Landmark Accident: Sideways Slide

The takeoff

On December 20, 2008, Continental Airlines Flight 1404, a Boeing 737-500, was scheduled to fly from Denver to Houston around 6 p.m. Mountain Standard Time. Pushback and taxi were routine. The weather was VFR with strong westerly winds, although the current ATIS was only reporting 11 knots of crosswind. At 6:14 p.m. the tower controller cleared the Boeing to line up and wait on Runway 34R following a prior departure. At 6:16 p.m. one of the pilots asked about the winds.

At 6:17 p.m. the tower controller advised the wind was from 270 degrees at 27 knots, assigned a departure heading of 20 degrees, and cleared the flight for takeoff. The captain remarked, “Looks like…some wind out there.” The first officer replied, “Yeah,” and the captain stated, “Oh, yeah, look at those clouds moving.”

As the takeoff roll began, the captain said, “All right…left crosswind, twenty, ah, seven knots…all right look for ninety point nine.” This referred to the engine power setting. The captain focused outside to maintain the runway centerline. At 6:18:04 p.m. the first officer advised that the power was set and shifted attention to airspeed to make the standard callouts.

Walking the rudder

As the airplane accelerated, the flight data recorder indicated increasing right rudder while the elevator and ailerons remained in their neutral position. At 6:18:07 p.m., as the airplane accelerated through 55 knots; the nose began to move left; and the flight data recorder recorded a large right rudder input that peaked at 88 percent of available forward travel.

This was relaxed to about 15 percent by 6:18:09 p.m. as the left aileron deflected. The nose moved to the right; however, at 6:18:10 p.m., accelerating through 85 knots, it began moving left again about 1 degree per second. This continued for two seconds as the captain countered with strong right rudder of 72-percent displacement at 6:18:11 p.m. It then decreased to 33 percent at 6:18:13 p.m. Airspeed was now above 90 knots.

“During this second large right rudder pedal movement (at 6:18:12), the airplane’s left turning motion slowed for about 1 second, and then the nose began moving rapidly to the left again,” said the NTSB. “A fraction of a second later (at 6:18:13.25), the right rudder pedal was abruptly relaxed (reaching its neutral position about 1 second later). At 6:18:13.5, the CVR recorded one of the pilots exclaiming, ‘Jesus.’ According to the captain, after the airplane left the runway he subsequently rejected the takeoff stating they were ‘along for the ride.’”

Continental 1404 departed the left side of Runway 34R, reaching a maximum speed of about 120 knots; and crossed a taxiway, a service road, and some uneven terrain before coming to a stop. A postcrash fire ensued. The captain and five of the 110 passengers were seriously injured; the first officer, two cabin crewmembers, and 38 passengers suffered minor injuries. One cabin crewmember and 67 passengers (three of whom were lap-held children) were uninjured.

The crew

The 50-year-old captain was hired by Continental in 1997 and served as first officer on the Douglas DC–9 and Boeing 737/757/767 airplanes before becoming a 737 captain about 14 months before the accident. He completed a line check in April 2008, and recurrent training/proficiency checks in October 2008. Additionally, he completed a Continuing Qualification Syllabus, which included at least one takeoff and one landing in a 35-knot direct crosswind.

The captain came to the airlines with 4,500 hours of Navy flight experience. At the time of the accident, he had 13,100 total hours, including about 6,300 hours in the 737. He had flown about 915, 81, and four hours in the 12 months, one month and 24 hours, respectively, prior to the crash. Flying skills, training records and peer reviews were all rated as excellent.

The 34-year-old first officer, hired by Continental in March 2007, was type rated in the de Havilland DHC–8 and 737. He had flown for Horizon Airlines prior to joining Continental and had about 8,000 hours total time with about 1,500 hours in the Boeing 737. His training records and peer reviews were rated “above average.”

The aircraft

The Boeing 737 had been delivered in 1994 and accumulated just more than 40,000 hours and about 21,000 cycles. No discrepancies on the aircraft or its loading were noted. Boeing guidance listed a maximum crosswind of 40 knots. Winglets installed a month before the accident were tested to a crosswind maximum of 33 knots.

The weather

The National Weather Service (NWS) surface analysis charts showed a low-pressure system near the Colorado/New Mexico border, with a stationary front extending north-south through those states. Westerly winds across the Rocky Mountains suggested downslope winds with moderately strong wind gusts and mountain wave activity.

To assess whether mountain waves could have played a role in the gusty surface wind conditions, the National Center for Atmospheric Research (NCAR) simulated conditions around that time using a high-resolution numerical model. The model showed a well-defined wave that increased significantly shortly before the accident, with winds of 40 knots to 68 knots at the airport between about 6:08 p.m. and 6:18 p.m.

The NTSB estimated the winds during the accident sequence “varied between 30 and 45 knots from the west, resulting in an almost direct crosswind for Runway 34R and a crosswind component that varied from 29 to 45 knots.”This resulted in strong, localized, intermittent gusts at the airport’s surface. NCAR’s model depicted generally stronger westerly flow to the north of the airport, with large regions of relatively lighter winds over the center and southern portions of the airport. Embedded in the overall flow structure were many gusts, which moved from west to east across the airport area. A gust estimated at 45 knots moved across the center of the airport, crossing the accident site, between about 6:14 p.m. and 6:16 p.m. This was apparently captured by one of the airport wind sensors but not reported to the crew.

“At the time of the accident, of the nine possible runway configurations available to ATC, the ‘Landing North/West’ configuration was in use,” according to the NTSB. “In this configuration, traffic was landing on runways 35L, 35R, 34R, and 26, and traffic was departing on runways 34L, 34R, and 25. The...local controller, who was responsible for departing traffic on all three departure runways, cleared seven airplanes for takeoff in about 9 minutes before the accident occurred. (Two more airplanes were holding in position on runways awaiting takeoff clearance when the accident occurred.) The pilots of the other departing airplanes did not report any crosswind-related issues or difficulties.”

Analysis

Using data from the flight data recorder, the NTSB estimated the winds during the accident sequence “varied between 30 and 45 knots from the west, resulting in an almost direct crosswind for Runway 34R and a crosswind component that varied from 29 to 45 knots. A peak gust of 45 knots occurred at 6:18:12 p.m., about the same time the flight data recorder recorded the right rudder pedal beginning to move aft from a position about 72 percent of its available forward travel, reaching a near neutral position at 6:18:13.75 p.m. Although the pilot briefly made a small right rudder pedal input at 6:18:14.25 p.m., the flight data recorder did not record any more substantial right rudder pedal inputs as the airplane continued to veer to the left.”

The NTSB also determined that the rudder was capable of producing enough aerodynamic force to offset the weathervaning tendency created by the existing wind gusts. Ostensibly, that means applying full rudder for as long as it took to counteract the gust.

Denver airport is well-equipped with low-level windshear-monitoring equipment, but the FAA guidance at the time was not to report the most adverse wind but only the wind at the departure end of the runway for takeoffs. One of the NTSB findings: “If the accident pilots had received the most adverse available wind information (which was displayed as airport wind on the Denver International Airport air traffic control tower local controller’s ribbon display terminal and indicated a 35-knot crosswind with 40-knot gusts), the captain would likely have decided to delay the departure or request a different runway because the resultant crosswind component exceeded Continental’s 33-knot crosswind guidelines.”

After the accident, Continental used data from its operational database to measure crosswind components seconds after takeoff for all 940,000 recorded fleet departures in an eight-year period. According to the NTSB, that report showed that only 250 departures encountered a crosswind of 25 knots or greater and only 62 had crosswinds of 30 knots or greater. Crosswinds that strong are truly outlier events.

Additional postaccident testing showed that Continental’s Boeing 737 simulators were not accurate in replicating gusty conditions. They were not programmed to produce gusts, only steady-state winds, below 50 feet agl.

Probable cause

The NTSB determined the probable cause was “the captain’s cessation of right rudder input, which was needed to maintain directional control of the airplane, about four seconds before the excursion, when the airplane encountered a strong and gusty crosswind that exceeded the captain’s training and experience. Contributing to the accident were the following factors: 1) an air traffic control system that did not require or facilitate the dissemination of key, available wind information to the air traffic controllers and pilots; and 2) inadequate crosswind training in the airline industry due to deficient simulator wind gust modeling.”

Commentary

I empathize with the captain. In this unusual and well-documented accident, the winds were close to the aircraft’s limit. Critical wind information was not provided and prior training did not and could not duplicate the conditions encountered.

There are similar situations in GA—the actual winds in the runway zone may be unknown and, frankly, effective training in crosswinds is often lacking. GA loses many aircraft annually to wind in much less extreme conditions, largely because of training limitations. Most crashes occur well within the aircraft’s capability, so there’s a great opportunity to improve. Practice with a good CFI will help. Occasionally, however, weather surprises and then both luck and skill are helpful.

In light aircraft, the pilot’s operating handbook provides demonstrated crosswind capability—the maximum the manufacturer encountered during testing. The actual aerodynamic limits are higher but unknown. Getting to know your aircraft may allow higher limits, but proceed cautiously because the penalty for foulups will always be expensive and sometimes severe.

Gustiness introduces variables, by definition. On some days, it will be unflyable, and on others, it will take significant and sustained control correction. For both takeoff and landing, extra airspeed provides additional control—up to a point. We owe no allegiance to any runway on either arrival or departure. Short runways need not apply because the extra speed needed to cope with strong crosswinds may invalidate the book numbers. If you’re uneasy with wind, advise ATC that a different runway is needed. Sometimes a flight should be delayed or canceled. Exercise that PIC authority, as needed. When inbound, diversion to a different airport may be the smart move. Got gas?

It’s often easy to find fault after an accident, but remember when Lady Luck has smiled on us and be reminded that she’s fickle.