History of Flight
On July 17, 2014, at 0142 mountain daylight time, an Agusta A109E helicopter, registration N507CF, struck a mesa near Newkirk, New Mexico. The commercial pilot, flight nurse, and paramedic sustained fatal injuries. A postimpact fire consumed the helicopter, which was destroyed. The helicopter, owned and operated by TriState CareFlight LLC, was on a positioning flight under 14 Code of Federal Regulations Part 91. Night visual meteorological conditions prevailed at departure, and a company visual flight rules flight plan was filed. The flight departed Santa Fe Municipal Airport (SAF) at 0051, en route to a hospital in Tucumcari, New Mexico.
The helicopter was reported missing and located by a local resident. The wreckage rested on the north side of a mesa, approximately 150 feet above the surrounding terrain. Fire damaged most of the fuselage. All major airframe and engine components were present at the site.
Personnel Information
The pilot, aged 46, held commercial pilot certificates for helicopter and instrument helicopter, along with flight instructor certificates for helicopter and instrument helicopter. He was issued a second-class medical certificate on September 17, 2013, with a limitation requiring corrective lenses.
Hired by TriState CareFlight on September 10, 2009, the pilot initially flew Agusta 119 helicopters, later transitioning to Eurocopter AS350s. On December 13, 2012, he was assigned as pilot in command of Agusta A109 helicopters. His estimated total flight time was about 6,167 hours, including 208 hours in Agusta A109s, over 410 hours of night flying, 75 hours of simulated instrument conditions, and 0 hours of actual instrument conditions. He had been stationed at the Santa Fe base for over 18 months and had flown extensively in the mountainous desert regions of Arizona, Nevada, and New Mexico since 2009.
His most recent check flight occurred on December 16, 2013, a combined Part 135.293 and 135.299 evaluation lasting 1.2 hours. The flight was conducted entirely under night vision goggles (NVGs) and included normal operations, emergency procedures, unusual attitude recovery, inadvertent IFR procedures, and NVG failure scenarios. Astrological conditions at that time featured clear skies and a full moon.
The pilot was qualified for NVG operations, with at least 162 hours of NVG flight time. His last recurrent NVG training was on December 16, 2013, with ground training completed that day. On July 16, he inspected and tested his NVGs, reporting no discrepancies. In the 90 days before the accident, he flew 8.3 hours of NVG time, with recent flights on April 23 (0.4), April 25 (2.4), April 27 (1), May 22 (2.6), May 26 (1), May 27 (0.2), June 18 (1.1), June 19 (0.6), and July 16 (0.8).
The pilot worked a 7-day on-call, 7-day off-call schedule. Prior to the accident, he had been on-call for 15 days. From July 2 to July 8, he was on-call from 0700 to 1900. After 24 hours off, he began a night shift on July 9 from 1900 to 0700. The accident occurred on his eighth consecutive shift, during which he had previously flown 0.6 hours.
Aircraft Information
The Agusta A109E helicopter, manufactured in 2000, was modified for helicopter air ambulance (HAA) operations. It was powered by twin Pratt & Whitney Canada PW206C turboshaft engines, each rated at 549 horsepower. The helicopter was certified for instrument flight rules operations. The last inspection, completed on July 4, 2014, combined 25-hour, 50-hour, and 150-hour inspections. After that, the helicopter flew 17.3 hours with two discrepancies: weak wheel brakes (replaced on July 7) and a transmission oil chip light illumination (maintenance cleaned slight fuzz from chip detectors on July 11). The helicopter was equipped with a Garmin GNS-530 GPS/NAV/COM, a SkyConnect Transceiver, and a panel-mounted Garmin GPSmap 396. The cockpit was modified for NVG use via supplemental type certificate, and the helicopter also featured an enhanced ground proximity warning system, auto flight system, and radar altimeter.
Meteorological Conditions
An NTSB Senior Meteorologist conducted a weather study. The National Weather Service Surface Analysis Chart for 0000 mountain daylight time (MDT) showed a surface trough over the accident site, a stationary front south of the site, and three low-pressure areas to the northwest, south, and southeast. Potential existed for clouds and precipitation due to combined lifting mechanisms. The Storm Prediction Center predicted a 15% chance of damaging thunderstorm winds or gusts of 50 knots in the vicinity from July 16 at 1845 to July 17 at 0600.
Aviation weather reports indicated a southerly wind component initially, shifting to northeast after passing south-southeast of Las Vegas, New Mexico. Cloud ceilings would have lowered as the flight approached Tucumcari. Upper air soundings suggested cloud formations between 5,000 and 8,000 feet msl, with possible rain showers and thunderstorms. Strongest wind speeds associated with microbursts or outflow boundaries were estimated at 45 to 53 mph. Low-level wind shear was possible between 5,000 and 6,000 feet msl, with clear air turbulence from the surface to 10,000 feet msl.
Doppler radar at Cannon Air Force Base (43 miles southeast) detected two outflow boundaries. The first moved east to west from 2124 MDT on July 16 to 0030 MDT on July 17. A second outflow boundary and wave pattern was visible on a later scan at 0235. Backwards trajectory analysis estimated that the first outflow boundary and associated convective activity would have been over the accident site at the time of the crash, increasing potential for wind shear, strong updrafts and downdrafts, reduced ceilings, and reduced visibility.
AIRMET Sierra, issued at 2045 MDT on July 16 and valid at the accident time, forecast instrument meteorological conditions with ceilings below 1,000 feet and visibility below 3 miles in precipitation and mist. The Tacumcari TAF, issued at 2334 MDT, forecast wind from 030 degrees at 11 knots, visibility 6 miles, vicinity rain showers, few clouds at 2,000 feet agl, and a broken ceiling at 8,000 feet agl.
The moon was waning gibbous with 65% illumination, but mid- and low-level cloud cover likely obscured moonlight at the accident time.
Wreckage and Impact
The helicopter came to rest midway up a 300-foot-high mesa at an elevation of 4,654 feet msl. The wreckage was concentrated in one area, with main rotor blade fragments and light debris scattered nearby. Postimpact fire consumed most of the structure. The main wreckage included main rotors, fuselage, tail boom, and tail rotors. Exterior panels and medical equipment were strewn around the site.
All linkages between the swash plate and pitch change horns were intact and cotter pinned. Tail rotor blades rotated when the tail rotor drive shaft was turned manually. Tail rotor linkages were intact. A boulder in contact with the tail rotor exhibited machining consistent with blade strikes. A boulder near the helicopter nose had circular scuffing and a main rotor blade tip wedged into the rock.
The airspeed indicator needle pointed at 190 knots (exceeding Vne of 168 knots). The left vertical speed indicator showed between 2,500 and 3,000 feet per minute descent. Other gauges were destroyed or unreadable. The lower portion of the helicopter was partially buried, and gear positioning indicated retracted landing gear. The emergency locator beacon was damaged by impact and heat. A Garmin GPSmap 396 was recovered and sent to NTSB laboratories.
Due to slope and thermal damage to rocks, the wreckage could not be fully examined on-site. A postaccident examination at a storage facility revealed flight control fracture signatures consistent with overload and/or thermal damage, with no preimpact anomalies. Engines showed no preimpact anomalies.
Additional Information
Company pilots recognized the area from Moriarty, New Mexico, to Amarillo, Texas, and from Wagon Mound to Fort Sumner, as having potential for low visibility even when weather stations reported clear weather. A photo of the Santa Fe base map is in the docket.
The operator used Generation III ANVIS 9 night vision systems, with a 40-degree field of view and 2,000 to 3,500 times light intensification. The pilot and one crew member were to use NVGs during the flight. Meteorological modeling indicated the helicopter likely operated without lunar or celestial light due to cloud cover. The helicopter lacked an infrared spotlight; a regular spotlight was incompatible with NVGs and likely unused.
Damage to the NVGs prevented determination of whether they were in the down position at impact.
The United States Army Field Manual 3-04.203 notes that low moon illumination or new moon cycles make desert night flying challenging, especially with night vision devices, and that unaided night flight is not recommended. The manual also describes NVG limitations: distorted depth perception, difficulty detecting adverse weather, reduced peripheral vision, and risk of spatial disorientation during aggressive maneuvers.
Communications
TriState CareFlight used the SkyConnect satellite system for communication. At 0050:48, the pilot reported departing for Tucumcari to the operations center. No further communications were received from the helicopter. The flight was conducted under company VFR and was not in contact with air traffic control.
Tests and Research
SkyConnect data tracked the helicopter from 0052, reporting positions every 30 seconds. The helicopter initially tracked at 100° heading, 9,500 feet msl (about 3,800 feet agl), and 160 knots groundspeed until about 25 miles southeast of Las Vegas, New Mexico. At 0120:13, the Garmin GPSmap 396 began recording, providing higher resolution data.
According to Garmin data, the first complete point at 0120:26 showed the helicopter tracking 106° at 9,237 feet msl, groundspeed 158 knots. At 0121, it began a descending left 360° turn, descending to 1,225 feet agl before resuming course. It then flew at about 6,600 feet msl (1,200-1,800 feet agl) at 130 knots. At 0136:17, about 30 miles west of Tucumcari, the helicopter turned right to south-southeast, descending to 6,000 feet msl, groundspeed increasing to 160 knots then reducing to 140 knots. At 0140:09, it continued right, flying south-southwest while climbing to 7,763 feet msl. At 0140:49, it turned left. As the turn passed 90 degrees, descent began. At 0141:21, descent rate was 940 fpm, increasing. The helicopter continued a tightening left 360° turn. The last point at 0141:47 showed 4,840 feet msl (200 feet agl), groundspeed 93 knots. Estimated final descent rate was 14,760 fpm.
Weather overlays with the GPS track indicated that when the helicopter performed the descending turn at 0121, it was near a predicted outflow boundary wave. It later flew between two boundary waves, and at 0141 likely entered the first wave as it began the final left turn.
Agusta engineers analyzed the GPS data for the final spiral turn. They concluded the helicopter remained within structural limits, with bank angles increasing to nearly 70° and load factor near 2.7 Gs. The helicopter was flying a high-speed, high-load descending turn, with the main rotor system approaching or reaching its aerodynamic limit.
The helicopter was tasked after a hospital requested a patient transfer. The operations center contacted the closest helicopter in Amarillo, Texas, but the pilot declined due to thunderstorms. At 0042, the accident pilot was contacted at Santa Fe, assessed weather for about 20 seconds, and accepted the flight. At 0050, he reported departure.
When tracking stopped, the operations center received a notification. Attempts to contact the crew via satellite and cellphones failed. At 0209, the on-call supervisor was informed. At 0234, search and rescue was initiated. Three company helicopters were dispatched; two were unable to proceed due to low visibility. One flew near the site but saw nothing, heard no ELT. At 0322, a ranch hand reported a fire. At 0355, New Mexico State Police confirmed the wreckage location.