Eleven years after the October 1994 crash of American Eagle Flight 4184 in Roselawn, Ind., the FAA proposed a revision to Part 25 certification regulations that aims to prevent such icing accidents. The comment period for the notice of proposed rulemaking (NPRM) closed on February 2. Now the new rules will begin to wend their way through the FAA rulemaking process.

The proposed changes to 14 CFR Part 25 outline an ambitious goal: to revise certification standards for transport-category airplanes so that handling character-istics are the same in both icing and non-icing flight conditions.

The proposed rules answer some NTSB recommendations stemming from the Roselawn and other icing accidents, but they neglect airplanes certified under Part 23 regulations, which, according to one report, account for at least 80 percent of icing-related fatalities.

An NTSB study of icing-related accidents between 1982 and 2000 found that 819 people were killed in 583 airframe-icing accidents during that period. The study’s authors divided the accidents into three types of operation–Part 91, 135 and 121. The majority of the 583 accidents, 80.6 percent, involved Part 91 operations. Part 135 accounted for 17.6 percent and Part 121 only 1.7 percent.

Unfortunately, this study did not quantify the number of accidents by airplane certification category. It is likely that the Part 135 accidents involved aircraft certified to both Parts 23 and 25, while the Part 121 accidents probably all involved Part 25 aircraft. With such a high number of fatalities–470–attributable just to Part 23 (and older CAR 3) airplanes, the FAA’s rulemaking effort to improve Part 25 icing certification clearly focuses on operations that don’t have the majority of icing fatalities.

The proposed new Part 25 icing rules are the result of an aviation rulemaking advisory committee flight-test harmonization working group, which the FAA tasked with recommending “new or revised requirements and compliance methods related to airplane performance and handling qualities in icing conditions,” according to the NPRM.

The task was to focus only on transport-category airplanes, and one important piece of information the group discovered was the cause of icing accidents. Icing accidents, it determined, are not a result of degraded aircraft performance but happen because of “loss of control of the airplane due to the effect of the ice on airplane handling qualities.”

FAR 25.1419 requires airplanes to be able to operate “safely” when flying in icing conditions defined in the icing envelope found in Part 25 Appendix C. But as the NPRM pointed out, “Part 25 does not have specific airplane performance or handling qualities requirements for flight in icing conditions, nor does the FAA have a standard set of criteria defining what ‘to safely operate’ in icing conditions means in terms of airplane performance and handling qualities.”

The testing required to show that an airplane meets 25.1419 applies only when a manufacturer wants to include its airplane’s ice-protection systems during certification. Thus, an airplane without ice-protection features doesn’t have to be tested for airframe icing effects. European regulators address icing certification from the viewpoint that any airplane–whether or not it has ice-protection systems installed–can be affected by icing conditions.

In the early 1990s the Joint Aviation Authorities released draft Advisory Material-Joint 25.1419 (an AMJ is similar to an FAA advisory circular), which advises manufacturers on how to determine if an “airplane’s performance and handling qualities would allow the airplane to operate safely in icing conditions,” according to the AMJ.

That AMJ became, in April 1993, Notice of Proposed Amendment 25F-219, “Flight in Icing Conditions–Acceptable Handling Characteristics and Performance Effects.” This viewpoint and the knowledge that icing accidents are handling-related and not performance-related led the working group to recommend changing Part 25 to address handling qualities in icing conditions.

The proposed new icing certification rules generated few comments, most of them positive. The NTSB submitted a lengthy comment that was mostly supportive of the NRPM, except regarding one issue. FAR Part 25 Appendix C includes an icing certification envelope, but as the NTSB noted in 1996 in Safety Recommendation A-96-54 following the Roselawn crash, the envelope doesn’t include the type of icing American Eagle Flight 4184 encountered.

In its comment, the NTSB suggested that the FAA “expand the Appendix C icing certification envelope to include freezing drizzle/freezing rain and mixed water/ice crystal conditions, as necessary. The proposed rules do not address this recommendation, and the Safety Board continues to await the FAA’s proposed regulatory changes to 14 CFR Part 25 in response to A-96-54 to include SLD [super-cooled large droplet] conditions in the icing certification envelope.”

Nevertheless, the NTSB concluded, “The Safety Board believes that the proposed rulemaking is an essential step in improving the safety of flight in icing conditions for airplanes certified under Part 25.”

Ice Accretion Rule Changes

The Air Line Pilots Association (ALPA) submitted a comment supportive of the intent of the new regulations. But it expressed concerns, one in particular, about the NPRM’s proposal that handling characteristics meet certain standards during some amount of time while ice accretion occurs before pilots take action to remove the ice.

As ALPA noted in its comment, the FAA in 1999 issued 17 Airworthiness Directives mandating a revision to flight manuals, which requires pilots to activate ice-protection systems as soon as ice begins to accrete. ALPA pointed out that this advice is contrary to legendary aviation lore that claims pilots must wait until ice accumulates to a certain level before activating de-icing boots; otherwise the ice will form an irremovable “bridge” over the boot.

In 1999 the FAA convened a workshop on the subject of ice bridging, and the conclusion was that ice bridging doesn’t occur in modern de-icing systems and that there is no reason to delay activation of ice-protection systems as soon as ice begins accumulating.

ALPA’s point is that the proposed rules should recognize that no delay be allowed and that any new icing certification rules should incorporate this knowledge. ALPA also noted that in 2001, “After great investment of time and effort, the Ice Protection Harmonization Working Group developed and forwarded to the FAA an FAR Part 121 operating rule that eliminated delayed activation of ice-protection systems.

“This operating rule was based on and justified by an exhaustive review of accident and incident data. The FAA then developed an FAR Part 25.1419 certification rule change that did the same thing for future aircraft certifications. This document was eventually approved by ARAC [rulemaking committee] and returned to the FAA for action. Both of these actions were in response to NTSB icing recommendations following several significant aircraft accident and incident investigations. However, to date, neither of these proposed safety-based rule changes has been implemented by the FAA.”

The NPRM noted that the FAA is addressing supercooled large droplet icing conditions, both in terms of the icing environment and an airplane’s ability to operate in supercooled large droplet icing, through a rulemaking committee.

The FAA’s goal for this rulemaking is to make new airplanes handle and perform such that the flight crew doesn’t notice a difference between flying in good weather and in icing conditions. “There should not be any special procedures or methods used for operating in icing conditions,” the NPRM stated, “other than activating ice-protection systems.”

The FAA estimates that manufacturers will spend $66.4 million to comply with the new Part 25 icing regulations. This assumes that the rule becomes final in two years and that four new aircraft certifications would be affected by the new rules during a 10-year period.

While the NTSB study of 1982 through 2000 icing fatalities mentioned that the general aviation icing accident record is improving, icing remains a serious issue for Part 23 airplanes. And so far, there is no regulatory activity for Part 23 icing issues on the scale of these proposed new rules.

Under the NPRM manufacturers will have to:

• investigate the susceptibility of their airplanes to ice-contaminated tailplane stall;

• provide for adequate warning on the flight deck of an impending stall in icing conditions;

• demonstrate that their airplanes meet the same maneuvering capability and
handling characteristics requirements in icing conditions as in non-icing conditions; and

• show that their airplanes have adequate performance capability in icing conditions.