Event Description
A pilot preparing for departure from a hospital's second-story rooftop helipad reported that after both engines were running, he placed both engine control switches to FLIGHT and verified no caution lights. He completed the DO/VERIFY checklist and set the rotor rpm switch to 102%. The pilot then pulled collective to hover, verified 102% rpm, and applied takeoff power with forward cyclic. As the helicopter initiated forward motion, the pilot heard a "ROTOR LOW" warning in his headset, and the aircraft began to settle. The tail rotor struck the helipad's concrete barrier, causing the helicopter to descend over a dirt courtyard bordered on the east by a building approximately 40 feet away. The aircraft impacted the building, spun 180 degrees, and came to rest on its right side. There was no post-crash fire. The pilot shut down both engines and assisted in evacuating the occupants.
Post-Accident Findings
An FAA inspector examined the accident site shortly after the event and found both engine control switches in the OFF position. Subsequent testing of various components and both engines revealed no anomalies that would have precluded normal operation. Data from the Enhanced Ground Proximity Warning System (EGPWS) indicated that the number 1 engine produced normal and abnormally high torque values during takeoff, while the number 2 engine showed zero torque but proper Ng values, consistent with the number 2 engine control switch remaining in the IDLE position.
Pilot Background and Training
The pilot had a total of 7,923 hours in all helicopters, with 15 hours in the accident make and model. He reported that during his transition ground school and flight training on the accident helicopter type, he had questioned instructors about previous accidents involving the position of engine control switches during the start sequence. A comparison of the Agusta Rotorcraft Flight Manual and the Agusta General Familiarization Training Manual revealed inconsistent instructions regarding when to move the engine control switches from IDLE to FLIGHT. The pilot also noted that the operation of the limit override switch was never demonstrated during his training and that it was in an awkward position; this switch was not activated during the takeoff sequence as the pilot was unsuccessful in locating it.