MIA95LA084

The pilot stated that while maneuvering at 250-300 feet, video taping a boat, he heard three dull thumps and the helicopter began to settle and the engine appeared to lose power. He lowered collective and nosed over. The engine power seemed low and he elected to land on the water. Postcrash examination of the aircraft structure, aircraft systems, flight controls, engine assembly, and all engine controls revealed no evidence to indicate precrash failure or malfunction. Uncontaminated Jet-A fuel was found in the airframe and engine fuel systems. About 2 minutes of audio information from the video camera that was being used by the passenger-photographer on the helicopter was subjected to frequency analysis. This analysis showed the engine N-1 speed was steady at 92% until 6.99 seconds before water impact when it decreased steadily to flight idle. Rotor speed remained steady at 99.8% until the N-1 speed began to decrease to idle. Rotor speed then dropped to 92.3% and the low rotor warning horn activated at 96.3% rotor speed. No thumps as reported by the pilot could be heard.

On March 6, 1995, about 1600 eastern standard time, a Bell 206L, N164BH, registered to Biscayne Helicopters Inc., ditched into Biscayne Bay while maneuvering on a 14 CFR Part 91 aerial observation flight. Visual meteorological conditions prevailed at the time and no flight plan was filed. The airline transport- rated pilot and two passengers were not injured and the aircraft sustained substantial damage. The flight originated from Miami, Florida, about 15 minutes before the accident.

The pilot stated that while maneuvering at about 250-300 feet in a hover, to video tape a boat, he heard three dull thumps and the helicopters nose yawed. Engine power at this time was set at 55-65 percent torque with a 670-700 degree F turbine out temperature. The helicopter began to settle and lose power. He reduced collective and nosed over. The engine power seemed low and he elected to make a controlled landing on the water. As he flared to decelerate for landing he rolled the throttle to idle. He rolled the helicopter to the left to make rotor contact with the water to stop the rotor blades. After the helicopter came to rest in the water he and the passengers exited as it sank.

The right rear passenger-photographer stated they departed Watson Island and climbed to 500 feet. They flew to Virginia Key and observed the boat they were supposed to video tape. They descended to 250-300 feet near the boat and he began taking video. The helicopter kept descending after this and when they were 20 feet off the water the pilot called out that "they were going to get wet." The helicopter touched down on the water smoothly and the main rotor blades contacted the water. The helicopter rolled to the left and when it came to a stop he exited the broken left door window and helped the other passenger out. Before the accident he stated the flight was smooth and normal and he did not hear any unusual sounds.

Postcrash examination of the helicopter after recovery from the water indicated there was no evidence of failure or malfunction of the helicopter's structure or flight control systems. Each fuel tank contained Jet-A fuel. The airframe fuel filter and engine fuel pump filter contained uncontaminated Jet-A fuel. The left airframe fuel pump operated normally and the fuel transfer system operated normally. The right airframe fuel pump did not operate due to water damage.

Postcrash tear down examination of the engine assembly revealed no evidence to indicate precrash failure or malfunction of the engine assembly. All fuel and air line fittings were found secure. The engine fuel control, engine fuel pump, engine fuel nozzle, engine bleed valve, and engine governor were removed from the engine assembly and operationally checked on a test stand. Each of these components operated within normal specifications.

Fuel samples from the left front fuel tank and the airframe fuel filter were submitted to Panair Laboratory, Inc., for testing. The samples met the specifications for Jet-A fuel. Some water contamination was present in the left front fuel tank sample. The water contamination had chloride levels consistent with sea water. See attached Panair Laboratory reports.

At the time of the accident the passenger-photographer was operating his video camera. A copy of this video tape was forwarded to the NTSB, Office of Research and Engineering for analysis. The video tape contained 1 minute 40 seconds of video information and about 2 minutes of audio information. Frequency analysis of the audio portion of the video tape indicated the engine N-1 or gas producing compressor operated continuously at 92 percent until 6.99 seconds before water impact. At this time the N-1 speed began to decrease slowly and steadily to 62.5 percent, reaching this speed at 2.1 seconds before water impact. The engine speed remained stabilized at this speed until 2.2 seconds after water impact, when the engine shutdown. N-1 speed of 62.5 percent is the flight idle speed of the engine.

The main rotor system speed was measured at 99.8 percent until the N-1 speed began to decrease. The main rotor speed dropped to 92.3 percent, reaching this speed 2.3 seconds before water impact. The main rotor speed then increased to 95 percent until water impact.

At 5.7 seconds before water impact the low rotor warning horn activated. The N-1 speed was 78.2 percent and the main rotor speed was 96.3 percent. The horn normally activates at 97 percent main rotor speed.

Analysis of the video tape to identify "3 dull thumps" as reported by the pilot was performed. No thumps are heard until 2.4 seconds before water impact. At this time 2 moderately loud thumps are heard and the video scenes indicate these thumps occurred as the camera was being moved into the helicopter from outside. The engine N-1 speed during these thumps was about 62.5 percent and rotor speed was about 92.3 percent. See attached Sound Spectrum Study.

The two passengers were given NTSB passenger statement forms. The completed forms were not returned to NTSB as requested.