The third meeting brought together experts from all icing-related disciplines to discuss airframe icing in general. (The ATR-72 was fitted with improved wing de-icing boots and returned to service, and as a result of the first two meetings, the FAA issued 18 airworthiness directives (ADs) covering 29 different airplane types and their operation in the icing conditions for which they were certificated.)
The FAA aimed these airworthiness directives at turboprops with wing boots and unpowered control surfaces. The agency took this action because the accident investigation findings concluded that an ice dam had built up behind the boots, disrupting airflow over the ailerons, which caused "hinge moment reversal," the uncommanded movement of the aileron upward into the area of reduced air pressure that the ice dam created. With the autopilot in use, the pilots could not detect the degradation in roll control. Advisory Circular 91-61A, issued in late 1996, recommends that autopilots not be used in icing conditions.
In September, 1997, the agency issued Notices of Proposed Rulemaking for further issuance of ADs covering a larger group of airplanes with unpowered control surfaces. This group included some singles, some twins, and some business jets. If these ADs become law, they will result in changes to the pilots operating handbooks for the affected aircraft.
The conferees broke up into five working groups: Ice Protection and Detection, Operational Regulations and Training Requirements, Forecasting and Avoidance, Icing Environment Characterization, and Requirements For and Means of Compliance in Icing Conditions. I sat in with the Forecasting and Avoidance group because, in my opinion, its objectives apply to "little airplanes" as well as to turboprops. Each working group had specific objectives, but as we'll see they overlapped when the recommendations came out.
The buzz word at all working group sessions was SLD, for supercooled large droplets. It had become apparent that Flight 4184 had encountered icing conditions far worse than those contemplated by the regulations that govern icing certification. Those regulations, written in 1946, assumed that the largest water droplet that might be encountered in a cloud had a mean volumetric diameter of 40 microns, or 40 one-millionths of an inch. Modern scientists are well aware of the fact that larger droplets can exist in a cloud at temperatures below freezing. Several working groups requested in their recommendations that the NWS develop a commonly accepted definition for a supercooled large droplet and disseminate it to meteorologists and pilots. Such a definition would form a foundation for developing new certification standards as well as meeting operational requirements.
The existing certification standard, Appendix C to Part 25 of the Federal Aviation Regulations (which also applies to Part 23 aircraft), takes into account temperature and droplet size but does not mention time of exposure. A distance of 17.4 nautical miles in a stratus cloud consisting of 40-micron droplets constitutes maximum continuous icing conditions for certification purposes. Flight 4184 was holding at the time of the accident, which exacerbated the aircraft's time of exposure to the SLD conditions. Two conference recommendations address exposure to icing - improve the air transportation system to decrease the probability of a catastrophic icing event, and establish specific controller actions when icing conditions are reported.
The working groups generally agreed that developing common icing terminology is just as essential to operational safety as was the Pilot/Controller Glossary when it became apparent that pilots and controllers were interpreting the same words differently. When a meteorologist includes the words "light," "moderate," or "severe" in a forecast or warning, that assessment is based on looking at data and inferring possibilities. National Weather Service meteorologists do not yet have the ability to measure droplet size directly, although researchers are now testing several technological approaches. The forecaster does not know what types of airplanes will be flying into the forecast area and how ice accretion might affect each type; still, each pilot must accept the forecaster's characterization of severity at face value.
Like it or not, as pilots we are still getting our fundamental icing information from the "Four Panel Chart," or Moisture-Stability chart, that is based on twice-daily balloon runs from about 100 locations in the United States. Using these charts we can see graphic representations of freezing levels, air mass stability, liquid water content (precipitable moisture), and relative humidity - or we can ask the flight service station briefer to interpret the charts for us.
Twice a day has worked for decades while airplane speed and range have increased dramatically. Fortunately, the meteorologists at the NWS Aviation Weather Center in Kansas City receive information continuously from participating airplanes, from NEXRAD sites, from automatic observation stations, and from human observers. The NWS uses that continuous flow to develop AIRMETs and SIGMETs for our use throughout the day.
Two Web sites give pilots additional insight into the icing hazard, although neither is "official" enough to use as a defense if you find ice where none was supposed to be. They are www.awc-kc.noaa.gov/awc/Neural_Net_Icing.html and www.rap.ucar.edu/largedrop/.
Several working groups recommended that pilots be provided with some objective measurement of icing severity, the equivalent of a "Trace-Light-Moderate-Severe" meter on the instrument panel. At present, the Aeronautical Information Manual (AIM) defines "severe icing" as "a rate of accumulation such that deicing/anti-icing equipment fails to control the hazard," while FAR 91.527(c) says that no pilot may fly a (large or turbojet) powered airplane into known or forecast severe icing conditions. The logical inference is that the pilot of an icing-certificated piston twin can fly into conditions that are off-limits to a turboprop or jet and not have to divert until the boots can no longer handle the ice.
Manufacturers of deicing/anti-icing equipment and ice detectors say the development of an STC for a detection device that can activate deicing equipment when a certain level of icing occurs is nearing completion. There is some question in the minds of the Protection/Detection folks about spot sensors as compared to area sensors. This concern is appropriate because bridging behind boots and ice run-back away from spot detectors can be operationally critical.
The Operational Regulations and Training Requirements working group put a lot of emphasis on pilot reports in their recommendations, specifically recommending increased emphasis on making pilot reports clear and concise, adding precipitation type and altitude, and ensuring coordination between ATC, the FSS, and the company. (See sidebar on pilot reports.)
The Forecasting and Avoidance working groups consisted largely of meteorologists and icing researchers. A NASA pilot who flies a de Havilland Otter used for research said it's not difficult to find areas where SLD icing conditions exist. A common perception at the icing conference was that SLD conditions are rare and localized. The Otter is based in Ohio, and increased emphasis is being placed on research in the Great Lakes region.
Pilots in the working group were surprised to learn how much the NWS depends on pilot reports to validate their forecasts and computer models. Most flight instructors push pilot reports as a means of informing other pilots, but the NWS plots pilot reports on maps of AIRMET and SIGMET warnings to establish whether its forecasts turn out to be correct and adjusts its computer models accordingly. Long term, this means that if the general aviation community increases its output of PIREPs, forecasts will become more accurate. The Aviation Weather Center manager specifically requested null reports - reports that forecast weather was not encountered.
Icing Research
The Air Force decommissioned the KC-135 tanker that had been used for whole-airplane icing certification tests in April, 1996. That left the Icing Research Tunnel (IRT) at NASA's Lewis Research Center as the only public facility available for icing research. Unfortunately, the IRT is not yet capable of producing supercooled large droplets, and it is limited to investigating ice accretion on representative shapes.
Several universities (and NASA Lewis) operate airplanes for icing research in natural ice, and they are capable of measuring droplet size and the liquid water content of the atmosphere. The fruits of this research ultimately reach the operational level, of course. Pilots have no way of measuring droplet size or liquid water content and thus cannot establish whether the conditions they are flying in (or are forecast to fly in) exceed the known-icing certification envelope.
This is the major reason why the conferees made it clear that no immediate change to the certification regulations would be forthcoming as a result of the Flight 4184 crash - it is unacceptable to tell pilots not to exceed certain parameters without providing a means to measure them (although that is the state of affairs today). The proposed new ADs mentioned earlier tell us that if rain is visible or if water droplets splash or splatter when outside air temperatures are below 0? C, pilots should interpret those visual clues to mean they are encountering out-of-the-envelope conditions.
To this non-academic layman, the conference was a surprise. It seemed to me on occasion that members of one academic discipline reported working feverishly on research without solid knowledge about what another group was working on. Participants who visited other groups would return with "They don't even know what you're working on!" or "They're suggesting research that you're already doing!"
This opportunity to exchange information face-to-face is bound to accelerate development of in-flight icing strategies. Also, the international makeup of the conference emphasized the fact that icing conditions know no political boundaries, and that the results of research must be disseminated to other countries more quickly.
For the pilot of an airplane not certificated for flight into known icing conditions, the concerns about icing conditions expressed by the operators of fully certificated turboprops and jets should come as a wake-up call. Only now are we discovering how little we know about severe icing conditions.
For the pilot of an airplane that is certificated for known icing, the conference findings make it clear that icing approval is not blanket permission to fly for long periods of time when the ice begins to collect.
The bottom line is that pilots will see no immediate change in icing certification standards or certification regulations. Avoidance strategies are already being worked out by the ATC and traffic management people, and you can be sure that the word "ice" will start the controller in your sector searching for a way to get you out of it.
Reporting Ice
In discussing pilot reports in the Forecasting and Avoidance working group, participants noted that many pilots of airplanes not certificated for flight into known icing conditions are afraid to report icing encounters for fear of certificate action.
I asked Tony Broderick, the conference's keynote speaker, who at that time was the associate FAA administrator for regulations and certification, whether this was a valid concern. He said that a radio report of an icing encounter could not and would not result in certificate action unless the pilot continued to fly in icing conditions.
A spokesman for the National Air Traffic Controllers Association told me that controllers have no interest in the certification status of airplanes or pilots, and if an icing encounter were received the controller would take no action other than to pass it on to the flight service station. So, a pilot has no reason to hesitate if he encounters icing. First, take action to get out of it. Second, report the encounter to ATC without fear of reprisal.
What is "Known Icing?"
A fundamental question that arises when instrument pilots get together is "What is the definition of 'known icing,' and where can that definition be found?" The truth is that the term is not defined in Part 1 of the Federal Aviation Regulations, but the administrative law judges of the National Transportation Safety Board have developed a solid definition of "known icing" in their decisions on icing-related certificate actions over the years. That definition is what you will have to live with if a load of ice contributes to an accident or incident while you are pilot-in-command.
Beginning with a case in 1957, the NTSB has stated in its findings that when temperatures are near or below freezing and visible moisture exists, those are icing conditions. They have said further that because the flight service station network states the existence of those conditions in reports and forecasts that are available to pilots both before flight and while enroute, when those conditions are forecast by the National Weather Service and disseminated to the FSSs, the icing conditions become 'known" to pilots who are required to check such reports and forecasts while planning a flight.
In a 1993 case, in upholding a certificate action against a pilot who relied on pilot reports in making his go/no go decision, the Board made it clear that official NWS weather reports and forecasts take precedence over "anecdotal" (their words) pilot reports.
So it is clear that a forecast meeting the NTSB definition of known icing would have the effect of grounding all airplanes not certificated for flight into known icing conditions. That doesn't happen, of course, because icing conditions sufficiently severe to knock a small airplane out of the sky occur infrequently and because almost all pilots know enough about weather to stay on the ground when conditions conducive to icing are forecast at their planned flight altitude and along their planned route.
Related Articles
