This is the second of a three-part article series on analyzing a Bell 407 crash. The first part outlined the conditions before takeoff and the fatal flight.
Representatives from the FAA, Survival Flight, Rolls-Royce engines and Woodward were parties to the investigation. In addition, the Transportation Safety Board (TSB) of Canada was an accredited representative and Bell Textron a technical advisor to the investigation.
An airworthiness group documented the accident site.They removed the engine electronic control unit (ECU), some of the cockpit instrument gauges, and the main rotor and tail rotor hydraulic actuators for further investigation. Data downloaded from the ECU showed there was no engine failure and examination of the rotor actuators showed no abnormalities.
The Bell 407 helicopter, N191SF, was manufactured in 1996 and was powered by a single turbine engine. It was purchased by Survival Flight in 2017 and modified to an air ambulance configuration in 2018. It had 1,179.7 hrs. of flight time and 5,058 landings, and it had last been inspected a month before the accident.
The helicopter had no autopilot and was not certified for IFR flight. However, it was well equipped with avionics. It had a radio altimeter and a dual Garmin GTN 650 system. The Garmin GTN 650 incorporated a helicopter terrain awareness and warning system (HTAWS). There was a Garmin G500H electronic flight information system with primary flight instrumentation, navigation, a moving map, and flight information service–broadcast (FIS-B) weather. Also installed was a Garmin GTS800 traffic collision avoidance system, a Garmin GTX 345R transponder and a night vision system.
The two Garmin GTN 650s were recovered and an NTSB vehicle recorder lab specialist was able to power them up. The only information stored was the location of one airport. There were no stored flight plans in either device.
14 CFR 135.607 requires air ambulance helicopters to be equipped with an approved flight data monitoring (FDM) system. To comply, the accident helicopter had an Outerlink IRIS lightweight flight data recording system. It provided satcom communications, an internal attitude heading reference System (AHRS), and recording of certain flight parameters and intercom voice information. The IRIS system can also record video, but no cameras were installed on the helicopter.
The system was not designed to be crashworthy and the company did not routinely download and analyze the FDM data. However, much of the recorded data was recovered from the accident flight.
Unfortunately, the audio recording quality was poor. A specialist downloaded the audio file and a flight data group examined it. The Outerlink file provided only short and intermittent pilot remarks. Radio communication between the OCC and the pilot were integrated into the resulting transcript and it helps to show the timeline of events.
The flight parameter files were of better quality. Data was recovered from both onboard and satellite uplink sources. There was a two-minute data dropout near the end of the recording.
An NTSB performance specialist used data from the Outerlink FDM system, the ECU, and automatic dependent surveillance-broadcast (ADS-B) to reconstruct the flight path. In the last minute of flight, there were changes in speed, large control inputs and large changes in pitch attitude. Despite the data dropout, the specialist was able to construct a 5,900 ft. arc ending near the crash site, which showed the helicopter was in about a 22-deg. left bank as it descended during the final moments of the fight.
The performance study also compared collective pitch and rotor torque to determine if icing was affecting the rotor blades. Because that relationship did not change, the study concluded that icing was not affecting the rotor blades.
There was a 10 second whining sound recorded near the end of the audio recording. A sound spectrum study compared the noise to common helicopter sounds such as the rotors and gearbox to the sound of a bird striking the aircraft. The study concluded that the sound was aerodynamic in nature, either from a plenum, a horn type sound, or an open window.
The 32-year-old pilot had been hired in April of 2018, nine months before the accident. At the time she had helicopter and instrument helicopter ratings on her commercial pilot certificate, and she held a helicopter and instrument helicopter flight instructor certificate, as well. She had a second-class medical certificate with no limitations.
When hired, she had logged 1,855 flight hours, and since completing training at Survival Flight she had logged an additional 83.3 hours, all in the Bell 407. During her initial training at the company she had successfully performed a GPS approach, a missed approach, instrument navigation, unusual attitude recovery, and flat light, brownout and whiteout recovery procedures.
She had been working day shifts from 0700 to 1900 for the six days before the accident, and had spent each night at home. She was the base safety representative and was described by other pilots as very approachable and proactive about safety.
Survival Flight is headquartered at Batesville, Arkansas, and operated 15 bases in six states at the time of the accident. The company operated 16 helicopters and one Pilatus PC-12 airplane and employed 70 pilots. Only the PC-12 was authorized by the FAA to operate under IFR. The Grove City base, Base 14, opened in May 2018, seven months before the accident. It was staffed with four pilots.
Helicopter air ambulance operators with more than 10 aircraft are required by 14 CFR 135.619 to have an operations control center. Survival Flight’s OCC was located at the company’s headquarters in Arkansas. Staff at the OCC take flight requests from hospitals and first responders and forward those requests to pilots at the appropriate base.
On the morning of the accident, the OCC took a call from the Holzer Meigs medical facility requesting patient transportation to a hospital in Columbus, Ohio. The technician at Holzer Meigs had already called two other helicopter air ambulance companies and had been turned down. The pilot for MedFlight, called first, turned the flight down because the icing probability was greater than 75% and because snow squalls were present on the HEMS (helicopter emergency medical services) Weather Tool. The pilot for HealthNet Aeromedical Services turned the flight down due to low ceilings and icing.
A free website, weatherturndown.com, provides a means for medical transport services to see if other companies have refused to take a requested flight. The technician at Holzer Meigs did not mention she had been turned down twice, and the Survival Flight OCC specialist did not ask. The specialist said she was monitoring the weather turndown website, but apparently the two turndowns had not yet been logged by MedFlight and HealthNet.
When the Survival Flight evening duty pilot at Grove City took the phone call from his OCC, he took only 28 seconds to decide to accept the flight. He then contacted the accident pilot, who was on her way to the base, and asked if she would take the flight. He did not brief her on the current conditions, except to say there was “good weather.” He said in an interview he had been sleeping when the OCC call came in.
There is no evidence that either pilot checked the weather conditions for the flight before it took off. The accident pilot apparently accepted the judgment of the evening shift pilot and deferred completing the flight risk assessment she was supposed to do.
The final part of this article series will evaluate the company’s safety culture.
Related Content
