by Robert L. Sumwalt

Ground Deicing and Anti-Icing Issues in Air Carrier Operations

In the past 25 years there have been 35 air carrier accidents worldwide that have been attributed to inadequate ground deicing/anti-icing [Endnote 1]. Nineteen of these accidents have occurred in the United States. Following an air carrier ground-based icing-related accident in March 1992 at New York's LaGuardia Airport, renewed attention was cast upon the issue of air carrier deicing and anti-icing. Major symposia were held to discuss the subject. New regulations were implemented, along with revised air carrier operating procedures.

This increased attention motivated ASRS analysts to conduct a detailed study of ASRS reports involving air carrier deicing incidents. This study analyzed 52 reports that were submitted to ASRS between January 1986 and January 1993. Eighty-one percent of the reviewed reports involved air carrier jet aircraft and 19 percent involved air carrier turboprops. Although the reports reviewed by our analysts were limited to air carrier operations, the findings of this research should be beneficial to all operators and pilots.

Findings

A takeoff with contaminated wing/tail surfaces occurred in 52 percent of the reports in this study. Reporters described ensuing problems such as engine damage and/or failure due to ice ingestion, aircraft control difficulties and rejected takeoffs. Revealed one reporter after a takeoff with contaminated wings:

• "...the tower supervisor...said he [had] observed 2 bright flames coming from the rear of our 2 engines. He [then] immediately closed Runway 13 and called Departure Control to see if we were airborne and in radar contact. He said he was ready to alert emergency personnel and equipment if he did not get a positive report from Departure Control." (ACN 81196)

Others also described the narrow margins that kept their flight from becoming an accident statistic. For instance:

• "Upon liftoff (VR) on Runway 14, aircraft immediately went into right roll...Full left aileron required to stop roll...Stick shaker occurred [at] approximately V2 + 10 KIAS (157 KIAS)." (ACN 51525)

The majority of the problems found in this study could be classified into three major categories:

1. Problems with detecting/inspecting for ice during preflight inspection;
   
   
2. Problems with ice removal, or with initially verifying successful ice removal after deicing; and
   
   
3. Difficulties assuring that aircraft critical surfaces were free of frozen contamination before takeoff.

Following is a brief look at each of these categories and some of the problems we found.

Problems Detecting Ice During Preflight

Hey, I'm Only 6 Feet Tall

One quarter of the problems noted in our study were the result of difficulties detecting ice on aircraft wings during preflight. This ASRS report excerpt exemplifies one such problem:

• "The Captain and I both made one last walkaround [inspection]...We both agreed there was no frost. The takeoff was normal. However, as the flaps [were] retracted the control wheel began to shake."

The crew decided to make a precautionary landing.

"On the base turn the flaps were lowered...All vibration and control wheel shake went away. The landing was normal...Inspection of the horizontal stabilizer found frozen snow/ice on both sides near the fuselage, approximately 1 inch thick and about 12 inches from the fuselage outboard. This area cannot be seen from the ground." (ACN 104785)

We found six other reports in the data set that cited the elevated height of wing and tail surfaces as a major factor in ice inspection/detection difficulties. The solution is obvious, but essential to flight safety: Operators must ensure that ladders and work stands are readily available for ice inspections. For those areas that cannot otherwise be accessed, consideration should be given to using "cherry pickers."

Ice Hard to See

Shortly after rotation, the crew of one transport jet heard several loud engine compressor stalls. Following a precautionary landing the crew discovered engine foreign object damage (FOD). From the ASRS report:

• "F/O had inspected wings (tops) for ice and could see none, however it was dark during preflight." (ACN 79693)

In this instance, the suspected culprit was undetected clear ice that had peeled off of the wings at rotation and was ingested by the engine. Clear ice that forms on upper and lower wing surfaces on some aircraft as a result of supercooled fuel lowering the temperature of adjacent wing surfaces can be a serious problem because it is difficult to see with the naked eye.

Eleven percent of the reports in this study cited ice detection problems such as crews being unable to see ice due to poor lighting conditions, the transparent nature of clear ice, or ice that was otherwise hidden from view. Not being able to reach ice during a tactile wing inspection was also cited.

When looking for clear ice, the surest detection method is a close visual examination combined with a "hands on" check of the wing. This, combined with vigilance, proper lighting and ice-inspection equipment should ensure a high probability of detecting clear ice.

Schedule Pressure

• "This is another case of a Captain thinking she is helping the company out by cutting corners in an attempt to save time on a flight that is operating behind schedule...I was outside, ready to perform the preflight walkaround. I immediately noticed that the aircraft had picked up considerable ice [on the inbound] flight. I informed the Captain of my findings and suggested [that] she take a look and have it deiced before departure. She went outside, came right back in and to my amazement said, 'It looks OK to me.' "(ACN 107073)

A sound corporate safety philosophy, reinforced by clearly written policies, procedures, and management attitudes, can help relieve a crewmember's self-imposed (or management imposed) schedule pressure. It is human nature for many people to hurry their tasks in order to "get the job done," so it is imperative that management establish a corporate culture that encourages crews to set safety as their top priority.

Problems with Ice Removal

Half the reports in our study mentioned problems with ice removal and/or verification of ice removal. Thirteen reports cited problems of ice still remaining on aircraft critical surfaces after deicing was completed.

• "...the first officer discovered that even after deicing was accomplished there was 1/4 inch of ice adhering to the...flaps (trailing edge) so he called the appropriate parties and requested that the...wings be done again. The deicing procedure was repeated...and upon inspection we found that they had again missed the [ice] on the flaps." (ACN 61983)

In another incident, the aircraft was reportedly deiced three times. But as evidenced by this narrative, the third time is not always a charm:

• "The takeoff acceleration was normal and at about 105 knots we heard a pop and the ENG FAIL lights illuminated...The Captain initiated an abort...[At the ramp] maintenance looked at the number 3 engine and found damage and chunks of metal missing from several compressor blades...Later a passenger informed us that it appeared that the deicing crew had trouble removing ice from the...right wing. He said that 1 or 2 long strips of ice were left on the wing after the deicing crew failed to remove them." (ACN 78762)

Checks and Balances

We noted 12 cases where the flight crew relied on the deicing crew's statement or hand signals that deicing had been completed, and therefore, failed to verify ice removal for themselves. Relying on the deicing crew's assessment sometimes led to takeoff with contaminated wings. In other cases of improperly deiced airplanes, the situation was caught before takeoff.

One solution is to have an additional post-deicing/anti-icing check accomplished by someone other than the deicing crew. This quality control measure is similar to maintenance practices, where one mechanic performs work, but final inspection of the work is accomplished by another person. The following report narrative illustrates the need for an independent post deicing/anti-icing inspection.

• "It is company policy to accept the deicing crew's inspection when deicing is accomplished...[After deicing and] enroute to the runway a flight attendant...advised me that...the left wing had not been deiced...I went back to personally view the left wing. I was incensed to find [it] encrusted with approximately 1/2 inch of rime ice." (ACN 103567)

Similar as this next report is to the previous report, they nonetheless occurred four years and some thousand miles apart:

• "...deicing personnel via external intercom advised aircraft had been deiced and all surfaces were 'clean...' Before we...[reached] the runway...a passenger notified the flight attendant [that] the left wing was never sprayed." (ACN 229944)

Can this happen at your airline? Without a system of "checks and balances," the answer is...???

Procedural Problems

• "Number 3 engine failed at rotation. Strongly suspect improper glycol procedure caused failure...We now believe untrained personnel used improperly prepared solution to remove ice from aircraft." (ACN 79411)

Procedural problems were noticed in 13 reports. Referenced were problems such as failure of deicing crews to follow prescribed procedures, inadequately designed procedures for deicing and/or post deicing checks, poor communication between deicing crews and flight crews, improperly prepared deicing fluids, lack of reliable equipment, and inadequate staffing to conduct deicing. Another report excerpt:

• "The local policy is to use [an] unheated mixture of glycol in a 2.5 gallon garden sprayer to apply to the wing surfaces. This procedure does not have the heat to remove and melt ice formed on the wing..." (ACN 198247)

As for staffing, the following report addresses the hazards of understaffing:

• "Following deicing...the deicing unit hit the [aircraft's] horizontal stabilizer...throwing the driver out of the [deicing] vehicle. The vehicle continued to ram into the stabilizer for approximately 10 seconds until the driver was able to climb back in and regain control...The deicing was accomplished by only one [person], contributing to the likelihood of a mistake." (ACN 82502)

Assuring Wings are "Clean" Before Takeoff

Before takeoff, the pilot-in-command is required by Federal Aviation Regulations (FARs) to ascertain that his/her aircraft critical surfaces are free of frozen contamination. One- quarter of the reports in this study referenced problems with making this determination.

Oh, Say Can You See

• "The value of inspecting the wing for ice from inside the cabin, especially at night, is questionable. Type II deicing fluid is the consistency of warm honey and when it covers the cabin windows, very little can be seen through them." (ACN 229944)

In another ASRS report, a pilot commented:

• "[After being deiced with Type I fluid and then anti-iced with Type II fluid]...we started engines and taxied to Runway 4L. Five minutes prior to takeoff...the First Officer [went]...to the cabin to inspect the wings, as now required by FAA regulations. When the First Officer returned, he said it was impossible to see through the Type II fluid on the cabin windows. Not only could you not tell if there was snow on the wings, but you couldn't tell if the wings were on the aircraft...The requirement to look at the wings from the cabin five minutes prior to takeoff, and [having] deicing fluid on cabin windows, are not compatible. This is a requirement that cannot be accomplished by cockpit crew members from inside the aircraft." (ACN 235382)

As these reports indicate, inspecting wings from inside the aircraft can be difficult, sometimes impossible. To take the "guess work" out of making this determination, FARs require that the "pre-takeoff contamination check" be conducted from outside the aircraft. However, many airlines have received FAA approval to conduct this check from inside the aircraft. ASRS reports indicate two problems with requiring flight crewmembers leave the flight deck to check the wings. First is the previously cited problem of not being able to see the wings adequately. The following report highlights the second potential problem:

Cockpit Distraction

• "[After deicing and] approaching Runway 9...sent F/E aft to inspect wing and control surfaces. Tower cleared [our large turbojet transport] onto Runway 9 for takeoff. F/E absent. F/O responded to tower, 'need 2 minutes.' Tower responded with 'position and hold'...F/E returned with wings clear report as we positioned on Runway 9...To my best recollection, [we] received takeoff clearance... Advanced power toward takeoff thrust. F/O released brakes. Aircraft moved with slow forward movement...Then received transmission from tower...to discontinue takeoff."

As it turns out, this aircraft had begun an unauthorized takeoff -- a takeoff that risked collision with a landing aircraft on a one-half mile final to an intersecting runway. In retrospect, the reporters reflected:

"We were busy, occupied with [checking] deiced surfaces...Crew was trying to get aircraft out shortly after deicing and was apparently too centered on that task." (ACN 135674)

He's OK, We're OK: The Ice-Status Myth

Three reports indicated that some pilots try to gauge the amount of snow/ice accumulation on their wings by simply observing the wings of other aircraft. One of these reporters, a 20,000- hour air carrier pilot, was deadheading in the passenger cabin when he learned that his passenger seat gave him a better vantage point of the wings than those seats in the cockpit.

• "[After deicing] we then took our position in line and waited perhaps 45 minutes ...before the pilot announced we were number 2 for departure. I...stared closely at both wings...until I could make out that the wings were completely covered with a blanket of snow and ice, or both."

The airplane was then flown, apparently without incident, to its destination. But the reporter decided to ask a few questions.

"As I got off the plane...I asked if the Captain would...talk to me for a minute...I said that I just wanted to know how they knew that they were free of ice and snow without a crewmember checking. He replied...[that] they were observing the wings of the aircraft in front of them, and [that aircraft] was clean, so [his aircraft] had to be clean, also...I asked if he would be surprised to learn that his wings were completely covered when he made his takeoff..." The reporter concluded, "There would be no way for a crew to determine their own condition by trying to see what was on another plane 150-500 feet in front of them...Each crew [should] check [their] own situation before attempting takeoff."[Emphasis added.] (ACN 133082)

Other Findings

Oh, It's Just a Little Snow

Twelve reports indicated that sometimes pilots see snow/ice on their aircraft surfaces but erroneously believe that its amount is inconsequential, or that it will blow off during taxi or takeoff.

• "We elected to taxi out and takeoff believing the snow would quickly blow off when the takeoff roll began." (ACN 194669)

Another pilot report reveals the same line of thinking:

• "It was a dry, powdery snow which would easily blow off during taxi and takeoff and not adhere to the wings...We determined we did not need to get deiced." (ACN 199436)

In both of these reports, the pilots learned after takeoff that their assumptions were incorrect. The safest policy is to have all contamination removed before takeoff. Often, loose dry snow will not blow off during takeoff roll but may instead freeze solidly onto wings. Due to the venturi effect, airflow accelerating over the wing's upper surface will sustain a rapid temperature drop. Thus, loose snow may be quickly transformed into frozen wing contamination.

Crew Resource Management

This study also highlighted that Crew Resource Management (CRM) can have a valuable application for ground icing situations. In ASRS reports where ground icing problems were caught after the aircraft had left the ramp, usually it was the cabin crew who notified the cockpit crew of the problem. To increase the likelihood that problems are caught before takeoff, consideration could be given to training cabin crewmembers to recognize wing ice formation. Furthermore, all crewmembers could be taught and encouraged to clearly voice their concerns. Consideration could be given to developing an easily remembered "statement of concern" that could be employed by any crewmember. As an example, "Captain, I am concerned that ice is on the wings." Once this "statement of concern" was voiced by any crewmember, the Captain would be required to fully appraise the situation before takeoff.

Summary

Preparing an aircraft for takeoff when icing conditions exist or are suspected requires vigilance, careful planning and adherence to prescribed procedures. Awareness of potential pitfalls is also helpful by knowing what to expect. Hopefully this study will help to enlighten the unsuspecting.

Management must resolve to help flight crews and deicing crews by providing them with suitable tools for them to perform properly. These tools come in the form of hardware such as equipment and supplies, but also includes such things as well thought-out policies and procedures. And a healthy, well-advertised, and consistently practiced corporate philosophy of total commitment to safety is absolutely, positively essential.