Regulation on Flight in Freezing Drizzle to be Issued ‘Soon’

Fri, Feb 26, 2010 — David Evans

Articles, Featured

The glacial progress being made updating the standards for certificating aircraft to fly in icing conditions was criticized at a House Aviation Subcommittee hearing 24 February. The hearing came just a week after the National Transportation Safety Board (NTSB) held its deliberations on the “Most Wanted” safety recommendations. [See Aviation Safety Journal, “Time to Revamp ‘Most Wanted’ System”]

The FAA promises a regulation on SLD soon.

The FAA promises a regulation on SLD soon.

After numerous promises and delays, the Federal Aviation Administration (FAA) now appears on the cusp of issuing a proposed regulation regarding flight in supercooled large droplets (SLD), which are not now covered by regulations. As a result, aircraft are routinely dispatched into SLD conditions, for which on board ice protection systems have not been proven to cope.

It is not likely that the House committee will recommend legislation forcing this action, as the FAA appears ready at long last to issue its proposed regulation.

Below, extracts of witness testimony, outlining the unseemly long history of delay, obfuscation, bureaucratic concern for process, and apparent indifference to a known safety deficiency (SLD) into which airplanes are routinely dispatched for flight.

Testimony of Deborah Hersman Chairman, National Transportation Safety Board (NTSB) [Extracts]:

“From 1998 to 2007, the NTSB has investigated 50 Part 121 and Part 135, and 214 [Part 91] General Aviation accidents involving airplane icing, resulting in 202 fatalities. The accidents have involved aircraft, powerplants, aircraft systems (excluding carburetor icing) and/or runway and surface conditions. During the same period of time, the NTSB has issued 48 recommendations [re icing] addressing various safety issues … to improve aviation safety. While the NTSB relies on others to implement our recommendations, we have worked to educate the pilot community about some of the hazards associated with icing condition through Safety Alerts 9Ground Aircraft Icing – December 2006; De-ice Boot Activation – December 2008).

“The following two tables represent the range of investigations that the NTSB has addressed between 1998 and 2007. However, my testimony today will focus on the areas addressed in our Most Wanted list.


NTSB Accident Records 1998-2007
Involving U.S.-Registered Airplanes
Aircraft/Powerplant/Systems Icing (other than carburetor icing)
  Total Accidents FatalAccidents Fatalities
Part 121 1 0 0
Part 135 36 15 27
Part 91, 137, and 125 163 79 175
Total 200 94 202


NTSB Accident Records 1998-2007
Involving U.S.-Registered Airplanes
Runway/Surface Icing
  Total Accidents FatalAccidents Fatalities
Part 121 4 0 0
Part 135 10 0 0
Part 91, 137, and 125 54 0 0
Total 68 0 0


“A significant icing accident occurred in 1994 in Roselawn, Indiana, involving an American Eagle ATR-72 in-flight icing encounter and subsequent loss of control, resulting in a crash that claimed 68 lives. That accident investigation prompted the NTSB to re-examine the issue of airframe structural icing and conclude that the icing certification process continued to be inadequate because it did not require manufacturers to demonstrate the airplane’s flight handling and stall characteristics under a realistic range of adverse ice accretion conditions, including supercooled large droplets (SLD). [See Aviation Safety Journal, ‘Response to Crash Recommendations Characterized by Regulatory Foot Dragging’]

“The consequences of operating an airplane in icing conditions without first having thoroughly demonstrated adequate handling/controllability characteristics in those conditions are sufficiently severe that they warrant a thorough certification test program, including application of revised standards to airplanes currently certificated for flight in icing conditions.

“On January 9, 1997, a Comair Embraer EMB-120 departed controlled flight and crashed in icing conditions over Monroe, Michigan, while on approach to Detroit Metropolitan Airport. All 29 people on board were killed. The investigation of this accident, and of several control upset incidents that occurred prior to the Monroe accidents that involved a delayed activation of the deice boots, revealed a widespread and firmly held, but incorrect, belief within the aviation community that the activation of deice boots should be delayed until ¼ to ½ inch of ice builds up to prevent ‘ice bridging.’ As a result, in many cases, flight crews do not activate the deice boots until after the build up of dangerous accumulation of ice on the aircraft.

Aftermath of the Comair icing crash.

Aftermath of the Comair icing crash.

“In both the Roselawn and Monroe accidents, the pilots were using the autopilot before the icing-induced upset began. Because the pilots were not manually flying the aircraft, they were not aware that the autopilot was having increasing difficulty maintaining stable flight until the autopilot suddenly disconnected, and the airplane entered an uncontrollable flight regime.

“NTSB recommendations stemming from the Roselawn and Monroe accidents called on the FAA to use current research on freezing rain and SLD to revise the way aircraft are designed and approved for flight in icing conditions; to apply revised icing requirements to currently certificated aircraft; and to require the pilots of airplanes with pneumatic deice boots to activate the boots as soon as the airplane enters icing conditions. The FAA referred this work to an Aviation Rulemaking Advisory Committee (ARAC) more than 10 years ago. Six years later, the ARAC approved a concept to revise the icing design requirement for new airplanes. In December 2005 the ARAC completed its final report and recommended appropriate revisions to the design and operational requirements for flight in icing conditions. In the more than four years since that report, the FAA has yet to issue a notice of proposed rulemaking (NPRM) to require consideration of more realistic icing conditions. …

“Unfortunately, the NTSB continues to investigate accidents and incidents in which in-flight icing has been a factor. Each of the following incidents are accidents reinforces the need for the FAA to address SLD in icing certification:

— On March 19, 2001, an Embraer EMB-120, operated by Comair Airlines as flight 5054, encountered icing conditions while in cruise flight at 17,000 feet and departed controlled flight, descending to an altitude of about 10,000 feet. The pilots recovered control of the airplane and diverted to West Palm Beach, Florida, where they landed without further incident. The 2 flight crewmembers, 1 flight attendant, and 25 passengers were uninjured, but the airplane sustained substantial damage to the elevators and horizontal stabilizer due to the high forces encountered during the recovery from the uncontrolled descent.

— On February 16, 2005, a Cessna Citation 560, operated by Martinair for Circuit City Stores, crashed about 4 nautical miles east of Pueblo Memorial Airport, Pueblo, California, while on approach in icing conditions. The 2 pilots and 6 passengers on board were killed, and the airplane was destroyed by impact forces and post crash fire.

— In January 2006, an American Eagle flight 3008, a Saab-Scania AB SF340B+, departed from San Luis County Regional Airport, San Luis Obispo, California, destined for Los Angeles International Airport … After it encountered icing conditions during the en route climb, the aircraft departed controlled flight and lost 5,000 feet of altitude before the pilots were able to recover control. …

“Finally, the NTSB is investigating an accident that occurred in January 2009 in Lubbock, Texas, involving an ATR-72 that was on final approach to the airport. During the NTSB’s two-day public hearing in September 2009, factual information established that this accident involved a flap asymmetry and the autopilot disconnected when the stick shaker activated. The captain declined to perform a go-around, even when it was suggested by the first officer. The captain took control of the aircraft and it subsequently crashed 300 feet short of the runway and was destroyed in a post-crash fire. Airframe icing was noted by the flight crew. The National Weather Service had forecast light freezing drizzle, which by definition is an SLD condition. However, because severe icing conditions were not forecast or reported, the air carrier operating specification, approved by the FAA, allowed the flight to be dispatched into such conditions. …


“Although not an in-flight aircraft icing recommendation, the Safety Board has been concerned with the broader issue of excursions due to runway contamination. Performing landing distance assessments, which ensure an adequate safety margin for landing, is another important issue included on the Safety Board’s Most Wanted list. The recommendation asks the FAA to require operators to incorporate a 15% safety margin for landing on contaminated runways and was issued as a result of the NTSB’s investigation of a fatal runway excursion involving Southwest Airlines at Chicago-Midway Airport in December 2005.

“When the NTSB issues a safety recommendation with an ‘urgent’ designation, it expects that the action can be completed within one year after the recommendation is issued. However, in this case, the FAA has only issued guidance and encouraged operators to conduct a landing distance assessment. The FAA has not made this a requirement and recent investigations have revealed that some of the FAA’s inspectors are not aware the guidance exists. Since the guidance was issued, the Safety Board has investigated several accidents involving runway overruns on wet or contaminated runways, including Shuttle America flight 6448 in Cleveland; Pinnacle Airlines flight 4712 in Traverse City, Michigan; a Hawker Beechcraft Part 135 flight in Owatonna, Minnesota; and, we are supporting the Jamaican authorities as they investigate the recent American Airlines runway excursion that occurred on December 22, 2009, in Kingston. (See Aviation Safety Journal, ‘Ten Years Between Accidents & Ten More Years of Half-Measures’)”

Testimony of John Hickey, FAA Deputy Associate Administrator for Aviation Safety [extracts]:

Ground icing: … Currently, the FAA prohibits takeoff unless the airplane’s critical surfaces are completely clear of wintry precipitation. …

In-flight icing: … According to FAA regulations, any pilot who finds himself or herself in icing conditions while operating an aircraft that is not approved for operations in icing must immediately exit the icing conditions…. [See Hersman above for FAA allowing aircraft to be dispatched into SLD conditions.]

“(W)e have steadily worked to address two types of in-flight icing phenomena outside of the existing icing certification envelope: supercooled large droplets (SLD) and ice crystals … We expect to issue a Notice of Proposed Rulemaking (NPRM) to address this small area of vulnerability .… [SLD accounts for about 1% of all encounters, hence the reference to “small area of vulnerability,” but the effect of SLD is anything but small; notice that Hickey did not indicate when the NPRM will be issued]

“I am pleased to report that the SLD NPRM is now in executive coordination within the Department [of Transportation].” [FAA has repeatedly told the NTSB that the NPRM “is coming” and there have been delays for years now; the same promises are now being made to Congress.]

Testimony of Captain Rory Kay, Executive Air Safety Chairman of the Air Line Pilots Association (ALPA) [extracts]:

“We are all too familiar with the reasons why icing related accidents and incidents remain an important flight safety issue in the airline community. Historically pilots have been expected to decide when icing conditions exceed the capabilities of their aircraft and when it is acceptable to proceed into icing conditions. Such decisions would appear to be no different than the thousands of critical decisions made by professional airline pilots on every flight, every day, resulting in the extraordinary safety record we now enjoy. However, even the most cautious and experienced pilots have been involved in icing accidents or incidents. We must take note of this disparity and to continue to explore the reasons why. This is because the tools which pilots use to make these critical decisions have yet to be fully developed. In nearly every other aspect of airline operations, there is not only a significant body of knowledge, but also very advanced, redundant technologies which have eliminated ‘guess work,’ but that guess work is still inherent in flying safely in icing conditions or avoiding them altogether.

“We know the effects of flying too fast or too slowly and have safeguards, including specific speed limits, throughout the system to avoid those [flight] regimes. We know the impact of improper loading on performance, and have established firm loading limits to ensure our aircraft remain controllable. Limitations for in-flight icing, however, remain a difficult problem to solve for a number of reasons. First … we still do not fully understand the nature of icing in the atmosphere, how to assess the risk of a specific icing encounter from the flight deck, and most importantly, we do not yet have the means to avoid operating in conditions which exceed the capabilities of the aircraft’s ice protection system.

“Secondly, the nature of the actual atmospheric phenomena that we refer to generally as ‘icing’ vary widely and their effect on airplanes can be equally variable, making it extremely difficult to establish norms and limits for operations. Conditions acceptable for one airplane may very well prove hazardous for another airplane or even to the same airplane at a later time, but with only slightly different operating and atmospheric conditions. So we are faced with a dilemma. Just as inadequate training or limited experience may leave a pilot unaware of an icing hazard, considerable experience, but without hard data to know exactly what the experience was, can leave a pilot erroneously believing he or she is operating in conditions which appear to match conditions that have been previously safely negotiated….

“ALPA is concerned and disappointed that in the 15 years since that tragic icing accident near Roselawn new rules have not expanded to include all aircraft categories that are certified for flight in icing conditions. Despite available information from research and studies, pilots are still facing the same dilemma of having to make subjective assessments about flying into an icing environment ….

“An icing environment of particular concern is supercooled large droplets, or SLD icing which is defined as liquid droplets with diameters greater than 0.05 mm at temperatures less than 0º C. Research and testing have revealed that operations in an SLD environment can quickly deteriorate into a severe icing situation with ice forming in locations not normally prone to icing and perhaps overwhelming the ability of the airplane ice protection system to remove the ice from where it normally forms. Aircraft with inflatable deicing boots, typically used in small regional turboprop operations appear to be the most vulnerable to this phenomenon ….

“(W)e see operators routinely flying a wide variety of aircraft in conditions of reported freezing rain or drizzle without measures to prevent flight into SLD conditions that may lead to a severe icing encounter. For example, some carriers have a policy that takeoffs are prohibited in moderate or heavy freezing rain, and heavy freezing drizzle, while other carriers have no specific guidance or policy other than flight in severe icing as prohibited and a caution that flight in light freezing rain or freezing drizzle may exceed the capabilities of the aircraft’s ice protection system ….

“There are usually no specific or quantifiable limits provided to the line pilots to enable a determination when icing conditions have exceeded those evaluated during the certification process. This disparity creates a gap in information between icing flight conditions tested during the development and certification process versus in-flight icing conditions encountered during revenue operations. The line pilot is the one who must bridge this gap when he or she encounters environmental icing conditions that have not been evaluated in the design and certification process. ALPA firmly believes, and has repeatedly commented, that evaluation of these conditions should occur in the design and certification process [and] not on a revenue flight.”

Testimony of Gerald Dillingham, Government Accountability Office (GAO) [Extract]:

“NTSB’s February 2010 update on the status of its Most Wanted recommendations related to icing characterized FAA’s rulemaking efforts as ‘unacceptably slow.’ In December 2009, at FAA’s International Runway Safety Summit, NTSB’s Chairman [Hersman] commented, ‘How do safety improvements end up taking 10 years to deliver? They get delayed one day at a time … and every one of those days may b e the day when a preventable accident occurs as a result of something we were just about ready to fix.’ In particular, NTSB has expressed concern about the pace of FAA’s rulemaking project to amend its standards for transport category airplanes to address supercooled large drop icing, which is outside the range of icing conditions covered by the current standards.

“FAA began this rulemaking effort in 1997 in response to a recommendation made by the NTSB the prior year, and the agency currently expects to issue its proposed rule in July 2010 and the final rule in January 2012. However, until the notice of proposed rulemaking is published and the close of the comment period is known, it will be unclear as to when the final rule will be issued.”

In sum, assuming a final rule in 2012, it will take the FAA some 16 years from the issuance of the NTSB’s recommendation to cover SLD in aircraft certification criteria. This task should have been accomplished within six years. That time period would have allowed for due deliberation; taking 16 years defines the term “glacial progress.”

Comments are closed.

Nolan Law Group