Summary
On 29 October 2002, at 1715 UTC, a Boeing 767-304 (registration G-OBYB) operating a public transport positioning flight from Manchester to Birmingham International Airport experienced a runway excursion during the landing roll on Runway 15. The aircraft touched down normally and initially decelerated, reaching 65 kt with over half the runway remaining. The crew then released the brakes and allowed the aircraft to coast at approximately 60 kt toward the end of the runway. Braking was reapplied gently near the threshold of Runway 33, and a left turn was attempted at about 40 kt using rudder and differential braking. The aircraft failed to complete the turn, and the nosewheels departed the paved surface by about 15 metres; the main landing gear remained on the runway. No injuries were reported, and the aircraft sustained no damage.
History of Flight
The flight departed Manchester at 1657 UTC and climbed to 11,000 ft. The crew received a direct approach clearance to Runway 15 at Birmingham. Despite a high workload, the approach was stabilised, with the aircraft on localiser and glideslope from 2,000 ft above runway level. The final approach from about 900 ft was flown at 126–129 kt, with autobrake 3 armed. Touchdown occurred at 128 kt within the normal touchdown zone. Spoilers deployed, idle reverse was selected, and autobrake 3 activated. Deceleration appeared normal, and groundspeed reduced to 65 kt approximately 1,000 metres after touchdown, leaving about 1,200 metres of paved surface. At that point, reverse was cancelled, spoilers retracted, and brakes released. The aircraft continued at a constant 60–65 kt toward a 90° left turnoff at the extreme end of the runway. Just before the Runway 33 threshold markings, symmetric brake pressure was applied. When speed reached 40 kt, a left turn was initiated with rudder and differential braking. The aircraft did not complete the turn, and the nosewheels ran off the paved surface at about 12 kt, travelling a further 15 metres before stopping. The crew shut down engines and disembarked via stairs; the aircraft was later towed and returned to service after inspections.
Weather
The ATIS information received before approach reported Runway 15 in use, surface wind calm, visibility 4,000 metres with mist and slight drizzle, overcast at 900 ft, temperature +9°C, dew point +9°C, QNH 1009, and the runway advisory "WET, WET, WET" with a warning of bird activity.
Flight Recorders
The aircraft was equipped with a solid-state Digital Flight Data Recorder (DFDR) and a 2-hour Cockpit Voice Recorder (CVR). The CVR retained the record of the entire Birmingham flight, but a balance issue between ATC reception and crew microphones made some crew speech unintelligible during the landing roll; however, the area microphone provided usable audio. The DFDR provided good quality data. Analysis of recorded parameters showed that after touchdown and deceleration to 65 kt, the aircraft coasted with speed slowly reducing to 61 kt until about 30 metres before the Runway 33 threshold markers, where brakes were gradually applied. Brake pressure increased symmetrically, but retardation dropped from -0.25 g to -0.1 g despite increasing pressure. Further brake application raised pressure to about 1600 psi per side but only increased retardation to -0.16 g. At about 40 kt, a left turn was initiated using rudder and differential braking (nosewheel steering use unknown). Left brake pressure then rose to maximum (3000 psi) while right decreased to 400 psi. When 50 metres from the runway end, right brake pressure also increased to maximum, producing a peak retardation of -0.3 g. Symmetric full braking was held for 2–3 seconds before the nosewheel left the paved surface at approximately 12 kt. Brake pressure was then relaxed, and the aircraft stopped after 10–15 metres. CVR sounds possibly consistent with nosewheel slipping after it left the paved surface.
Runway Surface
Shortly after the incident, Birmingham Airport Authority measured runway friction using a Grip Tester, finding average friction coefficient greater than 0.8 in the area between the Runway 33 threshold and the runway end. A later classification by Cranfield Aerospace Limited (16 January 2003) per CAA CAP 683 showed an average friction coefficient of 0.85, above the design objective of 0.8. However, several areas over painted markings had friction below the minimum level (0.55), and one very small area beyond the threshold markings fell within the maintenance planning band (0.55–0.63) or below. The Cranfield report recommended recoating with friction paint, and Birmingham Airport subsequently renovated the threshold painted surfaces with a new formulation providing friction above the maintenance planning level.
Analysis of the Incident
The approach and landing were normal. After reaching 65 kt with 1,200 metres remaining, the aircraft coasted at constant speed toward the end of the runway. The crew considered vacating at the Runway 06/24 intersection but continued to the end turnoff. There was no pressure to vacate quickly; ATC informed a following aircraft that speed was unrestricted. Braking was reapplied gently near the Runway 33 threshold, but retardation decreased briefly despite increasing brake pressure, possibly due to localised low friction. The turn was commenced at about 40 kt, which was high for a 90° turnoff. The probable cause of this incident was the high speed at the start of the turn.