Safety Pilot Landmark Accidents: Ice Crisis

Never let your guard down

November 1, 2007

More pireps needed

The AOPA Air Safety Foundation and the FAA General Aviation Safety Committee (ASF is a participant along with a number of industry and association members) have established a program to expand the number of pilot reports during periods of inclement weather. This winter, the Seattle Air Route Traffic Control Center, several area Tracons, Flight Service, and the National Weather Service will all cooperate in an experiment to solicit pireps during periods when icing and/or low ceilings/visibility are forecast.

These are factors in many general aviation weather accidents where pilots may not have had good information on actual conditions or had heard overly conservative forecasts in the past and elected to proceed into poor flying conditions. More pireps should improve the accuracy of the forecasts and provide timely warnings to pilots who may be inclined to ignore forecast weather. It's hard to make good decisions with poor information and hard to make good forecasts without verification.

If the test proves successful this program will be implemented in other air traffic control facilities. Visit AOPA Online for updates and surveys.— BL

One of the chilling challenges to operating light aircraft in the winter is airframe icing. Those who have ever collected an appreciable amount of ice on their airplane will avow that it's nasty stuff and does horrible things to the flying qualities of every aircraft. Our ability to forecast airborne ice is improving; however, the omnipresent airmets from October to April in northern climates frequently miss the reality. The assumption is that ice will be present when the temperature, in clouds or precipitation, is below freezing—an overly simplistic view on the reality of ice. It is so frequently over-forecast that when there are no airmets for icing, some general aviation pilots let their guard down, and that attitude may lead to disaster.

No icing reported or forecast

The Cirrus SR22 pilot contacted the Reno Flight Service Station (FSS) at 4:20 p.m. Pacific Standard Time on February 6, 2005. He received a standard weather briefing and filed an IFR flight plan from Reno/Tahoe International Airport, Reno, Nevada, to Oakland, California. The filed routing was from Reno, via the Mustang VOR (FMG), Victor 200 to Truckee intersection, Victor 392 to Sacramento VOR (SAC), then direct to Oakland at 12,000 feet. The briefing included current and forecasted weather for the Reno area, the intended route of flight, and the destination. An east-west stationary front was located in the Reno area. The briefer advised the pilot that there were no pilot weather reports (pireps) for the intended route of flight and no airmets or sigmets. The freezing level in the Reno area began at 6,000 feet with no precipitation. There was an airmet about 50 miles north of the Reno area that did warn of icing from the freezing level that began at the surface sloping to 5,000 feet, up to 18,000 feet. The pilot indicated he might request 14,000 feet once airborne. The minimum instrument altitude for the initial segment along V200 and 392 was 11,500 feet.

The flight departed Reno about 5:50 p.m. The Reno weather at 5:56 was reported as: wind 020 degrees at 6 knots with 10 statute miles visibility, and overcast at 5,500 feet with temperatures at 6 degrees Celsius, dew point -1 degree Celsius, altimeter setting 29.83 inches. A 7-degree spread on the temp-dew point indicated relatively dry air at the surface, but below-freezing conditions were only a few thousand feet aloft based on the cool surface temperatures. There was still no mention of icing by flights in the area or in the forecasts. The National Weather Service's Aviation Digital Data Service (ADDS) Web site did show some potential for icing at 16,000 feet using Current Icing Product (CIP).

The Cirrus was acquired by radar immediately after takeoff, executed the Mustang Six departure, and climbed on a westerly course to the assigned altitude of 12,000 feet. At 5:57 p.m. the departure controller asked whether N286CD could accept 14,000 feet to allow for departing traffic out of Truckee. The pilot acknowledged, and the flight was then handed off to the North sector of Reno Departure. The acquiring controller issued a 270 heading to join V392, and at 6:05 p.m. the Cirrus was handed off to Oakland Center.


6:07:46—N286CD: "Oakland Center, Two-Eight-Six-Charlie-Delta, any chance I could go up to one six thousand, see if I can get above these clouds?"

6:07:53—Oakland: "Cirrus Two-Eight-Six-Charlie-Delta, affirmative. Climb and maintain one six thousand."

6:07:57—N286CD: "One four thousand for one six thousand, Six-Charlie-Delta."

6:12:24—Oakland: "Cirrus Six-Charlie-Delta cleared direct Sacramento."

6:12:26—N286CD: "Direct Sacramento, Six-Charlie-Delta."

6:13:40—N286CD: "Oakland Center, Cirrus Two-Eight-Six-Charlie-Delta. Uh, I guess this isn't gonna work. I'm still in the clouds, any chance of lower?"

6:13:47—Oakland: "Cirrus Two-Eight-Six-Charlie, uh, Delta, Roger. Stand by one."

6:13:50—N286CD: "Six Charlie Delta."

6:15:00—N286CD: "Oakland center, Cirrus Two-Eight-Six-Charlie-Delta. If I could go up at, uh, another two, three hundred feet I could get above these clouds."

6:15:06—Oakland: "OK. Do you want to go up or down?"

6:15:08—N286CD: "Uh, let me go up first so I could build up some airspeed if that's OK."

6:15:12—Oakland: "November Six-Charlie-Delta, climb, uh, actually, November Six-Charlie-Delta, maintain the block one six thousand through one seven thousand."

6:15:18—N286CD: "Maintain the block one six thousand one seven thousand, Six-Charlie-Delta."

6:17:19—N286CD: "Uh, I'm coming down, Six-Charlie-Delta."

6:17:20—N286CD: (unintelligible) "I'm icing up."

6:17:39—Oakland: "November Six-Charlie-Delta, uh, say again."

6:17:42—N286CD: "I'm icing up. I'm coming down."

6:17:56—Oakland: "OK, uh last transmission was, uh, unreadable."


At this point Oakland Center made several attempts to reestablish contact, which included contacting a Cessna 210 that was nearby to call the Cirrus, all to no avail. The Cessna pilot noted that an aircraft said he was icing up and going down, followed by static, but could not confirm whether it was N286CD. Center acknowledged and then asked the Cessna pilot whether he was picking up any ice. The Cessna pilot replied, "Negative, I went through snow and I'm on top right here....We just came out from underneath a layer about at we're between layers but all I got was snow. I did not get any ice the whole way out." The 210 pilot then attempted to contact the Cirrus on another frequency. Three minutes later another aircraft reported that it was picking up an emergency locator transmitter (ELT).

The Cirrus dropped off of radar almost immediately in mountainous terrain and hit a steep rock and snow-covered slope. The pilot, who was the sole occupant, was fatally injured. The ballistic recovery system (BRS) parachute was deployed prior to impact and separated from the airplane almost immediately under extremely high loads. The placarded deployment speed is 133 knots indicated airspeed, but the physical evidence of the parachute risers showed the airspeed at the time of deployment was much faster.

Pireps after the fact

At 6:40 p.m., 20 minutes after the accident, the Oakland Center Weather Service Unit issued "[n]umerous reports of moderate to severe rime ice between 13,000 and 17,000 feet. Conditions developing and increasing in the area." Ironically, two reports of severe icing came from the accident aircraft just prior to the accident, passed presumably by the controller to the Center Weather Service Unit. Pireps included:

"6 p.m.—25 miles SE of Mustang VOR, Boeing 737, moderate rime during climb from 15,500 feet to 17,000 feet.

"6 p.m.—7 north CNO, Embraer 120, moderate clear 8,000 feet.

"6:15 p.m.—Over RZS (San Marcus on the coast near Santa Barbara), Cessna 182, light mixed, 9,000 feet.

"6:45 p.m.—25 miles NE of Reno, Turbo Commander 690, light rime 16,500 feet to 13,500 feet."

With the exception of the Cessna 182, all these turbine-powered aircraft were approved for flight in icing. Numerous other pireps started filtering in after the accident within 100 miles or so of the accident site.

The pilot

The 54-year-old pilot held a private pilot certificate for single-engine airplanes with instrument privileges. He had started flying in July 2002, received his private certificate in April 2003, and received his instrument rating in February 2004, one year before the accident. He had logged 473 hours with more than 100 hours in the last 90 days. More than half his flight time was cross-country, with 43 hours of night and nearly 75 hours of instrument time with 11 hours of actual weather.

He had completed the Cirrus factory school, and it appeared that he had just returned home to the Oakland area two days prior to the accident with his new Cirrus. The pilot had an estimated 69 hours in make and model with a current flight review issued December 29, 2004, six weeks prior to the accident.

The aircraft

No pre-impact malfunction was noted during the accident investigation, and the aircraft had well under 100 hours of total flight time. The airplane was equipped with ice protection, which was certified as a "No Hazard" system. This product is intended to allow an escape from unexpected icing conditions. The system holds about three gallons of deicing fluid that flows along the wing, horizontal stabilizer, and propeller blades.

The ice-protection system does not allow flight into known icing, and the pilot's operating handbook (POH) makes the limitations very clear that, "No determination has been made as to the capability of this system to remove or prevent ice accumulation." While pilots may be tempted to rely on this system, the guidance in the handbook is unequivocal. "Flight into known icing is prohibited.... At the first indication of icing, the most expeditious and safest course of exiting the icing conditions should be taken."

POH guidance on the operation of the system: "Normal mode is selected when icing conditions are encountered and prior to ice accretion. Maximum system operating time is approximately 1 hour. Maximum mode is selected if ice has accreted to flight surfaces. Maximum system operating time is approximately 30 minutes. Warning! The ice protection system may not remove significant accumulations of ice if accretions are permitted to form with the ice protection system off."

We don't know from the data available in the accident investigation whether the system was serviced with fluid or whether the system was operational, since it is possible to airlock the system pump if it is allowed to run dry. Pump priming must be done on the ground prior to takeoff. If the pilot failed to activate the system early, the ice could easily have overwhelmed the aircraft.

The effects of ice on aircraft performance are notable. According to the POH, "During simulated icing encounters, stall speed increases of approximately 12 knots in the clean configuration and 3 knots in the landing configuration were observed. In addition, cruise speed was reduced by 20 KCAS and the airplane's rate of climb diminished by at least 20 percent." Since this aircraft was not turbocharged and was flying close to its operational ceiling of 17,000 feet, the decision to climb, even without ice, was a non-starter. In this case, going down really was the only choice. The aircraft also was equipped with a supplemental oxygen system, but it was impossible to tell whether it was in use.

By all of the facts that we have available from the NTSB, the pilot appears to have been current and proficient for the flight. There was nothing in the preflight weather briefing that would strongly lobby against taking the flight. Many mountain pilots do not like single-engine flight after dark because it puts a lot of eggs into the powerplant basket, and while there is no statistical correlation to engine failures and time of day, it's just something that they won't do.


Two observations: I'd like to see many more pireps offered and solicited regularly through the ATC system. When the weather is poor or unexpected, pireps can literally be lifesavers. We as pilots and ATC can and should do much better in this regard. Good and current information equates to flight safety—more trips safely completed and more trips appropriately cancelled or diverted.

Also, the pilot should have been much more aggressive with ATC when he began to encounter the ice. Given his low total time, it is unlikely that he fully understood the nature of his adversary. Only 10 minutes elapsed from the time the pilot asked for a higher altitude to the time when the airframe was so badly iced it could no longer maintain altitude.

In circumstances such as these, tell ATC you need a lower altitude immediately, the code word for "this is a big deal." Do not hesitate to declare an emergency if ATC balks even slightly. It is very seldom that certificate action is taken and even if it were, so what? We should be far more afraid of gravity and terra firma than of the FAA. Icing survival is something for factory schools and all CFIs administering primary IFR, transition, recurrent training, or flight reviews to include in their training scenarios.

Bruce Landsberg is executive director of the AOPA Air Safety Foundation.