Get extra lift from AOPA. Start your free membership trial today! Click here

Accident analysis: Pushing the limits?

Flying into the teeth of a winter storm requires confidence, skill, and reliable instrumentation

Experienced instrument pilots can become comfortable flying into conditions their earlier selves wouldn’t have thought of attempting—especially if they’ve also moved up into turbine-powered aircraft with known ice protection. Even so, truly severe weather warrants a second and maybe even third look, not to mention the mental flexibility to change plans at the last minute if things aren’t working out.
Turbine Pilot Accident Analysis
Illlustration by Alex Williamson

Ugly weather

On the morning of February 18, 2018, the National Weather Service office in Salt Lake City issued a winter storm warning for Uinta County, Wyoming, that forecast eight to 14 inches of snow in “the most significant storm for the region all winter season.” Just after noon, the Salt Lake City Air Route Traffic Control Center (ZLC) Weather Service Unit issued a meteorological impact statement warning air traffic control facilities that:

“Widespread snow and strong winds continue through Monday as a strong cold front and winter storm move south across ZLC. Expect widespread turbulence with mountain waves and LLWS with blowing dust ahead of the front and blowing snow behind the front. Widespread icing with low clouds...mountains obscured and IFR conditions in snow...and areas of blowing snow with lower visibility.”

Fifteen minutes later, the Storm Prediction Center issued a convective outlook predicting at least a 10-percent chance of thunderstorms over Uinta County. Airmets were issued for IFR conditions and mountain obscuration, moderate icing below Flight Level 200, and moderate turbulence below FL180. The terminal aerodrome forecast for Evanston, Wyoming, forecast southwest winds of 43 knots with two miles visibility in snow showers and mist under a 2,000-foot broken layer. After 1 p.m., the winds were predicted to decrease to 27 knots, gusting to 37, with visibility decreasing to three-quarters of a mile in blowing snow with a broken cloud layer at 500 feet.

Those conditions might not have seemed prohibitive to a sharp, proficient pilot in a very capable airplane, but any pilot flying into them would need to be on top of his or her game and confident of the reliability of the airplane’s essential systems.

Flight summary

At 12:10 p.m. Central Standard Time (11:10 a.m. Mountain time), a 1997-model TBM 700A departed from Tulsa International Airport, Oklahoma, on an IFR flight plan to Centennial Airport in Englewood, Colorado. A little more than 30 minutes after takeoff, the pilot requested a change of destination to Pueblo, Colorado, about 80 nautical miles south of Centennial. Forty-five minutes later, he requested another destination change, this time to Utah’s Provo Municipal Airport some 355 nautical miles west-northwest of Pueblo, for “weather.”

Just under an hour and a half later, the pilot requested his third change of destination—to Evanston, Wyoming, which he said was below minimums when they took off but “looks better now.” It’s not clear how he came by that information; his only documented weather briefing was a ForeFlight package downloaded at 7:34 the previous evening, nearly 17 hours before takeoff. The controller cleared him to Evanston before handing him off to Salt Lake Center. After checking in, the pilot confirmed having Evanston’s current weather and requested the Fort Bridger transition to the ILS approach to Runway 23. He was advised to expect that and cleared direct to the Fort Bridger VOR (FBR) at an altitude of 15,000 feet.

Eight minutes later, at 2:33 p.m. Mountain time, the controller issued a series of low altitude alerts. Radar showed the TBM at 14,400 feet, 500 feet below the sector’s minimum IFR altitude. The controller made five transmissions on two different frequencies before the TBM’s pilot acknowledged the alert, adding that they were “bouncing around” as the autopilot tried to return to 15,000 feet. At 2:42 p.m. he was cleared to descend to 12,000, given the phone number of ZLC’s operations manager, and asked to call after landing.

Two minutes later he was cleared for the ILS with instructions to cross FBR at or above 10,000 feet and permission to change to the airport’s advisory frequency. The pilot’s response that he’d like to remain on ZLC’s frequency during the approach was the last transmission received from him.

Radar track data show that the TBM intercepted the localizer and began to descend on the glideslope. At about the time it crossed the final approach fix 5.9 miles from the threshold, the Evanston ASOS reported one-quarter mile visibility in snow and freezing fog with a vertical visibility of 800 feet. The TBM reached its minimum altitude of 7,200 feet—143 feet below decision height, and just 57 feet above the airport elevation—just over a mile from the field, then began to climb. However, instead of flying the charted missed approach procedure (a climbing left turn to intercept the 213-degree radial from the Evanston VOR), the airplane continued turning left through about 270 degrees, attaining its maximum altitude of 8,600 feet on a northeasterly heading before veering left, then right, and disappearing from coverage at 3:03:43 p.m. The last radar contact was at 7,900 feet msl, about 800 feet above the ground.

At 3:14, the Salt Lake controller asked a SkyWest fight in the vicinity to try to contact the TBM on Evanston’s CTAF frequency. The SkyWest pilot got no response.

On the Ground

Just before 3 p.m., the manager of the Evanston airport heard five clicks on the CTAF frequency, which he assumed was the TBM pilot trying to turn on the airport’s pilot-controlled lighting. That was followed by three clicks, then “an open mic for about 10 seconds. I thought I could hear a woman’s voice in the background, then the transmission ended.” Several local residents heard the crash and alerted emergency services, but by the time first responders located the scene most of the cockpit and much of the rest of the fuselage and wings had been consumed by fire. The pilot and his only passenger (his wife) were killed, their deaths attributed to “multiple traumatic and thermal injuries.”

The wreckage was found just beyond several broken trees less than a quarter mile from the last radar hit. The angle of impact was estimated at 60 degrees nose-down, and the debris field was oriented almost exactly due north (358 degrees). The landing gear was retracted but the flaps were still in their landing configuration, extended 34 degrees. The pattern of damage to the fan discs indicated that the engine was producing power at impact, but the severity of fire damage made it impossible to examine the avionics or determine the position of the cockpit switches.

In the panel

The airplane had been retrofitted with a Garmin G600 avionics package featuring dual 6.5-inch-diagonal LCD screens for the primary flight display and multifunction display. It was also equipped with standby attitude instruments.

The airport manager told the NTSB’s investigators about an earlier conversation he’d had with the TBM’s owner and pilot. The pilot recounted that during an ILS approach in marginal VFR the previous autumn, the LCD displays had gone black just a few miles from the airport, disconnecting the autopilot. The pilot had been unable to contact ATC, and instead used ForeFlight on his iPad to navigate back to the northeast and fly the approach again, this time successfully. While he’d told the airport manager that he planned to have an avionics shop troubleshoot the failure, the airplane’s logs held no record of any avionics work having been performed. The possibility of a panel failure during this approach could not be assessed because of the severity of the crash damage.

On the record

The NTSB officially attributed the accident to spatial disorientation in low visibility and turbulence, noting in the body of the report that investigators lacked the evidence to assess the role of a possible avionics failure. The apparent turn toward the northeast was consistent with the airport manager’s account of the earlier instrument failure, but not necessarily inconsistent with spatial disorientation as a principal cause. The 71-year-old pilot held commercial privileges for single- and multiengine airplanes and single-engine seaplanes and had logged 4,154 hours total time. He had flown the accident airplane for 90 to 100 hours. His career instrument experience was not reported.

Regardless of his experience and currency, the decision to fly into a howling winter storm should not have been taken lightly. The available information doesn’t tell us whether this pilot neglected his preflight briefing responsibilities, overestimated his skills, or simply suffered an equipment failure at the worst possible time. It does reinforce that the reduced margin for error in difficult weather is reason to stamp out all other sources of uncertainty—or reschedule the flight if that cannot be done.

ASI Staff

David Jack Kenny

David Jack Kenny is a freelance aviation writer.

Related Articles