BOEING 737-8

Full Narrative

HISTORY OF FLIGHTOn March 8, 2024, about 0758 central standard time (CST), United Airlines flight 2477, a Boeing 737-8, N27290, departed taxiway SC after turning off of runway 27 at high speed after landing at IAH (see figure 1). The left main landing gear departed the paved surface and impacted a concrete structure, that was recessed in the ground, resulting in its separation (see figures 2 and 3). None of the 6 crew and 160 passengers were injured and deplaned via airstairs. Flight 2477 operated as a Title 14 Code of Federal Regulations Part 121 scheduled domestic passenger flight from Memphis International Airport (MEM), Memphis, Tennessee to IAH.


Figure 1. The accident airplane’s position data from ADS-B overlayed on a Google Earth image of runway 27 and taxiway SC. The green arrows show the airplane’s direction of travel.


Figure 2. The accident airplane after the taxiway excursion.


Figure 3. The concrete electrical manhole hit by the left main landing gear.
According to the flight crew, the captain was the pilot flying and the FO was the pilot monitoring. The captain said he observed the reported runway surface condition codes when checking the automatic terminal information system (ATIS) via his electronic flight bag (EFB). Specifically, the captain recalled seeing a runway condition assessment matrix (RCAM) condition code of 3/3/3 for runways 26L and 26R and a condition code of 5/5/5 for runway 27. At the time of the accident, the RCAM code broadcast by the ATIS for runway 27 was 3/3/3, indicating the entire runway was “slippery when wet” and that braking deceleration would be noticeably reduced for the wheel braking effort applied or that directional control would be noticeably reduced.
When the flight crew checked in with Houston Approach Control, they were told to expect a landing on runway 26L. The captain asked the FO to request a landing on runway 27 instead. The FO made, and the controller approved, this request, and issued instructions for the instrument landing system (ILS) approach to runway 27. The captain asked the FO to request approval to roll to the end of runway 27, which the FO stated he would do and the captain replied that, if approved, they would change the autobrakes setting from 2 to 1, which would reduce the deceleration rate for the autobrake system.
About 7 miles from the runway 27 arrival threshold, the FO contacted Houston Tower and asked “how’s our spacing looking? Can we roll it all the way to the end?” to which Houston Tower responded “uh…keep your speed up. That’s approved.” The captain subsequently changed the autobrake setting from 2 to 1. The approach was conducted in instrument meteorological conditions (IMC) and, according to the flight crew, the airplane broke out of the clouds between 800 and 1,000 feet msl. They reported that visibility under the clouds was good, and the captain recalled that the runway appeared dry. The FO recalled that the runway appeared wet.
The crew stated that the touchdown was uneventful, at an appropriate speed, and within the touchdown zone. The speedbrakes extended normally, and the thrust reversers were deployed to idle reverse thrust. The captain said that, shortly after touchdown, he disable the autobrakes by moving the speedbrake lever to its down and locked position, which also retracted the speedbrakes. He did not “slow too much initially” because the runway appeared dry, he wanted to expedite their time on the runway, and because he preferred decelerating gradually for passenger comfort.
The captain recalled applying wheel brakes manually, about 6,000 feet from the end of the runway, but felt as if the deceleration was less than normal. He recalled hearing the runway awareness and advisory system (RAAS) alert indicating 1,000 ft of runway distance remaining. He became concerned and began applying more pressure to the brakes. As he approached the end of the runway, he elected to attempt to turn onto taxiway SC by utilizing the steering tiller and rudder pedals, while pushing aggressively on the brake pedals.
The captain felt the fuselage and rudder/brake pedals begin to shake violently, as the aircraft turned onto the taxiway. He briefly released the brake pressure and the shaking ceased. He then reapplied aggressive brake pressure and the shaking resumed. The airplane departed taxiway SC and the left main landing gear tires and nose wheels tires entered the grass before the airplane came to a rest with its left wing low. PERSONNEL INFORMATIONThe captain was hired by United Airlines in April 1987 and was upgraded to captain in March 2007. His most recent training on landing performance and crew resource management was on February 26, 2024. He last completed recurrent ground school on January 2, 2024. His most recent Line Oriented Flight Training and Line Operating Evaluation were completed on August 6, 2023 and August 7, 2023, respectively.
The FO was hired by United Airlines in October 2019. His most recent training on landing performance and crew resource management were on January 31, 2024. Additionally, his last Line Oriented Flight Training and Line Operating Evaluation were completed on January 31, 2024. He last completed recurrent ground school on November 19, 2023. ADDITIONAL INFORMATIONUnited Airlines Boeing 737-8 FM stated that autobrakes are required to be set at 3 or maximum for a RCAM condition code of 3/3/3 or less. The FM section on landing contained guidance to use maximum reverse thrust after main gear touchdown, with a warning to “maintain maximum reverse thrust until stopping is assured.” A caution stated “when turning off from a wet or contaminated runway, slow to an appropriate speed (no greater than 30 knots) to prevent skidding and excursions. A note stated that “unless the speedbrake is raised after touchdown, braking effectiveness may be reduced initially by as much as 60% since very little weight will be on the wheels and brake application may cause rapid anti-skid modulation.”
United Airlines FOM stated that arrival briefings should be conducted by the crew during the last phase of cruise flight and prior to the top of descent point. The FOM stated that the captain starts the arrival briefing with a threat discussion, but to promote mentoring and crew engagement, encourages the captain to have the FO begin the discussion. Threats identified for the arrival, approach, landing, and taxi should be thoroughly discussed to include considerations and mitigation strategies. The list of topics for the threat discussion included landing performance and runway condition. DAMAGE TO AIRCRAFTPortions of the left main landing gear and its doors, as well as portions of the left wing flaps, partially separated from their normally installed locations on the airplane during the accident. The left main landing gear had rotated aft from its normal position and its tires were in contact with the inboard flap. The fuse pins for the forward trunnion bearing housing assembly of the left main landing gear had sheared. This bearing housing assembly remained attached to the outer cylinder of the left main landing gear. The aft trunnion pin of the left main landing gear had dislodged from the main landing gear beam and remained on the outer cylinder, with no damage found on the trunnion pin. The main landing gear beam aft trunnion spherical bearing remained installed in the main landing gear beam but was damaged and out of its normal position. The main landing gear beam exhibited gouging damage.

Figure 4. The left main landing gear was displaced aft of its normal position.
There was no visible damage to the forward and aft spars of the left wing structure. The integrity of the left wing fuel tank was not compromised. The lower-aft fuselage skin showed abrasion and deformation consistent with ground contact. The left engine nacelle exhibited abrasion damage to the fan cowl and thrust reverser consistent with ground contact. FLIGHT RECORDERSData from the accident airplane’s FDR and automatic dependent surveillance-broadcast (ADS-B) showed that the airplane touched down about 1,400 feet from the runway arrival threshold with a groundspeed of 158 kts. FDR data showed the speedbrake lever position was full up and the autobrake system activated after landing. About 3 seconds after the activation of the autobrakes, the thrust reversers showed fully deployed, but engine fan speeds trended toward idle through the remainder of the landing.
About 6 seconds after activation of autobrakes, the speedbrake lever position changed to down and locked, the autobrake system ceased applying wheel brake pressure, the autobrake disarm light activated, and the speedbrakes were stowed. The airplane was about 6,600 ft from the end of the runway and the airplane’s groundspeed was about 136 kts when the autobrakes were disarmed and the speedbrakes were stowed. Manual wheel braking did not begin until the airplane was about 5,000 feet from the end of the runway, at a groundspeed of 117 kts, though left and right brake pressure began low with a slowly increasing trend. With about 1,000 ft and 500 ft of runway remaining, the airplane groundspeeds were 65 kts and 45 kts, respectively. The airplane’s groundspeed was about 40 kts when the airplane turned onto taxiway SC from runway 27. The airplane departed the paved surface with a groundspeed of 22 kts. TESTS AND RESEARCHInvestigators evaluated airplane stopping performance for different combinations of runway surface condition and the usage of speedbrakes, maximum engine reverse thrust, and manual wheel brake application. The performance study determined that the Boeing 737-8 had adequate performance capability to stop on wet pavement within the IAH runway 27 landing distance available. Furthermore, for all wet runway simulation cases that included both maximum manual wheel braking and speedbrakes redeployed at least 2,000 ft prior to the departure threshold, the stopping distance remained less than 10,000 ft, the length of runway 27.
The airplane performance simulation used the accident airplane’s configuration and determined that it would require in excess of 10,000 ft of landing distance to stop. The accident airplane configuration included its estimated loading, energy state, main landing gear touchdown location, environmental conditions, deceleration device configuration schedules (timing, setting, and duration), estimated aerodynamic forces and moments, estimated engine forces and moments, and wheel braking using flight data. However, 24 of the 29 simulations using alternate deceleration device configuration schedules on wet pavement conditions enabled the airplane to slow to a safe taxi speed or to a full stop within the landing distance available for runway 27.
Differences between the accident flight configuration and the aforementioned 24 simulations that were major factors enabling successful deceleration were 1) speedbrake redeployment or not stowing speedbrakes prior to runway exit, resulting in reduced wing aerodynamic lift, increased aerodynamic drag, and increased main landing gear wheel loading to improve wheel braking performance; 2) timely application and duration of autobrakes and/or manual wheel braking; and 3) use of engine reverse thrust, including reverse maximum thrust, until a safe taxi speed or full stop was assured. ORGANIZATIONAL AND MANAGEMENT INFORMATIONAccording to the Line Operations Safety Audit (LOSA) Collaborative, which developed the method in the early 2000s, a LOSA was a “peer-to-peer observational methodology based on the capture of qualitative and quantitative threat and error management (TEM) performance data of frontline personnel in their natural work conditions.” The LOSA Collaborative assisted major airlines with LOSA implementation that could take the form of either a “snapshot LOSA” or a “continuous LOSA.” A snapshot LOSA collected, validated, and analyzed observational data within a defined timeframe, typically six months, concluding with a report on the results. A continuous LOSA provided a constant stream of observational data over time that was managed by the airline. The operator stated that LOSAs were not evaluations and that the program collected de-identified data on crew compliance with the Flight Operations Manual (FOM) and the fleet-specific Flight Manual (FM) SOPs.
Prior to 2023, United Airlines conducted snapshot LOSAs about every four years, with the most recent snapshot LOSAs conducted in 2016 and 2021. Each involved about 60 pilot-observers who were removed from the flight schedule for about 2.5 months to conduct observations full time. In mid-2023, United Airlines launched a continuous LOSA program in which a pool of about 50 trained pilot-observers performed about 25 flight observations per month. By early 2024, the continuous LOSA was actively collecting data across the United Airlines fleet. Analysis of the resulting data was to be performed in partnership with the LOSA Collaborative. At the time of the accident, the dataset was not large enough to provide a statistically relevant sample.