Unfortunately, many more aircraft would repeat the same mistakes and attempt takeoff with contaminated wings. Much more recently wing contamination led to yet another fatal mishap, this time involving a Challenger 604. Witnesses stated that they had seen frost or ice on the wing surfaces of the Challenger prior to its departure from Birmingham, England, on Jan. 4, 2002. Other aircraft that had also been parked overnight accumulated moderate to severe ice coatings, and were de-iced during the morning.

The wreck of the Ebersol Challenger in Montrose

 However, neither Challenger crewmember requested de-icing. The takeoff appeared normal up to the moment of liftoff. Immediately thereafter, the aircraft started to bank to the left. Two seconds after liftoff the aircraft had reached 50 degrees left wing down. The last recorded aircraft attitude was approximately 111 degrees left bank and 13 degrees nose-down pitch. After the left wing had scraped the ground for 120 meters, the ruptured wing tank ignited. The five people aboard the aircraft died in the crash that followed.

Bob McIntosh, an Investigator-in-Charge of the prestigious "Go-Team" at the NTSB, has investigated many icing accidents and notes that many of the pilots involved believed their wings were not contaminated. Investigators from Transport Canada and from Sweden (two countries that know quite a bit about operating in ice) also seconded this observation at the "SAFE Ground de-icing Conference" in Salt Lake City on June 15-17, 1993, which was held to examine the ground de-icing problem. This conference really was pivotal in standardizing the industry on anti-ice and de-icing.

Accident investigators from several nations also pointed out that pilots aren't adequately aware of the effects of minute amounts of airfoil contamination. Aeronautical engineers at Fokker conducted a series of tests in wind tunnels by rubbing sandpaper along the leading edge of a jet airfoil to simulate the rough surface of frost and found that the decrease in lift was a whopping 34 percent. However, that's just part of the problem. The critical angle of attack was decreased by 7 degrees, meaning that a contaminated airfoil would aerodynamically stall long before the artificial stick shaker activated. Furthermore, in a swept wing design, roll control would quickly be lost, as would be the natural pitch-down tendency of the wing. Instead, the wing would continue to pitch-up.

Boeing engineers found that frost on the slats produced nearly the same loss of lift as a fully contaminated wing, bringing out the immense importance of having a clean leading edge. However, before someone thinks that merely having a clean leading edge is "good enough," the Boeing engineers found that the loss of lift with a

Eyewitness Reports had this CRJ200's engines exploding at low-level

clean leading edge but frost on the rest of the wing still created an enormous loss of lift. Subsequent wind tunnel testing found that airfoils without leading edge devices are even more susceptible to lift degradation due to contamination than those with leading edge devices. This is especially true at times of rapid pitch changes such as those encountered at takeoff.

Aerodynamically speaking, the airfoil is not designed to carry any contamination on it without a loss of performance. To make matters worse, any asymmetric contamination can lead to one wing stalling before the other, leading to an uncontrollable roll at rotation. The "Clean Aircraft Concept," codified in the FARs (Part 121.629, 125.221, 135.227, 91.527) prohibits takeoff when frost, ice or snow is adhering to the wings, flight control surfaces, propellers, engine inlets or other critical surfaces of the aircraft such as flaps, windshields, stall vanes, pitot heads, static ports, ram air intakes, and any other probes or sensors.

Aircraft with fuselage mounted engines (which obviously includes most business jets) are particularly susceptible to shed ice from the inner root of the wing, as well as from the side and top of the fuselage. At rotation, the flexing of the wings can cause the ice to break loose and be immediately ingested into the engines. The problem of hard-to-detect clear ice forming over the wings (cold soak ice) occurs not only during overnights but also when the wing or the fuel is cold soaked and the aircraft descends into relatively humid conditions, or when rain falls onto the cold soaked surfaces. Clear ice indications are very difficult to detect from ahead and behind the wing during walk-around, especially when the lighting is poor or the wing is wet. Touching the wing's upper surface is necessary for a more detailed inspection.

In North America and Europe the common de-icing and anti-icing practice is to apply freezing point depressants (FPD), usually involving glycol-based fluids. Type I fluids can be composed of a mixture of various forms of glycol, and also contain wetting agents and corrosion inhibitors. Their primary purpose is for de-icing.

China Eastern's CRJ200 DFDR has been recovered and read

 They are effective for short taxi times with no prolonged delays and good in the absence of freezing precipitation. In cases when especially thick frost, ice or snow is adhering to aircraft surfaces, repeated application of Type I de-icing fluids may be required to completely de-ice the aircraft. This process can be accelerated by heating the fluids. The anti-icing protection provided by Type I fluids is minimal and takeoff should occur immediately after de-icing.

Type I fluids have relatively low viscosity, meaning that they flow off a surface rather readily. The freezing point of a Type I fluid varies with the content of water. A 50/50 mixture has a lower freezing point than pure concentrated glycol, and due to the lower viscosity, it flows off of the wing much better. Type I fluid should form a smooth, glossy film on the surface of the aircraft.

Type I fluids are effective when applied at heated temperatures. However, additional precipitation such as freezing rain or snow will tend to dilute this mixture and accelerate the refreeze process. For this reason, the anti-icing protection afforded by Type I fluids tends to be rather limited.

By contrast, Type II fluids are thickened to remain on the aircraft surface until the time of takeoff and normally serve as an anti-icing agent. Type II fluids have a minimum glycol content of 50 percent, with 45-50 percent water plus thickeners, wetting agents and corrosion inhibitors. The thickeners and wetting agents cause the fluid to disperse over the sprayed aircraft surface and acts like a protective cover. It is normally applied cold since this provides longer anti-icing protection.

One of the unique properties of Type II fluids is that their viscosity decreases dramatically as shear stress (created in the boundary layer by air flowing over the wing). When the aircraft is stationary or taxiing, the shear stress is minimal, so the fluid's viscosity remains high, keeping the fluid on the aircraft surfaces. However, as the aircraft accelerates for takeoff, the fluid viscosity rapidly declines and the fluid flows off the aircraft surfaces. These fluids were developed for transport category aircraft with a rotation speed in excess of 85 knots.

There's also Type IV fluids. These have additional thickeners, which allows the application of a thicker fluid film and thus have significantly better holdover time performance. These properties do not come without some penalty, however. Some Type IV fluids have been observed to form a thick jell during dry-out and when re-hydrated form a slippery film on the aircraft. Also, some have exhibited poor "flow off" at colder temperatures.

In light freezing rain or freezing drizzle, Type II and Type IV fluids produce far better anti-icing qualities than Type I. However, there are some limitations on the use of Type II/IV fluids. Since they were designed for jet transport rotation speeds, they are not recommended for general aviation aircraft and should not be applied unless the manufacturer has approved their use. Also, under no circumstances should Type II or Type IV fluids be applied to pitot heads, AOA sensors, control surface cavities, cockpit windows, aircraft nose, static ports, air inlets and engines. (Be advised that when extensive de-icing and anti-icing procedures are in effect at a busy airport, it can lead to fluid buildup on the runway takeoff end, which may significantly reduce runway braking action.)

Type II/IV fluids are particularly effective for protecting parked vehicles during extended periods of time when the fluid is applied to the top surfaces of wings, horizontal stabilizers, crown of the fuselage and vertical stabilizer. It should be noted that unless all surfaces are aerodynamically clean, the aircraft must first be de-iced with Type I fluid. All Type II/IV fluid must be removed with Type I fluid before departure if the parking time has exceeded four hours.

The "one-step procedure" is usually accomplished using a heated FPD and is intended to primarily remove frost, ice or snow. It should be noted that the one-step procedure will provide only a very limited duration of anti-ice protection. The heated mixture is used primarily for de-icing the aircraft surfaces, and has relatively limited anti-icing protection.

China Eastern CRJ200 Wreckage being pulled from the icy lake

Sometimes heated water is used alone to de-ice the aircraft. In this case caution must be exercised that the conditions aren't conducive to refreezing. For this reason, the industry has developed a practice of ensuring the remaining fluid film has a freezing point less than 18°F below the ambient temperature.

The "two-step procedure" involves both de-icing and anti-icing. The de-icing step is usually accomplished with hot water or a water/FPD mixture. The second (anti-icing) step involves applying a mixture of Type II or Type IV fluid to the critical surfaces of the aircraft.

Off We Go?

So de-icing an aircraft is a simple deal and off we go, right? Wrong! If improperly applied, de-icing and anti-icing fluids can cause potentially dangerous changes in aircraft performance, stability and control. Anti-icing fluids themselves are somewhat "thick" and are intended to remain on the aircraft surfaces at slow speeds to help prevent ice or snow from accumulating too quickly. As the aircraft speeds up to rotation speeds, most of the fluid should shear from the aircraft. However, some will remain and will lessen the lift by a small but distinct amount. Thicker solutions of Type II fluids result in approximately a 10 percent loss of lift at rotation to roughly 4 percent by the time of gear retraction. The stall speed of a de-iced/anti-iced wing does increase, thus reducing the stall margin (but the decrease is deemed far preferable than the danger posed by snow, ice or frost contaminating an airfoil).

If the temperatures are real cold, even Type I fluids would be so viscous that a typical regional turboprop would have a wing still contaminated by the fluid and would stall at rotation. Interesting enough, research by a Canadian manufacturer found that due to the degradation in airfoil performance with de-icing fluids present on the wing, an additional 2 to 2.5 degrees of AOA would be needed to climb. Hence, the manufacturer recommended keeping the aircraft on the ground until achieving a faster rotation speed to ensure enough shearing force cleaned off the wing. (Some aircraft manuals recommend increased rotation speeds when the aircraft has been de-iced. Don't forget this increases required field length.)

Optimally, de-icing should occur so that final treatments are applied just prior to takeoff. When possible, this means using remote sites near the takeoff position so that de-icing, anti-icing and final inspection is done as close as possible to takeoff time. Flight crews should use FPD fluid types and concentrations that will delay ice formations for as long as possible under the prevailing environmental conditions.

Under normal conditions, de-icing of the aircraft should begin at the leading-edge wing tip, sweeping in the aft and inboard direction. This avoids increasing the snow load on outboard wing sections, which could produce excessive wing bending stresses. However, if the tail has accumulated a heavy load of snow, it may be advisable to remove snow from the tail first to prevent tipping the aircraft.

The fluid stream should not be aimed directly at the aircraft surface. Rather, it should be applied at a low angle to prevent damage to surfaces. It should not be used at a high angle to physically ram ice and snow off of surfaces.

Consult the AFM to determine the appropriate slat and flap positions during de-icing. The same applies to the position of the horizontal stabilizer so that fluids will run off rather than into interior bays. Under no circumstances should the spray be directed at the trailing edges of the control surfaces. This can force partially melted contaminant into the hinge mechanisms with the risk of later re-freezing.

The fuselage should be de-iced from top down. de-icing the top of the fuselage is particularly important on aircraft with aft-mounted centerline and fuselage engines. Fluid flowing down from the fuselage top should be sufficient in most cases to de-ice the windows. Direct spraying of the windows is to be avoided since it can cause thermal shock resulting in cracking or crazing of the windows.

Balance bays, control cavities and gap seals should be inspected to ensure cleanliness and proper drainage. Contaminants must be removed from the seals to prevent freezing, which could subsequently impede control movement.

Engines deserve equal attention during icing conditions. Obviously installing engine covers is the best method for preventing ice and snow from entering the engine compartment in the first place when the aircraft is parked in icing conditions. Engines should first be inspected for any signs of ice or snow on the fan blades, first stage stators, cowling and sensors. Any snow or ice should be removed, and then the engine should be moved by hand to ensure freedom of rotation.

Since FPD fluid vapors can be irritating, crews are advised to turn off all air intakes during the fluid application process. Fluid residue on engine fan or compressor blades can reduce engine performance or cause surges, in addition to increasing the possibility of glycol vapors entering the aircraft. Several manufacturers recommend advancing each of the thrust levers individually to clear residual fluid from the engine.

The "holdover time" begins with the first application of the de-icing fluid. Holdover time factors include ambient temperature, type of fluid used, concentration, and rate and type of precipitation. However, many other variables influence the effectiveness of FPD fluids. For example, cold soaked fuel or wing temperatures significantly lower than the OAT can decrease holdover time. Blowing winds, jet blast, high relative humidity or improper fluid temperature can substantially lower the holdover time. Holdover times are published with high and low ranges. The low end of the range represents estimates in heavy precipitation conditions. The high end of the range is an estimate for light precipitation. There are no holdover time guidelines for moderate and heavy freezing rain, for obvious reasons. Remember that holdover times are only estimates and cannot be used as the sole indicator that the aircraft is clean and safe for takeoff.

A post de-icing/anti-icing check should be performed during or immediately following the ground de-icing/anti-icing process. Some aircraft manufacturers specify which areas need to be inspected. These often include the wing and tail surfaces, to include the leading edges and flight control surfaces. It should also include the engine inlets, windshields, antennas, fuselage, fuel tank caps and vents, landing gear, APU and air intakes and exhausts, as well as instrumentation such as AOA vanes, pitot-static probes and static ports. Once it is determined the aircraft is clean and adequately protected, the aircraft should ideally be released for takeoff as soon as possible.

A before-takeoff check is mandatory any time an aircraft has been de-iced. This is a vital step to ensure that the critical surfaces are indeed free of contamination. This is a visual and/or physical inspection of critical aircraft surfaces and components as close as practical before takeoff, regardless of how much of the holdover time has elapsed. This is the final confirmation that the aircraft is contamination-free. Flight crews should directly examine the condition of the entire wing, if possible, from whatever position the crew believes will give an accurate indication of the condition of the critical surfaces. The wing should appear glossy, smooth and wet. Some indications of the loss of effectiveness include progressive surface freezing or snow accumulation, random snow accumulation, or loss of surface gloss. The flight crew should look for any evidence of melting and refreezing, as well as any ice that might have formed during taxi. If these signs appear, then the aircraft should be returned for additional de-icing.

Can the crew take off if the holdover time has been exceeded? That depends. The flight crew must conduct a Pre-Takeoff Contamination Check. Takeoff must occur within five minutes following the completion of that check. If takeoff is not initiated within five minutes, the crew must repeat the check. On the other hand, if the contamination check reveals contaminants on critical surfaces, then the aircraft should return to the de-icing pad. Can the de-icing crew just apply another coat of Type II/IV fluid on top of the existing film? No. If the holdover time is exceeded and the crew returns because they believe the anti-icing fluid layer had lost its effectiveness, the surfaces must first be de-iced with a hot de-icing mixture of Type I fluid and water before another application of Type II/IV fluid is made.

The air transport industry has come a long way in developing and instituting better methods for de-icing aircraft since the Air Florida accident. The current method in use by the industry has worked sufficiently for many years now.

I do worry about "gaps" in business aviation's de-icing procedures, however. I question whether linemen at some FBOs are adequately trained, particularly at locations that don't de-ice often, and whether the fluid is heated to the proper temperature, is the correct mixture, applied with the correct spray nozzle, and is being sprayed at the proper angles. Each of these is key to proper de-icing, and it's in a flight crew's self-interest to assure that those questions get the right answers every time the spray trucks roll.