Probable Cause #59: Failure To Identify

An engine failure in the clouds becomes even more challenging if the pilot misdiagnoses the problem.

This article originally appeared in IFR Refresher, July 2007.

A turbocharged engine provides sea-level power to a higher altitude than a normally aspirated powerplant and is the main difference between the two. However, if a pilot is not totally familiar with the operation of turbocharged engines, it is entirely possible to misdiagnose a problem.Could that have happened here? On Feb. 22, 2006, at 1750 (PST) the pilot of a Beech 58P pressurized Baron contacted the Seattle AFSS requesting a standard briefing for an IFR flight from Bellingham, Wash., to Ogden, Utah.The pilot told the briefer he was planning on leaving Bellingham 20 minutes later and requested the winds for 12,000 and 18,000 feet.The briefer detailed a northerly flow aloft with an area of high pressure off shore and moisture due to a trough of low pressure at the lower levels in western Washington. AIRMETs for the entire route of flight called for occasional moderate turbulence below FL180, with turbulence above FL180 beyond Boise, Idaho.The area forecast predicted occasional moderate rime and mixed icing in clouds and precipitation from 2000 to 14,000 feet. The briefer also advised that icing could be expected in the descent from 12,000 feet nearly to the surface. Departure weather at Bellingham was reported as 3500 overcast and 10 miles visibility.

Lost an Engine

The pilot called the Bellingham ground controller at 1835 for his clearance to Ogden and departed at 1843. The aircraft was handled briefly by the Whidbey Approach Control but quickly passed on to Seattle Center.The Baron pilot contacted Seattle Center leaving 7800 feet for 9000 feet as the controller cleared the aircraft to 15,000 feet. Two minutes later the pilot attempted to contact the center controller. The transmission was mostly unintelligible, and the controller asked the pilot to repeat it."Mayday, Mayday, Mayday. Just lost an engine."The controller asked the pilot to "say your intentions." No more transmissions were received from the aircraft. The aircraft crashed about two miles south of Bow, Wash., approximately 15 miles southeast of the Bellingham Airport. The pilot, the sole occupant, was killed in the crash.There were several witnesses to the accident. One said he was about a half mile from the crash site. He told investigators that the "engine noise was excessively loud and the aircraft was flying in fast, tight, horizontal circles. The flight path seemed erratic."The Baron dove to within approximately 150 feet of the ground and pulled up, repeating the same maneuver, getting even closer to the ground. The third time the aircraft crashed into the ground. Another witness said that the aircraft flew away from him during its maneuvering, and that the engine noise was consistently loud as "if it were at full throttle."A third said she was driving east when she saw a light going straight "down to the ground at a very rapid speed." It was raining at the time. She said her view was blocked by buildings and trees, and she assumed it was an airplane. She was not sure how far it might have been from her.From the front porch of his house, another witness said the ceilings were low and that it was raining at the time of the accident. The first indication of trouble, he told investigators, was the "engine overspeeding." He observed the aircraft do a series of uncontrolled rolls and deep diving maneuvers. The aircraft seemed to level out, "but then went into a nose-down spin and disappeared from view." He said he heard the impact several seconds later.

At the Controls

The pilot held a private certificate with a single-engine land rating, a commercial certificate with a multi-engine land rating and a commercial instrument-airplane rating. His medical certificate was current. His most recent medical application listed total flight time as 7500 hours, with 150 hours in the prior six months.No pilot or aircraft logbooks were located. However, documentation confirmed the pilot had completed a flight review and instrument proficiency check on Oct. 22 of the previous year.Investigators were able to determine that the last annual inspection was completed in December 2005, when the aircraft had 9401 hours in service. Both engines had been top overhauled at the previous annual inspection. The right engine had 621 hours since a major overhaul, and the left engine had 695 hours since major.The closest weather reporting facility to the accident site was Skagit Regional Airport, about 4.5 miles south. The AWOS at Skagit recorded the weather at 1850 PST as winds 180 degrees at 4 knots, visibility 8 statute miles, overcast skies at 3400 feet, and a temperature of 4 degrees C. At 1910 it was 7 miles visibility, scattered clouds at 100 feet, broken clouds at 2500 feet, and overcast skies at 3400 feet.All aircraft flight control surfaces and major system components were located at the wreckage site prior to recovery efforts. The impact crater was full of water and oil. The landing gear and flaps were in the retracted position.Both engines were torn down at the manufacturer's facility with NTSB representatives present. Both had extensive, impact-related damage. However, disassembly and inspection revealed no inordinate wear and no evidence of the failure of any internal components.Such was not the case when the turbochargers were examined. The right turbocharger had sustained damage to the compressor wheel and compressor housing shroud. Foreign material was found in the compressor housing and turbocharger after-cooler. The material consisted of aluminum fragments, a rivet head, a piece of brass, paint chips, rubber, gasket material and dirt.The turbocharger was sent to another facility for further examination where additional damage to the turbine/compressor input shaft, possibly from insufficient lubrication, was found. The shaft journal, on the turbine end of the shaft, had contacted the journal bearing. Additionally, the bearing's bores "were out of round and heavily tracked from contact with the shaft." An imbalance in the shaft, caused by foreign object damage to the compressor wheel and a lack of lubricant, "may have caused" the unit's failure.

Why?

The NTSB blamed the accident on the pilot's failure to maintain aircraft control during climb to cruise. Contributing factors included a loss of engine power to the right engine following turbocharger failure, instrument and night meteorological conditions.There's no question the pilot did not handle the engine failure effectively, especially since the engine failure occurred nearly 8000 AGL. The pilot should have flown the airplane first, increasing power on the good engine while using the necessary control inputs to maintain level flight and a safe airspeed. That also meant leveling off and handling the emergency, including feathering the propeller of the dead engine, while considering a place to land, all of course, while on solid instruments. That's a lot of work for one person, but something for which a single-pilot flying a multi-engine aircraft IFR must be prepared.Many pilots don't practice emergency procedures often enough. When the crunch comes, they are not prepared. We don't know for certain if the pilot's lack of proficiency was the causes, although the reports from witnesses of erratic flying make it a likely possibility.There is another possible theory to explore regarding the Baron. If the turbocharger failed, the engine may have still put out some power. Investigators did not find any reason for a total failure of the engine. If the pilot recognized that the engine was not dead, but that the turbocharger had failed as the aircraft descended, he could have gained more power from the bad engine. That certainly would have helped to avoid what seemed like the description of a minimum-controllable airspeed (VMC) roll by one of the witnesses just before the aircraft crashed. Did he simply not notice what was really causing the emergency?The accident report also does not indicate whether any of the propeller blades were feathered, which leads me to wonder if the pilot failed to feather the propeller at all.But perhaps he recognized he had some power left on the bad engine, and if he did, he might have had a problem with flying the gauges and trying to handle the emergency at the same time.If you recognize a turbocharger failure, you must consider that perhaps the exhaust system failed, not the turbocharger itself. If that happens, the potential for fire is high, and it probably is best to shut the engine down to reduce that possibility.I have had two engine failures while flying multi-engine aircraft. The first occurred because a very nervous passenger kicked the right-engine fuel selector to the off position in a Cessna 310 while we were in level flight at 7500 feet directly over the airport at Portland, Maine. There was no need to rush as we were in no danger. Had we been in IMC, I probably would have thought about it a little differently.The oil pressure on the engine was almost normal as the engine wind-milled, so I looked down at the fuel selector and realized what had happened. I turned the fuel back on and the engine came back to life a few seconds later.I explained to my passenger what had happened, and he settled down after an explanation that the airplane could fly quite nicely on one engine. It was a good demonstration for him, though I don't recommend anyone do it intentionally to prove the point.The second failure was in a Cessna 421 at low altitude shortly after takeoff from Providence, R.I. I had just dropped my passengers off and was heading back to my home airport 12 miles away. The engine quit with little warning.The runway for my home airport was about 7 miles in front of me and the good engine kept the airplane at 1000 feet AGL. I feathered the propeller on the bad engine and flew straight-in to an uneventful landing.There are all kinds of stories about the poor handling characteristics of twin engine airplanes with an engine malfunctioning. VMC accidents often occur when the pilot allows the airspeed to bleed off because they are trying to keep the airplane level or make a turn when a controlled descent would be a much better idea. The worst VMC-related situation is an engine failure shortly after takeoff with the airplane close to gross weight. If the pilot senses a loss of control, even after the engine has been feathered, they are far better off trying to land under control straight ahead.How would you handle an engine failure in IMC? Are you current and proficient at flying the gauges with some distraction? Remember, fly the airplane first. But you must be able to multi-task to some degree, at least in order to secure the faulty engine, declare the emergency and consider alternative landing sites.If you don't practice engine failures under the hood, if and when it does happen, you likely will become nothing more than a passenger along for the ride.


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.