DCA06IA007

The flight was level at Flight Level (FL) 400 for approximately 30 minutes. Air traffic control (ATC) cleared the flight to FL380, where it remained for about 15 minutes. Further clearance was issued for pilot's discretionary descent to FL330. The flight was operating in visual meteorological conditions in the vicinity of cumulonimbus buildups. Upon reducing power to initiate descent the crew heard a "popping" noise and both engines rolled back. The crew declared an emergency, elected to divert to Jacksonville Airport, and initiated emergency procedures. All attempts to restart the engines were unsuccessful, and the crew elected to save battery power for navigation and landing. The crew executed a successful emergency landing. After landing, the Captain attempted to restart the engine, and did observe some rotation, however stopped when the temperature did not rise.

Post incident examination of the aircraft, and review of crew actions indicated that the cause of the dual-engine flameout was not due to fuel exhaustion or fuel starvation. There were no open maintenance items, and post-incident evaluation and flight test of the aircraft revealed no mechanical malfunctions. The chemical composition of the fuel was correct, anti-ice additive was present, although it was marginally below the recommended ratio, and no contamination or excess water content was found. There was no evidence of core lock, a phenomenon in which differential cooling of engine components following a rapid in-flight rollback causes a mechanical binding of the rotating components of the engine. In-flight testing of the right engine of the incident airplane was unable to cause a core lock condition.

This flight and other BE-400A engine flameout events before and subsequent to this incident occurred in similar environmental conditions. Pilot reports and meteorological records indicated the flights were in or near instrument meteorological conditions with convective weather cells in the area and the pilots had just retarded the power levers to initiate a descent.

Research revealed that convective storms can lift significant amounts of water into the upper atmosphere and that the blowoff from the tops of these storms can contain significant amounts of ice crystals. A post-incident study showed that the ice crystals could partially melt passing through the low-pressure compressor of the JT15D-5 engine due to the increase in temperature of the air being compressed. Further, the study determined that with the engine anti-ice turned off, that it was possible for the ice crystals to accrete on the leading edges of the front inner compressor stator leading edges. The study also determined that if a significant build up of ice had occurred, any change in the airflow angle of incidence that would occur as power was reduced would cause any ice that had accreted on the leading edges of the stators to break away and would result in the engine surging or possibly flaming out. The study also revealed that after the engine had flamed out, the radiant heat from the oil tank, which is in the core of the engine between the low and high pressure compressors, could cause the ice on the front inner compressor stators to melt and the water could run back and refreeze in the high pressure compressor impeller, acting like a wedge to prevent engine rotation and restart.

Research, flight tests, and bench testing revealed that a temporary ice blockage in an orifice in the combustion chamber pressure signal line could cause a sudden drop in fuel flow, much more rapidly than normal, and reduce flow to a level that would not support combustion. The orifice was added to eliminate trapped water from the line, however this orifice was small enough to be thermally overwhelmed in the incident conditions.

Pilot interviews revealed that pilots do not perceive high altitude ice crystals to be a threat to the airplane or the engines. The Beechjet's airplane flight manual (AFM), and flight crew training, did not address high altitude ice crystals prior to the dual-engine flameout events. Post-incident, Raytheon developed guidance for Beechjet 400A flight crews on high altitude ice crystals that was distributed to the 400A community via a safety communiqué. The NTSB recommended that the guidance provided in the Raytheon safety communiqué be incorporated into the Beechjet 400A AFM as well as the AFMs of other JT15D-powered airplanes (A-06-57 and 58). The NTSB also recommended that the FAA require the operation of the engine anti-ice system whenever flights are near convective weather (A-06-56).

Crew actions after the flameout were appropriate to attempt to restart the engines, notify and coordinate with ATC for an emergency diversion, and make a forced landing. In fact, investigators found the crew's coordination and actions to be exemplary. Investigators also found ATC's assistance to the crew was also exemplary and contributed to the successful outcome of the flight.

History of flight

On November 28, 2005, about 1340 eastern standard time, a Raytheon Beechcraft 400A, N691TA, lost all power from both of the Pratt & Whitney Canada (PWC) JT15D-5 engines while descending from flight level (FL) 380 (approximately 38,000 feet above sea level) near Jacksonville, Florida. The two pilots on board were not injured. The airplane safely landed at Jacksonville International Airport. The flight was operated by Flight Options in visual meteorological conditions on an instrument flight rules flight plan under the provisions of 14 Code of Federal Regulations Part 91 from Indianapolis, Indiana, to Marco Island, Florida.

The captain was seated in the right seat and was performing check airman and pilot monitoring duties. The first officer was seated in the left seat and was the pilot flying. The flight crew reported that after departing Indianapolis, it took approximately 25 minutes to climb to FL350 where they remained for about 40 minutes. The airplane then climbed to FL400 where it remained for about 30 minutes. At about 1316, air traffic control (ATC) cleared the flight to descend to FL380. The crew did so, and they flew level at that altitude for about 15 minutes. At 1331, ATC cleared the crew to descend at pilot's discretion to FL330, the crew delayed descent briefly until the planned top of descent point. The pilots reported that they were cruising in visual meteorological conditions at the time, near cloud tops. They reported the static air temperature was -57º Celsius and the total air temperature was -37º Celsius. The crew monitored fuel temperatures during the flight and noted a minimum of -23º to -24º degrees Celsius. The captain remembered looking outside for signs of airframe ice, and saw none. They were not operating engine anti-ice. The pilots reported that they had been cruising at about .73 to .74 Mach, and there were no abnormal indications throughout the flight.

The pilots reported that when retarding the throttles to initiate the descent, they heard a loud pop from the right engine followed about 10 seconds later by a loud pop from the left engine. The captain also termed the sound a "bang" and said he had heard a compressor stall at another time in a Beechjet, on the ground, and this sound was not similar. The pilots observed the engine indicators quickly roll back, N2 dropped immediately to 46% and then to zero. The pilots donned their oxygen masks and declared an emergency and advised ATC that they had lost both engines and needed to descend to a lower altitude. The captain said he purposely switched electrical power to the emergency battery bus to save the battery charge for an attempted restart.

The captain said there was brief loss of airspeed indication, and he said he used the "peanut" (backup) gyro attitude indicator and altimeter to control airplane pitch. About one to one and a half minutes later the airspeed indication returned to normal. (The captain stated he knew of a service bulletin that was related to this phenomenon.)

After both engines flamed out, the crew set the airspeed for 170 to 180 knots for best glide. The captain planned on attempting a battery start at 180 knots, per the restart envelope on the emergency checklists. When he first attempted to restart the left engine using the battery-assisted restart procedure at about FL280, he did not observe any rotation, the voltage dropped from 28 volts to 8 volts, and he heard a "humming or groaning" sound. As the airplane passed through FL260 he reported that he pitched further downward to increase speed to 230 knots and to attempt an air start of both engines, but he did not observe any N2 rotation on either engine and the attempt was aborted. The crew then switched back to normal battery to attempt a battery-assisted restart on the right engine, with no success.

During the descent, ATC provided vectors to the ILS [instrument landing system] approach to runway 7 at Jacksonville. The flight was in clouds during the descent, with moderate to heavy rain beginning at about 10,000 feet. As the airplane neared the airport, ATC provided continuous callouts of the distance remaining to the runway that the pilots later stated was very helpful in managing their descent and approach to the airport. The captain took over the controls at about 9,000 feet for the landing. The crew manually extended the landing gear and broke out of the clouds at about 1,200 feet. The pilots stated that at about 500 feet, they began to observe N2 indications between 1% and 2%. When on short final, they saw the left engine N1 indicate 1.1%.

After they landed and rolled off the runway onto a taxiway, the right landing gear tire deflated. The pilots stated that while the airplane was stopped on the taxiway after the airplane had landed and they were waiting to be towed into the ramp, they pressed the left engine starter button and noted engine rotation on the cockpit gauges but aborted the start. The captain stated that the engine would rotate, but "wouldn't take." Instruments indicated 24 volts on the battery, and a positive fuel flow, but no rise in turbine inlet temperature.

Post-event examination revealed that the fuel gauges indicated 1400 and 1350 pounds of fuel in the left and right wing tanks, respectively, and visual examination noted fuel level up to the flapper valves in each tank. The center tank was empty.

Personnel information

The captain held an airline transport pilot certificate and had approximately 8,200 hours total time, with 1,800 hours in the Beechjet 400A. The first officer held an airline transport pilot certificate, and had approximately 3,100 hours total time, with about 20 hours in the Beechjet 400A. The first officer was newly hired by Flight Options and was receiving Initial Operating Experience training from the captain during the incident flight.

Aircraft information

According to Flight Options maintenance records, the airplane, fuselage serial number (SN) RK-317, had operated 4,010.3 hours and 3,359 cycles since new. The airplane underwent an "A" check on November 15, 2005, and had operated 31.4 hours and 26 cycles since that check.

The airplane was equipped with two PWC JT15D-5 dual-spool turbofan engines with a takeoff thrust rating of 2,965 pounds. The No. 1 (left) engine was SN PCE-JA0521 and had operated 662.3 hours and 604 cycles since new and the No. 2 engine was SN PCE-JA0407 and had operated 4,010.3 hours and 3,359 cycles since new.

According to the Flight Options load manifest for the incident flight the gross weight of the airplane at departure was 16,061 pounds and the maximum allowable takeoff weight was 16,300 pounds. The airplane was loaded with a center of gravity (CG) of 28.1% mean aerodynamic chord (MAC), with an acceptable CG range of 17.2 - 31.7 MAC.

There were no open maintenance items in the airplane's logbook.

Meteorological information

There were widespread thunderstorms with cirrostratus outflow over the southeastern US at the incident time. Weather data obtained by a Safety Board meteorologist indicated the airplane overflew an area of convective activity, with tops above FL420 to the southwest of the route. The pilots reported that while cruising at FL400 they were clear of all clouds, and during an intermediate cruise at FL380 for about 15 minutes prior to the loss of power, they were about 400 to 500 feet above the cloud tops with good visibility and no turbulence. The captain reported that at the time of the flameout, they were in the clear, but within about five miles of a convective cloud buildup.

Communications

No communications problems with ATC were noted at any time during the accident sequence. The crew promptly and effectively advised ATC of the situation, and ATC provided vectors and additional information to assist in the forced landing. Inter-crew communication was hampered by a partially extracted microphone plug on one of