DES MOINES, IA, USA
N608AW
British Aerospace BAE 146
While operating in clouds at flight level 290, in an area with thunderstorms and turbulence, the airplane experienced a loss of power in all four engines. Information from the flight data recorder indicated that during descent from flight level 290, an elevator oscillation occurred. The #1 and #2 engines were shut down due to high turbine gas temperatures. The #2 engine was restarted before making an emergency landing at the Des Moines Airport. The captain reported that the airplane had been flown into severe icing conditions. The aircraft flight manual (AFM) prohibited flight in icing conditions above 26,000 feet (except when necessary for descent). Also, the AFM stipulated the minimum engine speeds which needed to be maintained while operating in icing conditions; it also required that the thrust management system (TMS) be disconnected in icing conditions. There were indications that the TMS was disconnected about 230 seconds after engine roll-back began. Additionally, the AFM did not provide accurate information concerning 'in cloud' correction factor for the outside temperature thermometer.
HISTORY OF FLIGHT On May 26, 1996, at 2120 central daylight time (cdt), a British Aerospace, BAe 146, N608AW, operating as Air Wisconsin flight 5558, experienced a loss of power on all four engines. The airplane departed from the Denver International Airport, with an intended destination of Quad City Airport, Moline, Illinois. The airplane was operating at flight level 290, when the loss of power began. Radar showed that at 2109 cdt, the approximate time of the descent from flight level 290, flight 5558's position was North 40 degrees 51 minutes 42 seconds, and West 94 degrees 20 minutes and 22 seconds. The pilot shut down the number one, and the number two engines, due to high turbine gas temperatures. The pilots also reported that the aircraft was experiencing an uncommanded pitch oscillation, but did not report this event until the day after the incident. The crew restarted the number two engine before making an emergency landing at the Des Moines, Iowa, Airport. The four flight crew members, and 54 passengers were uninjured in the incident. The 14 CFR Part 121 flight was operating on an IFR flight plan, and instrument meteorological conditions prevailed at the time of the incident. METEOROLOGICAL INFORMATION A weather package was obtained by the investigator in charge (IIC) for the time period of approximately one half hour before until one half hour after 0209 universal time coordinate (UTC). This information displays the cloud base reflectivity, wind profile and the cloud tops, around the approximate location of N608AW, when it began its descent from 29,000 feet. This weather data is included as a supplement to this report. COMMUNICATIONS The en route communications for the flight of N608AW were obtained by the investigator in charge from both Minneapolis Center and Chicago Center. The company dispatch tape for the flight was also obtained by the IIC. FLIGHT RECORDERS The aircraft was equipped with both a cockpit voice recorder (CVR) and a flight data recorder (FDR). No usable data was available from the CVR. Most of the data from the incident flight was retrieved from the FDR. Some gaps existed on the FDR data, which would be consistent with the flight crews reports of power interruptions occurring during the flight. TESTS AND RESEARCH The FDR read out was preformed by the NTSB and the Air Accidents Investigations Bureau (AAIB) with the assistance of AVRO. The complete textual description of the flight as prepared by the AAIB in conjunction with AVRO is included as a supplement to this document. The FDR data showed that when the aircraft was at 29,000 feet the static air temperature varied from -28.8 degrees celsius to -30.4 degrees celsius. The FDR data also showed that while at 29,000 feet the N1's split apart from each other, similar to what would happen if the thrust management system (TMS) was disconnected while in an N1 sync mode. Before the descent was begun from 29,000 feet the FDR showed signs which appear consistent with the beginning of the loss of power on all four engines. The FDR data indicated that during the descent from 29,000 feet the elevator transducers showed movement of the elevators, similar in nature to what has previously been seen before on this aircraft, and referred to as a pitch oscillation. The investigators who looked at this data classified this pitch oscillation as the mildest pitch oscillation ever recorded. During the captain's radio transmissions with Minneapolis Center he requests a lower altitude from flight level 290. Minneapolis Center responded by clearing flight 5558 down to flight level 240. Control of flight 5558 was transferred to Chicago Center after the descent to flight level 240. During radio transmission with Chicago Center the captain of flight 5558 said "Chicago Wisconsin 558, 240, we need lower." Chicago Center replied "Wisconsin 558 roger, stand by." The captain replied "Wisconsin 558 we're descending we need lower altitude at this time." Chicago Center told the pilot that they were unable to approve a descent at that time, at which time the captain stated "Wisconsin 558 we are descending, we are descending." After this transmission Chicago Center cleared the flight to descend to flight level 190, and further cleared the flight down to 12,000 feet. The captain declared an emergency with Chicago Center at 18,000 feet, and reported "lost all power at this time, were descending, 558." The Chicago Center controller then gave flight 5558 a direct heading of 350 degrees, for the Des Moines Airport, Des Moines, Iowa. The pilots of the aircraft were interviewed by the IIC in Denver, Colorado on June 27, 1996. The captain remembered being clear of clouds at 24,000 feet, at which time the decision was made to climb to 29,000 feet. The captain thought that the airplane was at least 50 miles from any thunderstorm cells. The captain said the airplane was traveling through green on the radar at the time of the incident. The captain recalled having the engine and tail anti ice systems on while at 29,000 feet. He said that the converted SAT temperature was -40 degrees celsius. The captain did not recall any heavy moisture, or any indications of ice while cruising at 29,000 feet. He said that the pitch oscillation began before the engine rollback occurred. The captain thought that the airplane had been in the clouds for approximately 10 minutes before the pitch oscillation occurred. He did not remember any TMS arrows or the engine power settings while operating the airplane at 29,000 feet. The copilot on the flight also remembered being clear of clouds before climbing the airplane to 29,000 feet. He reported that he disconnected the autopilot while at 29,000 feet. The copilot reported that the pitch oscillation lasted approximately 15 seconds, and that the maximum amplitude and frequency occurred at approximately 10 seconds. He reported that the pitch oscillation occurred before the descent from flight level 290 was begun. The copilot did not remember any ice detection light being on until the airplane began descending from 29,000 feet. He believed that the airplane was in the clouds between five and 10 minutes before the pitch oscillation occurred. He said he would have been the one to disengage the TMS at 29,000 feet, but did not recall disconnecting the TMS. The copilot did not remember the power setting while at 29,000 feet. During the captain's conversation with dispatch immediately following the flight he reported to the company dispatcher that "We got in some severe icing, and we lost power to all four engines." He also stated that "we had a lot of ice all the way down." When company personnel asked the pilot if he would classify the in flight conditions as severe icing with moderate turbulence the captain agreed. When the company personnel asked the captain if they had reduced the power on the engines, the captain was unsure, but thought they might have reduced the power on the engines due to turbulence. The aircraft flight manual (AFM) stipulates minimum engine speeds which need to be maintained while operating in icing conditions, and requires that the TMS be disconnected. The AFM also prohibits flight in icing conditions above 26,000 feet, except when it is necessary for descent. During this conversation there was no mention made by the captain about a pitch oscillation. When the company personal told the captain that they were not going to find any problems with the airplane, and this incident sounded like a classic case of engine rollback, the captain did not disagree. The IIC obtained both the AFM correction to outside air thermometer, and the Air Wisconsin Operating Handbook Correction To Outside Air Thermometer. The AFM states "In cloud the value of the correction factor is reduced, that is the instrument reads more closely to the true temperature." The AFM does not say how much the temperature correction is reduced by, when in clouds. When the IIC questioned the manufacturer about the temperature correction factor while in clouds, the manufacturer stated that in heavy moisture the outside air thermometer could indicate static air temperature. The Air Wisconsin Operating Manual did not make any reference to different temperature correction factors when operating in clouds. However, the supplement states: "For a more detailed chart, refer to Aircraft Flight Manual, Section 5.01.00 page 11." ADDITIONAL INFORMATION Following this incident the aircraft manufacturer modified the procedures involved in dealing with a pitch oscillation. The new procedures require the use of the airbrake to slow the airplane in the event of a pitch oscillation. A warning was also added to the pitch oscillation procedures which requires minimum power settings to be maintained if above 26,000 feet in icing conditions. Parties to the investigation were AVRO, Air Accidents Investigation Branch, Allied Signal, Airline Pilot's Association, Air Wisconsin and the Federal Aviation Administration.
failure of the flight crew to follow proper procedures, which resulted in an accumulation of ice and subsequent loss of engine power. Factors relating to the incident were: thunderstorms, turbulence, icing conditions, and the inaccurate method that was provided by the manufacturer for determining 'in cloud' air temperature and icing conditions.
Source: NTSB Aviation Accident Database
Aviation Accidents App
In-Depth Access to Aviation Accident Reports
