Description

On 8 June 2022, a Boeing 757-200 (N949FD) operated by Federal Express on a scheduled domestic cargo flight from Fort Worth Alliance to Tulsa as FX1170 was cleared to land on runway 18L at destination in night VMC. However, it landed instead on the significantly shorter runway 18R without ATC intervention. Since this was a much shorter runway than 18L, harder than expected braking was required to safely exit at the end. The two runways are just over 1500 metres apart. No traffic conflict in the air or on the ground was created as a result of the error.  

Investigation

An Investigation was carried out by the US National Transportation Safety Board (NTSB). The combined FDR/CVR recorded was removed from the aircraft and relevant data were recovered. The 57-year-old captain had a total of “over 10,000 hours” flying experience, which included 790 hours in command on type and stated that he had previously operated into Tulsa “at least a hundred times”, most recently about two weeks prior to the landing under investigation. The 50-year-old first officer had a total of “about 4,500” hours flying experience, which included 739 hours on type. It was his second time operating into Tulsa since being employed by FedEx two years earlier. Both pilots were Memphis-based but had not previously operated together.

What Happened

The crew reported for their night duty at Ontario, California, at the equivalent of 2050 Central Daylight Time (CDT) on 7 June, following which they operated to Fort Worth Alliance, arriving there at about 0030 CDT the following day. They then each went to separate crew rest rooms until due to prepare for their next flight three hours later. The first officer stated that he had been able to get “about 30 minutes rest,” and the captain stated that although he had attempted to nap, he “couldn’t get to sleep." However, the captain added that despite this, he had not felt tired during the subsequent flight to Tulsa. At approximately 0330, the aircraft pushed back for the less than one-hour flight, with the captain acting as PF.

Once airborne, the flight climbed to FL310 in nineteen minutes and then began the descent into Tulsa two minutes later. The first officer obtained the destination weather and landing data, preset the ILS frequencies, and inserted a waypoint about 30 miles from the airport because they wanted to be at 11,000 feet at that position. The captain then briefed the expected visual approach to the 3,048 metre-long runway 18L, backed up by the ILS. He noted a VREF of 123 KIAS and the approach lighting for this runway. He also briefed that he would not use the autobrake, as he wanted to “roll long” because the required parking position was at the far end of the runway. 

At about 10,000 feet, still in IMC, the flight was transferred to Tulsa APP and given a radar heading of 360° to take the flight to the west of the airport. Further descent followed and took the flight below the cloud base, with the first officer stating that although he could not see the runway, he had seen the airport beacon and said their position at that time “looked like a normal downwind”. The Captain stated that at this time, he had seen the lights for runway 08/26 and that they were “normal”. The crew extended the runway 18L centreline in their FMS, and this was shown on their respective NDs and confirmed that ILS 18L frequency had been correctly set. 

On request from the APP controller, the first officer confirmed that they had the airport in sight, and the flight was then cleared for a visual approach to runway 18L, and then cleared to land. The AP was still engaged, and the right turn from downwind to final approach was commenced. The first officer set the PSA to 2,400 feet final approach fix (FAF) for the 18L ILS. During the turn onto final approach, the AP was disengaged, flaps 5 selected, and the aircraft was subsequently configured for landing by selecting flaps 30 and gear down. The first officer stated that at this time, he had noted that “the aircraft appeared low visually” and that he had brought this to the captain’s attention. He added that whilst the ILS GS indication appeared to be “normal," the PAPI lights indicated they were below the correct glideslope. He also stated to the Investigation that the deviation bar on his horizontal situation display (HSI) had been deflected to the left, but that he had not brought this observation to the captain’s attention. 

The captain adjusted the rate of descent to return to the correct PAPI, and at “about a 2.8 nm final and when about 800 feet AGL, the RAAS annunciation “Approaching 18R” occurred simultaneously with crew communications and “neither pilot acknowledged or recalled this call." The captain stated that initially the HUD was showing the runway “slightly off to the left” but as he transitioned visually he had been more focused on the PAPI display. The flight touched down on the 1,860 metre-long runway 18R, and the RAAS subsequently announced that there was 3,000 feet (914 metres) of runway remaining. The captain stated that when he applied the brakes, he “came on them harder initially because he was confused.” After decelerating, had asked the first officer if they were on the correct runway, then said “we landed on the right [hand] runway." The controller, who had not noticed the error, was advised of it and responded with taxi instructions to the ramp. 

The Tulsa parallel north/south runways with the aircraft ground track shown in green. [Reproduced from the Official Report] 

Why It Happened

How both pilots and the controller involved failed to recognise that the aircraft was not complying with their approach and landing clearance until after touchdown was examined. The presence of multiple salient visual cues for the pilots, both external and internal, was noted. It was clear that having set up the correct ILS, both pilots had appeared to “discount the information their instruments were providing” in favour of the view they had of the runway they had already visually acquired with the “multiple visual cues that they were misaligned” unrecognised. Once the captain had disengaged the AP, the crew had continued to the landing without engaging in any further confirmatory actions.

Such behaviour was noted to be consistent with the psychological phenomenon known as Plan Continuation Bias - an unwillingness to deviate from a previously determined course of action, despite indications that change is required. It was noted that “plan continuation bias is exacerbated by fatigue,” and that the error had occurred within the window of circadian low, with the corresponding potential for increased vulnerability to error. Although the captain stated that he had not been fatigued, it was noted that he had been continuously awake for more than 15 hours and was therefore “likely (to be) experiencing fatigue due to both chronic and acute sleep debt,” resulting from limited sleep over several days preceding the flight. It was therefore concluded that both pilots’ lack of recognition of their error was probably affected by fatigue, plan continuation bias, and their inability to perceive and efficiently integrate available information.

The aircraft operator’s approach to Fatigue Risk Management was examined, and it was found to rely on the Karolinska Sleepiness Scale (KSS) to determine potential risk for fatigue. This was used for each intended crew pairing during the preparation of pilot rosters. The assessment of the flight pairing using this scale assumed that the crew would nap for 30 minutes during their hub layover, but it was noted that neither the assessments nor the assumptions on which they were based were disclosed to crews. The Investigation found that by failing to obtain a nap during the hub layover, the captain’s KSS score had increased from "within limits" to "high risk." It was noted that the operator also “collaborates with ALPA” on rostering assessments of fatigue risk and ALPA's method also incorporated the same nap assumption.

The approach as flown was programmed into the operator's FFS, and the HUD and PFD views of the misaligned approach were recorded and are reproduced below. Both clearly indicate the lateral and vertical offset from the intended ILS 18L approach, which was also affected by the threshold of runway 18L being approximately 800 metres closer than the threshold of runway 18R.

The failure of the controller to make any attempt to monitor the flight’s compliance with its clearance was also examined given that “it provides an additional barrier” in the avoidance of aircraft misalignment on approach. This inaction on the part of the controller was attributed to expectation bias, and it was noted that “past experience or repetition can exacerbate this issue." During her interview, the controller stated that after conducting “a visual sweep of the runway when she initially cleared the flight for the approach and to land [...] she did not look back at the aircraft again." She added that “the last time she observed the aircraft on the radar was when it was on a “dog leg” turn to base”.

The view of the approach on the Captain’s HUD. Reproduced from the Official Report]

The view of the approach as flown on the PFD. [Reproduced from the Official Report]

The Probable Cause of the event was determined as "The flight crew’s misidentification of the intended landing runway”. 

Two Contributory Factors were also identified:

1. The flight crew’s failure to perceive and correctly interpret visual and auditory indicators - including electronic guidance - that they were approaching the incorrect runway, which was likely the result of a degradation in cognitive function brought on by working within their window of circadian low, increased workload, and fatigue.
2. The air traffic controller’s failure to monitor the arriving flight after issuing a landing clearance. 

The Final Report was published on 3 August 2023. No Safety Recommendations were made. 

Related Articles

• Wrong Runway Use
• Runway Incursion
• Plan Continuation Bias
• Fatigue
• Fatigue Risk Management System
• Circadian Rhythm Disruption
• ATC Expectation Bias
• Vigilance in ATM
• Flight Path Monitoring
• Pilot Flying (PF) and Pilot Monitoring (PM)
• Head Up Display (HUD)
• Runway Awareness and Advisory System (RAAS)
• Visual Approach
• Flying a Visual Approach