Flight History
On November 11, 2014, about 1355 central standard time, a Bell 407 helicopter, registration N373RL, was landed on the Gulf of Mexico, Louisiana, following a droop in engine power. The pilot and three passengers were uninjured. The helicopter sustained no damage during the water-landing. The helicopter was registered to and operated by Rotorcraft Leasing Company LLC, under the provisions of 14 Code of Federal Regulations Part 135, as a passenger flight. Day visual flight rules (VFR) conditions prevailed for the flight, which operated on a company VFR flight plan. The flight originated from Viosca Knoll 989 (VK 989), an offshore platform in the Gulf of Mexico, and was destined for Main Pass 301 (MP 301), another offshore platform in the Gulf of Mexico.
According to the operator's incident report, the pilot reported that approximately 30 seconds after takeoff at about 400 feet above sea level, a series of compressor stalls and engine surges began. The pilot adjusted the collective pitch and began a slow descent. After lowering the engine power, the surges and stalls ceased and the pilot's plan was to attempt to fly the helicopter back to VK 989. At approximately 250 to 300 feet, the pilot began increasing the collective to regain some power, but the engine surges and stalls reoccurred. The pilot reported hearing the low RPM horn and when he observed the rotor RPM gauge (Nr), the Nr was about 90 percent and the power turbine gauge (N2) indicated it was running high, at or near redline. At this point, the pilot made the decision to land the helicopter in the water. He fully lowered the collective to salvage the RPM. The engine was still surging at flat pitch so he rolled the throttle to idle and entered an autorotation. The pilot prepared the passengers for the landing and radioed a mayday notification to the operator's flight following station. He then activated the float inflation handle, pressed the aircraft quick position button, flared the helicopter, and landed on the water. The pilot estimated from the time of the initial compressor stall to water contact was approximately 15 to 30 seconds. The helicopter did not sink. The pilot subsequently deployed the life rafts and got a verbal response from all the passengers that they were "ok." He directed the passengers to collect their belongings and a first aid kit. They got into the life raft on the left hand side of the helicopter.
Personnel Information
The pilot, age 30, held a commercial pilot certificate with airplane single engine land, rotorcraft-helicopter, and instrument helicopter ratings. His most recent second-class medical certificate was issued on December 9, 2013, with no limitations. The most recent pilot's flight review was accomplished on January 20, 2014. According to the operator, the pilot had accrued a total of approximately 2,211 hours of flight time, including 1,064 hours as pilot-in-command in the Bell 407. He accumulated 145 hours of flight time in the Bell 407 in the 90 days prior to the accident and 67 hours in the 30 days prior.
Aircraft Information
N373RL was a 1999 Bell 407 helicopter with serial number 53373. The single-engine helicopter was powered by a Rolls-Royce model 250-C47B turbo shaft engine (serial number CAE847835) driving a four-bladed main rotor and a two-bladed tail rotor. The engine's type certificate data sheet indicated a takeoff rating of 650 shaft horsepower for five minutes and 600 shaft horsepower for continuous operations. The helicopter was configured to carry one pilot and six passengers. The operator reported a maximum gross weight of 5,250 pounds and a weight of 4,306 pounds at the time of the incident. The helicopter was maintained in accordance with an approved inspection program; its last inspection was completed on October 27, 2014. Total time at the incident was 7,216 hours. A FAA major repair and alteration form dated February 6, 2012, showed the helicopter was fitted with a FDC Aerofilter engine inlet barrier filter.
The Rolls-Royce Model 250-C47B engine incorporates a Triumph Engine Control Systems model EMC-35R Full Authority Digital Electronic Control (FADEC) system that controls engine fuel flow via a Hydro-mechanical Unit (HMU) and Electronic Control Unit (ECU). The FADEC ECU contains two embedded processor systems. The primary system senses the pilot-controlled collective twist-grip throttle position (Power Level Angle) and other inputs to determine fuel flow. A reversionary system provides backup control. The HMU includes a gearbox-mounted fuel pump, motor-driven fuel metering valve, backup fuel control system, PLA input shaft, and feedback position sensors.
Meteorological Information
At 1350, the recorded weather at Houma-Terrebonne Airport near Houma, Louisiana, was: wind 190 degrees at 7 knots; visibility 7 statute miles; sky condition scattered clouds at 4,200 feet, broken clouds at 6,500 feet; temperature 26°C; dew point 16°C; altimeter 29.96 inches of mercury.
Tests and Research
Under the supervision of the National Transportation Safety Board (NTSB) investigator-in-charge, the incident engine was examined at Rolls-Royce near Indianapolis, Indiana, on December 2, 2014. The examination revealed that the engine was shipped without its ECU and T1 sensor. The engine was subsequently fitted with exemplar components and test-run in a test cell; it was found operational, with no request for maximum fuel flow observed. The bleed valve was found to close early prior to its schedule, and the engine was seven percent below new engine performance standards, which is two percent below the minimum required by the power assurance chart for in-service engines. Accelerations and decelerations did not produce surging.
The incident ECU (part number 115220-2A5-24, serial number JG8ALK0486) was sent to Triumph Engine Control Systems for repairs and examined under FAA supervision. The ECU sustained no physical damage and both tamper seals were intact. An acceptance test procedure was completed with no discrepancies; the unit passed a vibration test. However, a pressure check was not passed initially; the effect would be a lower acceleration schedule and an approximate five percent reduction in the maximum fuel flow limit. The sensor was recalibrated and subsequently passed the pressure test. Recorded incident data indicated an engine surge condition on the last flight, with N2 and Nr rpm joined throughout (no split). N2/Nr initially drooped to about 89 percent. Message coding showed the ECU requested maximum fuel flow but no expected increase occurred. When collective was reduced, N2/Nr returned to 100 percent. The data also showed the throttle was not rolled to idle until after touchdown on the water.
The bleed valve was bench tested by the operator, revealing early closure. Disassembly showed corrosion was present and corrosion flakes were found in the bleed valve exit orifice. Witness marks consistent with wear were found on the bleed valve bellows face near the exit orifice. According to the engine manufacturer, an early-closing bleed valve would only reduce surge margin in the Ng (N1) speed range where it would normally be open. The incident recorder data depicted Ng initially at 95 percent and decelerating to 91 percent, at which time the main rotor drooped enough to trigger the recorder (<92 percent Nr). Based on the bleed valve closure chart, the bleed valve would be closed at 95 percent and would not open until just below 88 percent for the given T1 (80°F). The bleed valve would have been closed per specification during the sequence. The engine was decelerating at that point, which is not a surge-inducing condition. The potential concern for surging due to a stuck-closed bleed valve is during engine acceleration. An early closing bleed valve could contribute to surging in the Ng range where it would normally be open; engine power at two percent below power assurance chart levels could contribute to surging; and a faulty ECU or T1 sensor could contribute to surging. Based on available information, it cannot be concluded that these factors had an effect on the subject engine during the incident.
Subsequent to the incident, the operator revised the bleed valve overhaul schedule from 1,500 hours to 750 hours to mitigate recurrence. Additionally, the operator's managers increased surveillance on engine water wash and chemical wash procedures to ensure proper techniques are used.