History of the Flight
On 12 December 2000, a Boeing 747-236B (registration G-BDXP) operated a scheduled passenger flight from Boston, USA, to London Heathrow Airport (LHR). En route, the crew obtained the 1800 hrs Terminal Aerodrome Forecast indicating surface winds could exceed the aircraft's crosswind limitation at the expected arrival time. They reviewed alternates and selected London Stansted Airport as a suitable diversion.
At 1917 hrs, the aircraft was instructed to hold in the London area. The crew inquired about the availability of Runway 23 due to wind conditions but were advised of a 20-minute delay. The commander elected to continue for Runway 27R, the scheduled landing runway until 2359 hrs, as the surface wind remained within limits and an ILS approach was preferred.
The First Officer flew the approach with the autopilot engaged. At about 1,000 ft AGL, the Flight Engineer reported an INS-derived wind of 206°/45 kt with 16.1° drift. ATC warned another aircraft of severe turbulence near the hangars on short final. At 700 ft AGL with visual contact established, the commander took control and disengaged the autopilot. Flap 25 with VREF 135 kt was used. Touchdown occurred at 150 kt on a heading 8° left of the runway QDM of 273°M, smoothly and near the centerline. Autospeedbrake deployed normally.
After touchdown, the aircraft initially tracked left despite the commander's full right rudder input. The left wing lifted to a 4° right roll attitude then returned level. The First Officer delayed reverse thrust selection until confident of control. At 110 kt, reverse thrust was deployed. The aircraft tracked back to centerline and then slightly right. A 70% left rudder input caused a rapid left swing. Reverse thrust was cancelled; the commander used the steering tiller but felt nosewheel scrubbing. The left wing briefly lifted again. Full right rudder was reapplied, but the aircraft continued left and departed the paved area onto a grass triangle between two taxiways.
The aircraft came to rest on the grass. The crew assessed for damage, consulted with the attending fire service, and decided against passenger evacuation, completing a normal shutdown. Emergency services confirmed no apparent damage. Disembarkation was delayed due to ground and wind conditions.
Meteorological Conditions
At 1900 hrs, a low pressure area over northwest Ireland with a cold front from South Wales to Western Scotland and a trough 40 nm west of London produced a strong southerly airstream over Heathrow. The 1920 hrs ATIS reported: landing Runway 27R, surface wind 190°/21 kt gusting to 32 kt, varying between 160° and 220°, visibility 10 km, moderate rain, cloud broken at 1,300 ft and overcast at 1,800 ft, temperature 13°C, dewpoint 11°C, QNH 998 mb, with turbulence and windshear warnings. This ATIS message was required by the Manual of Air Traffic Services when the surface wind is between 170° and 230° exceeding 15 kt. Anemometer traces near Runway 09R threshold showed wind speeds between 12 kt and 36 kt in a two-minute period around the incident. The Runway 27R touchdown sensor recorded a 2-minute mean wind speed of 25 kt, maximum 33 kt, minimum 15 kt, mean direction 198°M, maximum 220°M, minimum 180°M at 1938 hrs. The tower reported wind to the aircraft as 200°M/31 kt with landing clearance. The runway was wet. A post-incident grip test gave a braking condition 'good' reading.
Aircraft Information
The aircraft's centre of gravity was in the middle of the allowable range. Landing weight was 216 tonnes (max 285 tonnes). An outstanding technical defect was an unserviceable autothrottle system, causing the reference speed 'bug' to remain fixed at 290 kt; a manually positioned 'bug' was used instead.
The maximum demonstrated crosswind for initial certification on a dry runway was 28 kt; this was not considered the limiting crosswind condition. In 1995, following crosswind-related diversions, the manufacturer conducted simulator trials and analysis, resulting in revised crosswind guidelines for the 747-200: 32 kt on a wet runway, adopted by the operator in September 1996. The recommended landing technique for crosswind on a wet surface was to fly the aircraft onto the runway while maintaining centerline track, without necessarily eliminating the crab angle before touchdown. The operator's flight crew manual recommended delaying reverse thrust until directional control was assured.
The aircraft's rudder fine steering system had been de-activated for fleet commonality. If full rudder was insufficient, the manufacturer and operator recommended differential braking, though this technique was rarely required and not routinely practiced in the simulator. The operator's manual stated tiller steering could be used up to 100 kt but normally not required until 40 kt. The aircraft also had a body gear steering system used solely for taxiing; it was not selected during the incident.
Flight Recorder Analysis
Data from the Cockpit Voice Recorder (CVR), Flight Data Recorder (FDR), and Quick Access Recorder (QAR) were used to reconstruct the approach and landing. Rudder pedal position was recorded but not rudder deflection; tiller input data was unavailable. The aircraft remained accurately aligned on the localiser, slightly below glideslope on final approach. ATC transmitted wind reports to the aircraft, including 200°/25 kt gusting 30 kt at 8.5 nm from touchdown, and 200°/31 kt about 36 seconds before touchdown. The commander commented "that's on our limits." At 10 seconds before touchdown, wind was reported as 200°/28 kt occasionally gusting 30 kt. Throughout the approach, the Flight Engineer called INS-derived headwind and drift, consistent with QAR data: 216°/27 kt four seconds before touchdown and 191°/27 kt at touchdown.
The FDR and CVR confirm the commander had difficulty controlling after landing. In the first 10 seconds of the landing roll, first the left then right wing lifted. Full right rudder pedal was used to regain the centerline. The crew considered control regained as the aircraft decelerated through 110 kt, based on CVR comments and reduced control inputs, and deployed reverse thrust.
As airspeed reduced through 100 kt, the airspeed-groundspeed relationship changed consistent with an 8 kt relative increase in headwind component. The aircraft turned right; full left rudder and 80% left control wheel were applied. Two seconds later, the increased headwind component disappeared. After three more seconds, a 16 kt relative headwind increase occurred. The left wing began to lift, and the aircraft turned left. Full right rudder and full left control wheel were applied but the aircraft continued turning left, reaching a maximum roll angle of about 4° right. It left the runway on a heading of 242° at 50 kt. After leaving the paved surface, the heading increased and came to rest at about 258°.
Probable Cause
null