NTSB Factual Report

HISTORY OF FLIGHT

On August 27, 2009, about 1603 eastern daylight time (EDT), a Boeing 717-200, N994AT, operated by Airtran Airways as flight 163, encountered convective turbulence while climbing through flight level 250 near Inverness, Florida. One flight attendant sustained serious injuries. The other two flight attendants, the two airline transport pilots, and the 81 passengers were not injured, and the airplane was not damaged. Instrument meteorological conditions prevailed, and an instrument flight rules flight plan had been filed for the flight that departed Tampa International Airport (TPA), Tampa, Florida, destined for Hartsfield-Jackson Atlanta International Airport (ATL), Atlanta, Georgia. The scheduled passenger flight was conducted under Title 14 Code of Federal Regulations (CFR) Part 121.

According to the flight crew, after departure, the airplane was assigned a heading of 350 degrees, and was cleared to climb to its cruise altitude. The flight then entered instrument meteorological conditions (IMC), and encountered "very mild light chop." There were "no reports of turbulence from PIREPS or ATC and no cells were displayed on the radar." The seat belt sign was illuminated. After the airplane passed through 10,000 feet, the flight attendants began their cabin service. As the airplane passed through flight level 250, still in IMC, it encountered turbulence for approximately 15 seconds. The non-flying pilot checked the airborne radar, and changed the range settings to see if the airplane had "hit a cell." He saw "nothing" on the radar. The turbulence encounter occurred approximately 60 miles north of TPA.

According to the lead flight attendant, the cabin crew consisted of three flight attendants, designated as Lead, R1, and L2. The lead attendant was working business class at the front of the cabin, and the R1 and L2 attendants were working the service cart in the aisle at the rear of the airplane. The R1 attendant was forward of the cart, and the L2 attendant was aft of the cart. At the time of the turbulence encounter, the lead attendant was in the galley, and he "came off the floor about a foot." He reported that the turbulence lasted about 15 to 20 seconds, that coffee was "splashing out of the pots," and that beverage cans were turning over. According to the operator's incident report, during the turbulence encounter the L2 flight attendant and the service cart "came off the floor," and the cart struck the flight attendant's leg.

The turbulence subsided, and the L2 flight attendant was observed to be lying on the floor, and complaining of "intense pain" in her right leg. The flight attendant's leg was immobilized and a passenger who was a nurse helped her to an unused row of seats. The flight crew was notified of the injury, and informed that the injured attendant could not perform her duties. The flight crew and the dispatch department arranged for medical personnel to meet the flight in ATL. Cabin service was discontinued, and the R1 flight attendant occupied the L2 station at the tailcone door for the landing. The injuries were subsequently diagnosed as a fractured tibia and a crushed ankle.

PERSONNEL INFORMATION

Captain

According to records provided by the operator, the captain held an airline transport pilot certificate with multiple ratings including airplane multi-engine land, and a type rating for the Boeing 717. His most recent Federal Aviation Administration (FAA) first-class medical certificate was issued May 2009. The captain reported 18,700 total hours of flight experience, including 5,000 hours in the accident airplane make and model. The captain's length of employment with the operator was not determined.

First Officer

The first officer held an airline transport pilot certificate with ratings for airplane multi-engine land, and multiple type ratings. His most recent FAA first-class medical certificate was issued in February, 2009. The first officer reported 10,000 total hours of flight experience, including 3,000 hours in the accident airplane make and model. The first officer's length of employment with the operator was not determined.

Cabin Personnel

The lead flight attendant was employed by the operator as a flight attendant for nearly 5 years. The L2 (injured) flight attendant was employed by the operator as a flight attendant for 2 1/2 years. The R1 flight attendant's length of employment with the operator was not determined.

AIRCRAFT INFORMATION

The airplane was a variant of the Douglas Aircraft DC-9, and was manufactured in 2002. It was powered by two Rolls-Royce BR 715 series turbofan engines, and at the time of the turbulence encounter, the airplane gross weight was approximately 99,000 pounds. According to the operator, the airplane was configured for 2 flight crew, 4 cabin crew, and 117 passengers.

METEOROLOGICAL INFORMATION

General Forecast

Detailed meteorological information regarding the flight conditions was documented in a separate report by a National Transportation Safety Board (NTSB) meteorologist. That and other information is summarized briefly in this section.

The National Weather Service (NWS) data issued at 0800 depicted conditions which supported widespread upward vertical motion over the region of the flight track. The 1230 NWS Convective Outlook indicated that only general airmass-type convective activity was expected, and did not forecast any organized severe weather for the region.

The 1345 area forecast for northern Florida called for scattered clouds at 1,500 and 4,500 feet, broken cloud layer at 12,000 feet, with tops to 17,000 feet, broken cirrus clouds above, with scattered thunderstorms and light rain. Cumulonimbus cloud tops were forecast to reach 45,000 feet. The same forecast stated that after 2000, scattered clouds at 2,000 feet, a broken cloud layer at 12,000 feet, with widely scattered thunderstorms and light rain, were expected.

NWS information issued at 1400 depicted conditions in the immediate vicinity of the turbulence encounter that provided additional support for the generation of convective activity, including moderate to strong multicellular-type thunderstorms in lines and clusters.

AIRMETs and SIGMETs

The NWS website stated that "AIRMETs (AIRman's METeorological information) are issued by the Aviation Weather Center to advise of weather potentially hazardous to all aircraft but that does not meet SIGMET criteria." The NWS stated that "moderate turbulence" was one of the conditions that would result in the issuance of an AIRMET, and that "AIRMETS are also amended as necessary due to changing weather conditions or issuance/cancellation of a SIGMET." AIRMETs are routinely issued for 6-hour periods beginning at 2245 EDT.

The NWS website stated that a "SIGMET (SIGnificant METeorological information) advises of weather potentially hazardous to all aircraft other than convective activity," and are issued for several reasons, including "severe or extreme turbulence." SIGMETs are typically issued for 4-hour periods. According to the NWS, "if conditions persist beyond the forecast period, the SIGMET is updated and reissued."

Convective SIGMETs are issued for "severe surface weather," including "embedded thunderstorms, lines of thunderstorms, and thunderstorms greater than or equal to video integrator and processor (VIP) intensity level 4 affecting 40% or more of an area at least 3000 square miles." The NWS stated that "any convective SIGMET implies severe or greater turbulence," that they are issued hourly at 55 minutes past each hour, and are valid for up to 2 hours.

Between 1255 and 1555, the NWS issued three convective SIGMETs that overlaid the planned flight track in the vicinity of the turbulence encounter, and which forecast embedded thunderstorms, with tops above flight level 450, moving approximately south to north. At 1255, the NWS issued convective SIGMET 37E. That convective SIGMET was valid until 1455, which was about 1 hour prior to the planned departure time. At 1455, the NWS issued convective SIGMET 43E, which overlaid a portion of the area encompassed by 37E, and was valid until 1655. Convective SIGMET 43E forecast a diminishing area of embedded thunderstorms. At 1555, about 6 minutes after the flight took off, convective SIGMET 46E was issued. Convective SIGMET 46E was valid until 1755, forecast embedded thunderstorms, and included areas common to 37E and 43E.

Meteorological Conditions in Vicinity of Turbulence Encounter

The 1600 NWS radar composite reflectivity mosaic image depicted several defined echoes associated with thunderstorms north of the departure airport, including a short but intense line at the latitude and longitude of the turbulence encounter, but below the airplane's altitude. The 1620 NWS radar summary chart depicted an extensive area of thunderstorms in the vicinity of the flight track, and also depicted the solid line of thunderstorms in the vicinity of the turbulence encounter. Weather observations from several airports in the vicinity of the encounter included the presence of thunderstorms prior to, during, and after the encounter.

Correlation of the ground-based WSR-88D with the flight path information revealed that the airplane was penetrating a line of radar echoes when the turbulence encounter occurred. The data depicted radar echoes of magnitude 20-30 dBZ (decibels) at the airplane's altitude, and echo magnitudes of 50 to 55 dBZ below the airplane. The radar data also depicted echoes that overlaid the airplane's flight path, with tops above the airplane's altitude.

Infrared satellite imagery from 1602 depicted the presence of cumulus clouds, with tops near 36,500 feet, at and around the location of the turbulence encounter. The radar and satellite imagery indicated conditions that were consistent with the airplane flying through clouds at the time of the turbulence encounter.

"Tail-end Charlie" Phenomenon

The operator's submission to the NTSB suggested that the airplane encountered a "Tail-end Charlie," which was a common-use term for the trailing ("tail-end") cell in a line of build-ups or thunderstorms. Review of the available ground-based WSR-88D radar data revealed that a continuous line of echoes was oriented along a southwest to northeast axis. The line was translating to the north-northwest, and was expanding in length to the southwest. The line was depicted on the radar summary chart with reflectivity Levels 5 or 6, which were defined as "intense to extreme." The airplane penetrated the southwestern portion of the line, which was its advancing segment.

AIDS TO NAVIGATION

The initial segment of the planned route of flight was from TPA 101 miles north to the Cross City (CTY) very high frequency omni-range navigation facility (VOR), and then northwest to WYATT intersection. Although not cited in the flight plan as waypoints, two VORs which were equipped to broadcast hazardous inflight weather advisory service (HIWAS) were located within reception range of the airplane's flight track. The St. Petersburg (PIE) VOR was located about 10 miles west of TPA, and the Gators (GVN) VOR was located about 35 miles east of CTY VOR. Review of aeronautical charts indicated that at the time of the turbulence encounter, the airplane was about 60 miles north of the PIE VOR, and about 60 miles southwest of the GNV VOR.

COMMUNICATIONS

According to air traffic control (ATC) radio communications documentation provided by the FAA, the flight was issued taxi clearance at 1541, and takeoff clearance at 1547. Two minutes later, the flight was handed off to the Tampa TRACON (traffic control). About 1553, the flight made initial radio contact with Jacksonville Air Route Traffic Control Center (ARTCC, designated as "ZJX") sector R88. The turbulence encounter occurred at 1603, and the flight was handed off to ZJX sector R17 about 1 minute after that.

According to Chapter 2, section 6 of FAA Order 7110.65R (Air Traffic Control), "Controllers shall advise pilots of hazardous weather that may impact operations within 150 NM of their sector or area of jurisdiction. Hazardous weather information contained in HIWAS (hazardous inflight weather advisory service) broadcasts includes ... Significant Meteorological Information (SIGMET), Convective SIGMET (WST)...and Center Weather Advisories (CWA)." The section also stated that "Controllers within commissioned HIWAS areas shall broadcast a HIWAS alert on all frequencies, except emergency frequency, upon receipt of hazardous weather information. Controllers are required to disseminate data based on the operational impact on the sector or area of control jurisdiction."

Review of the communication transcripts revealed that none of the four ATC controllers who communicated with the flight prior to the turbulence encounter mentioned the existence of either SIGMET 43E or 46E, or broadcast a HIWAS alert, while the flight was on their frequency.

FLIGHT RECORDERS

The flight data recorder information was downloaded and provided to the NTSB Recorders Laboratory. The data indicated that when the airplane was climbing through an altitude of 25,300 feet, and at an airspeed of 307 knots, it experienced a rapid vertical acceleration increase that reached a peak value of approximately 2.5 g (acceleration due to gravity), and decreased to approximately 0.5 g within 1 second. The vertical acceleration values then fluctuated between 0.6 g and 1.2 g for the next 11 seconds, before they stabilized near the normal value of 1 g. During that same period, the data indicated pitch and roll excursions of approximately 5 and 20 degrees, respectively, and an airspeed decrease of approximately 15 knots. The altitude at the end of the encounter was approximately 25,800 feet.

The acceleration sensors were installed near the airplane's center of gravity (CG). However, since an airborne airplane has freedom of motion in six degrees, the accelerations measured at the CG may not be representative of those experienced elsewhere in the airplane. Accelerations in the cockpit or the aft cabin can be significantly less than or greater than those at the CG, and can differ significantly from each other as well.

ADDITIONAL INFORMATION

CAST, JSAT and JSIT

The Commercial Aviation Safety Team (CAST) was founded in 1998 as a joint industry -government effort to develop and apply an integrated, data-driven strategy to reduce the commercial accident fatality rate, and to improve commercial aviation safety in the United States. The CAST processes consisted of two sequential phases; analysis and implementation. The Joint Safety Analysis Team (JSAT) identified the top safety areas by the analysis of accident and incident data, produced "problem statements" which identified specific safety deficiencies, and identified and prioritized intervention strategies for use by the follow-on efforts of the Joint Safety Implementation Team (JSIT). The JSITs developed the means to implement the specified safety improvements through various equipment, process, operational and regulatory changes.

Several JSAT/JSIT teams, each assigned to a specific subject matter area, were formed; the Turbulence JSAT/JSIT focused on cabin safety issues related to turbulence. The Turbulence JSAT produced a total of 21 problem statements, and identified 18 safety enhancements for implementation.

The Turbulence JSAT developed the following three flight-crew-related problem statements and definitions:

No. 601: Flight Crew - Failure to process and interpret available, relevant data. Definition: Flight crew failure to process and interpret available, relevant data, including decisions arising from Collaborative Decision-Making.

No. 602: Failure to communicate with cabin crew/Passengers. Definition: Failure of flight crew to communicate to cabin crew and/or passengers information about impending turbulence or directions to be seated/restrained.

No. 604: Not adequately prepared for the task. Definition: Flight crew not adequately prepared with briefings, in assessment of weather factors or other information received, or not mentally prepared.

The Turbulence JSIT 's "Detailed Implementation Plan For Cabin Injury Reduction During Turbulence" recommende

NTSB Probable Cause

The flight crew's encounter with convective turbulence. Contributing to the accident was the failure of the operator's dispatch department to provide the flightcrew with current adverse weather information along the planned route of flight. Also contributing was the failure of the flightcrew to provide relevant forecast information to the cabin crew, and their failure to detect and avoid the existing convective conditions.