he began flying in 1966 and received his ATP certificate in March 1989. He had 3,563 total hours in flight, of which 2,580 hours were in multiengine aircraft. He had approximately 560 hours in the Cessna 340 and he had flown approximately 239 hours in the year prior to the accident. Both his medical and flight review were current.

In December 2000 he had attended Flight Safety's Cessna 340 training program and in September 2001 he completed their Cessna 441 program. He had flown the accident airplane on three occasions in January 2002 and flew a Cessna 441 five times in the same month. He was on vacation for most of the month of February, but flew the Cessna 340 five days before the accident flight.

The weather at Centennial Airport around the time of the accident was reported as winds from 050 degrees at 10 kts, visibility five miles, broken clouds at 900 feet and overcast clouds at 1,500 feet. The temperature was 25 degrees Fahrenheit and the dew point was 21 degrees Fahrenheit.

A King Air pilot who flew the ILS approach about 10 to 15 minutes behind the Cessna 340 said that she was in solid instrument conditions from 11,000 feet down to 6,500 feet or about 600 feet above the ground. She said she had all available anti-ice equipment turned on and experienced light to moderate icing conditions on the approach. She reported the visibility under the clouds as three to four miles.

A pilot who landed about 30 minutes after the accident said he picked up rime ice on his trip from Gunnison to Centennial, but that it dissipated. He also said he picked up light to moderate ice on his approach into the airport.

The accident pilot never made any mention of icing while he was on the approach.

A Salvageable Problem?

Cessna 340 Panel

So, what caused this airplane to crash when it was reasonably well positioned to complete the ILS approach, even with an engine not working? The NTSBs conclusions would indicate that the accident sequence started with the pilot's failure to select the main fuel tank for landing and the subsequent draining of the auxiliary tank. Or, he may have improperly selected the off position on the fuel selector, halting the fuel flow to the left engine.

The fuel system in twin Cessnas fitted with tip tanks is complex and requires a thorough knowledge and understanding of it. The tip tanks are also the main fuel tanks and fuel is pumped directly from them to the engines. Each tip tank holds 50 gallons of fuel, of which about 1.5 gallons is unusable.

The auxiliary fuel tanks in the accident aircraft could hold as much as 31.5 gallons each. The aircraft owner told investigators that he typically ran the tanks down to two or three gallons each to leave enough fuel to slosh around and keep the inside of the tank bladders wet. He would then fuel the auxiliary tanks before flight with whatever fuel load was required for the trip.

The Cessna 340 is powered by two turbocharged, fuel-injected engines. Fuel is pumped from the tank to its respective engine where a fuel control unit meters the amount of fuel going to the injectors. Any excess fuel is then always returned to the main fuel tank. That's why a pilot can't immediately switch to the auxiliary tanks when the main tanks are full. The Owner's Manual states that pilots should burn off 90 minutes of fuel from the main tanks before switching to the auxiliaries. That's to allow room for the returned fuel. Otherwise, the excess fuel it will be pumped overboard and lost.

Awareness of where the fuel selectors are placed and how much fuel is in each tank is essential when flying one of these aircraft. My experience with twin Cessna fuel systems is that the gauges are reasonably accurate. Checklists must be followed to insure that the correct selections are made for landing and takeoff.

But the lack of fuel being supplied to one engine was only one link in the error chain that led to this accident. An engine failure on an approach is a manageable situation if it's handled correctly. If the NTSB is correct and all that happened was that the engine ran out of gas, then the pilot still had ample opportunity to save the day.

The pilot reported the engine failure at 4:31 p.m., six seconds before radar contact was lost. The radar plot shows that by then the airplane was already in a left turn and off the localizer. It is likely the engine failure occurred about a minute earlier, when according to the radar plot the airplane was still maintaining the localizer.

When the failure occurred, had the pilot maintained directional control, feathered the left propeller and used power from the right engine as necessary, he should have been able to get the airplane on the runway. He was about 3.5 miles south of the airport and needed to clean up the airplane to keep it flying. He didn't do it. While the failure to insure that fuel was getting to the left engine may have caused the engine to quit, it can be argued that the pilot's lack of following emergency procedures likely contributed to the accident.

Flying an airplane on instruments is not a difficult task for an experienced pilot when everything is working well. But when problems occur it becomes necessary for the pilot to maintain control of the aircraft while handling the resulting situation. Did the Cessna 340 pilot have the knowledge to keep the airplane flying after the engine failure? His prior training and experience suggests that he did. Yet he may have gotten distracted by trying to solve the problem when he should have focused on securing the engine and completing a single- engine approach.

When an engine failure occurs in a twin-engine airplane close to the ground, you must act quickly to contain the situation. That means having certain items in memory and acting upon them right away.

Once you are past the final approach fix on an ILS approach, your focus must be on getting the aircraft to the runway. You can always figure out why the engine quit after you've landed. To do that you must clean up the airplane to the point where you can keep it in the air and going straight ahead. That means applying the correct rudder pressure to counter the asymmetric thrust and feathering the faulty engine's propeller to help maintain control over the direction the airplane is traveling. Unless the failure is catastrophic, you should be able to maintain the glide slope even on a single engine, so focus on flying instead of troubleshooting.

While our airplanes are aging, they are still highly reliable. Therefore, we don't practice enough those important life-sustaining procedures that will help us out of situations like the one the C-340 pilot found himself in. He didn't have much time to think about what he needed to do. Unfortunately, it appears that he did not use the time he had to keep the airplane in the air.


More accident analyses are available in AVweb's Probable Cause Index. And for monthly articles about IFR flying including accident reports like this one, subscribe to AVweb's sister publication, IFR Refresher.