Flying Seasons

Icing Accidents

January 1, 2007

Learning from the stats

Lectures and texts on the dangers of icing are certainly edifying, but nothing beats a look at the accident record to make a more visceral impression. It's one thing to discuss icing from a theoretical standpoint, but quite another to contemplate the wreckage.

It may come as a surprise, but general aviation accidents involving structural icing are quite rare. In 2005, the NTSB cited a mere six icing accidents — and only two were fatal. As of this writing (November 2006), the totals for 2006 amounted to just two accidents, neither fatal. This tells us that most of us use sound judgment and stay on the ground when icing conditions are advertised. Perhaps all those texts and theories are doing their jobs: They've scared us so thoroughly that we don't dare fly if icing is even a remote possibility. Good!

So where and how do those few unfortunate GA pilots go so wrong? The findings strongly suggest poor preflight briefing practices. They also clearly show that icing is most frequently encountered in the climb or cruise phase of flight, and that the accident chain of events evolved from poor situational awareness, failures to take appropriate evasive steps, and just plain lousy decision making. For some insight, let's look at 2006's accidents to date.

According to an NTSB preliminary report, on January 13, 2006, the pilot of a Cirrus SR22 (which was not certified for flight into known icing) took off from the Fulton County Airport in Atlanta and flew to Alabama's Birmingham International Airport. The next leg of the IFR flight was to have been from Birmingham to Orlando.

The pilot took off from Birmingham at 3:44 p.m., and was cleared to climb to 7,000 feet. With the autopilot engaged, the airplane entered clouds at 5,000 feet. But upon reaching 7,000 feet, ice began to accumulate on the Cirrus. The pilot asked for a climb to 9,000 feet, and the request was approved. By 8,000 feet, the Cirrus reached the cloud tops in visual meteorological conditions, the report stated. But the pilot reported that the airplane began buffeting, and that airspeed had dropped to 80 knots. The airplane stalled, entered a spin, then descended through the clouds. The pilot deployed the ballistic parachute, and fell into trees. The pilot and passengers survived; the report didn't mention any injuries.

The pilot stated that he obtained a full Direct User Access Terminal system (DUATS) briefing the night before the flight. That briefing wasn't valid for the time of the accident, but mentioned airmets for icing between 3,000 and 8,000 feet. He received another, abbreviated briefing prior to leaving Birmingham, but the icing advisories mentioned at this time were set to expire at 3 p.m. Another airmet for icing was posted at 2:45 p.m., but the pilot stated he wasn't aware of it.

The Cirrus was equipped with XM Satellite Weather datalink weather capability, and the airmet was transmitted from the National Weather Service to the cockpit. Airmets are broadcast over XM Satellite Radio every 12 minutes.

Pilot reports also documented icing conditions in the altitudes below 7,000 feet over northern Alabama.

The pilot's qualifications included an airline transport pilot certificate, ground and flight instructor certificates, 12,773 total flight hours, 681 hours flying Cirrus SR22s, and 617 hours as an instructor in Cirrus SR22s. The pilot completed the Cirrus Standardized Instructor Program in March 2005, according to the NTSB.

The NTSB's determination of the accident's probable cause names inadequate preflight planning, a failure to obtain a current weather briefing, and a decision to operate the airplane into a known area of icing outside the airplane's certification standards. It's easy to come up with a couple more hunches of our own, based on the information. Good thing the Cirrus had that parachute!

The other accident took place at 2:10 p.m. on February 17, 2006, during a landing at the Greeley-Weld County Airport in Greeley, Colorado. The pilot, who holds an ATP certificate, was cleared for an instrument landing system (ILS) approach to the airport, but then performed a missed approach. The reason, he told air traffic control, was that his Piper Twin Comanche wasn't showing a "landing gear down" indication. He circled the airport while he attempted to manually extend the landing gear, was unsuccessful, and then elected to land wheels-up.

The weather at the time was reported as: ceiling 400 broken, 5,500 broken, 7,500 overcast; 1-mile visibility; temperature 2 degrees Fahrenheit; dew point minus 2 degrees F; and wind 040 at 16 knots.

During the subsequent landing, the airplane struck a sign, and the pilot and his passenger received minor injuries. Airport personnel stated that the airplane was covered with ice measuring approximately three-quarters-inch thick.

Also, the pilot said that halfway into the IFR flight from Sioux City, Iowa, he heard something vibrate, but after 5 minutes the vibration stopped. He said that en route temperatures were "minus 5 degrees." (The NTSB report didn't say whether this was Fahrenheit or Celsius, but either way this is a temperature conducive to icing — assuming the airplane is flying in visible moisture.) A post-crash investigation showed no anomalies with any of the airplane's systems.

The NTSB said the accident's probable causes included the pilot's flight into adverse weather, airframe icing, and the gear-up landing. Contributing factors were listed as the icing conditions, frozen landing gear, and a frozen alternate gear extension system.

These examples illustrate a larger trend: Airframe icing accidents happening to high-time pilots flying high-performance airplanes in conditions they knew — or should have known — were conducive to icing. And yet, they pressed on. In 2005, a Bonanza crashed at Barnum, Wyoming, with fatal results. Same thing with the nighttime crash of a Beechcraft King Air E90 at Rawlins, Wyoming, in 2005.

The lessons are multiple: Become thoroughly familiar with the weather before you depart; don't press on in deteriorating conditions; don't let your experience breed overconfidence; and don't be afraid to cancel a flight if your experience level, the weather, or your airplane's capabilities aren't up to the tasks at hand.

E-mail the author at tom.horne@aopa.org.

Thomas A. Horne

Thomas A. Horne | AOPA Pilot Editor at Large, AOPA

AOPA Pilot Editor at Large Tom Horne has worked at AOPA since the early 1980s. He began flying in 1975 and has an airline transport pilot and flight instructor certificates. He’s flown everything from ultralights to Gulfstreams and ferried numerous piston airplanes across the Atlantic.