B738, East Midlands UK, 2020
B738, East Midlands UK, 2020
On 9 February 2020, a Boeing 737-800 rejected its takeoff from East Midlands from a speed above V1 after encountering windshear in limiting weather conditions and was brought to a stop with 600 metres of runway remaining. The Investigation found that the Captain had assigned the takeoff to his First Officer but had taken control after deciding that a rejected takeoff was appropriate even though unequivocal QRH guidance that high speed rejected takeoffs should not be made due to windshear existed. Boeing analysis found that successful outcomes during takeoff windshear events have historically been more likely when takeoff is continued.
On 9 February 2020, a Boeing 737-800, (G-DRTN) being operated by Jet2 on a scheduled international passenger flight from East Midlands to Tenerife, made a high speed rejected takeoff on runway 27 after encountering windshear during an adverse weather departure and was brought to a stop with approximately 600 metres of runway remaining. Damage attributable to the rejected takeoff was found to one main landing gear wheel and its brake unit which were replaced. No damage to any runway edge lighting was caused.
An Investigation was carried by the UK AAIB. The CVR and FDR were not available but the aircraft operator made relevant QAR data available.
It was noted that the 45 year-old Captain had joined the Operator 11 months previously, had a total of 12,300 flying hours which included 893 hours on type and had flown six departures from East Midlands. The Co-pilot was a Senior First Officer who had a total of 3,200 flying hours, all but 700 hours of which were on type, and had been promoted to Senior First Officer ten months previously. This promotion required in excess of 2000 hours and six months service with the operator as well as passing a suitability assessment by the flight training department. Unlike First or Second Officers, Senior First Officers were permitted to act as PF up to the same OM takeoff and landing limits as Captains which the operator stated was to allow them to gain relevant operational experience in challenging conditions ahead of promotion to command.
On the day of the event the prevailing forecast weather conditions at the departure airport included a strong surface wind (200° at 32 knots with gusts to 45 knots) which was close to the applicable crosswind limit (25 knots) for takeoff at the expected mean speed and well outside it in gusts. The crew stated that they had arrived early for their duty and had “discussed the weather situation at some length”. They were aware that the runway surface condition was being reported as wet and that light rain was continuing to fall. The Captain initially decided that he would do the takeoff because of the adverse weather conditions but, after being reminded by his Senior First Officer (SFO) that he was permitted by the operator to use the same weather limits and after discussing various anticipated weather-related customer service matters which he anticipated would require his attention in the lead up to departure, he changed his mind and decided he would act as PM for the takeoff and the briefing was completed accordingly.
Once the crew reached the aircraft, the Captain carried out the pre flight external check whilst the SFO carried out the necessary pre flight internal checks and set up for the departure. During an onboard departure briefing, it was reported that it was agreed that rated thrust would be used on the full runway length and that actions in the event of a rejected takeoff or windshear at various stages of the takeoff and the implications of windshear for performance calculations were all reviewed. The calculated speeds for the departure included a V1 of 134 KIAS and a VR of 149 KIAS, a VRMAX of 158 KIAS and a V2 of 156 KIAS. It was noted by the Investigation that VRMAX was an additional calculation required for windshear takeoff performance being the lowest of the maximum VR for the maximum allowable weight in the prevailing runway and weather conditions and the VR for the actual weight plus 20 knots. The “rotate” call is then made at VRMAX unless windshear is encountered at or beyond the normal VR and before reaching it in which case, rotation should commence immediately and “without hesitation”.
Since the fuel loaded was in excess of the planned requirement, it was decided to start up and taxi to the departure end of the runway to await an opportunity for takeoff within crosswind limits. Having converted their 25 knot crosswind limit to a wind speed which would apply if the relatively constant wind direction of 210° was maintained as 29 knots, this wind velocity was passed to the TWR Controller.
As the aircraft approached the A1 holding point for Runway 27, the flight were passed a wind of 220° at 32 knots (a crosswind component of 23 knots) and cleared for takeoff so the line up was continued into a rolling takeoff. The Captain reported noticing and calling out a transitory loss of airspeed of 10/15 knots as the aircraft passed 120 knots but observed that acceleration quickly resumed. He stated that as the speed approached V1 (134 KIAS), “the aircraft deviated dramatically from the centreline to the right” - he estimated the deviation was 20/30° - and saw a downward trend on his airspeed indication. He stated that he had “felt that the SFOs attempts to control the heading were ineffective” and perceiving that the indicated airspeed was reducing and the aircraft was still below V1, he had taken control, called “Stop” and commenced a Rejected Take Off (RTO). The aircraft came to a stop on the runway centreline between taxiways ‘M’ and ‘H’ (see the illustration below) with approximately 600 metres of the 2,893 metre-long 45 metre-wide runway remaining.
The takeoff runway with the two intersections between which the aircraft stopped highlighted. [Reproduced from the Official Report]
Once the aircraft had stopped, the Captain made a PA announcement to acknowledge what had happened and then spoke with TWR to advise that they intended to vacate the runway and suggesting a runway inspection in case any runway edge lights had been struck during the deviation. Once clear, the crew requested clearance to return to their stand and were so cleared. The Captain reported that he had been “concerned about the SFO who appeared to be preoccupied about what had happened during the takeoff” and also that he had suspected that “at least one brake was ‘binding’” since maintaining a normal taxi speed required more thrust that usual.
With aircraft back on stand, the passengers were disembarked and once a towbar had been attached, the parking brake was released. The QRH recommendation for remote parking to be considered after an RTO was not considered. Company maintenance personnel were then called to carry out an inspection of the landing gear. The damage attributable to the rejected takeoff was found to be limited to one main landing gear wheel and its brake unit which required replacement (as well as all the other wheels and tyres in accordance with company maintenance requirements).
Recorded flight data confirmed a transient speed loss of 23 knots at 131 KCAS and rapid recovery. It also showed that almost immediately after the automatically generated V1 callout, the second reported transient speed loss of 13 knots had occurred over 1½ seconds and had been accompanied by the reported deviation to the right of approximately 8°. There was a significant corrective left rudder input and between 5 and 6 seconds after V1, an autobrake-generated maximum brake pressure application was recorded, indicating that a rejected takeoff had been initiated. The approximate speed when the reject was initiated was 149 KCAS, some 15 knots above V1.
It was noted that there were four situations listed in the QRH in which a high speed rejected takeoff should be considered - fire/fire warning, engine failure, if the aircraft is unsafe or unable to fly or in the event that a predictive windshear warning is annunciated. The latter are only enabled during takeoff until passing 100 knots and none were received. It was found that reactive windshear warnings are not active until the aircraft is airborne.
The QRH also provides guidance if actual windshear is encountered at high speed, noting that even if it occurs prior to V1, “there may not be sufficient runway remaining to stop if an RTO is initiated at V1” and notes “if windshear is encountered near the normal rotation speed and airspeed suddenly decreases, there may not be sufficient runway left to accelerate back to normal takeoff speed”. It goes on to say that if there is insufficient runway left to stop, and that assuming maximum thrust is set, a normal rotation should be initiated at least 600 metres before the end of the runway “even if airspeed is low” with the possibility that higher than normal attitudes may be needed to lift off in the distance remaining.
On being asked to comment on the above guidance, Boeing observed that it was based on industry consensus and event data and that “successful outcomes of windshear and other emergency events during takeoff have been more likely if takeoff is continued than if takeoff is rejected late or without sufficient stopping distance” whilst also noting that “guidance cannot be given for every possible set of takeoff and windshear conditions (and that) the Captain must assess each situation to determine the safest course of action”.
The aircraft operator noted from their own investigation that the V1 call which is required to be made by the PM was not made. They observed that since this call is the cue for Captains to remove their hand from the thrust levers, it was possible that without it, their hand may have remained on the thrust levers, an action “which lends itself to a rejection”. The Investigation concurred and noted that this SOP to remove the physical connection to the thrust levers at V1 “is intended to remove the possibility that pilots will instinctively or impulsively close the levers in response to an adverse event”. The Investigation also recognised the possibility that the Captain’s response in rejecting the takeoff from above V1 may have been affected by the “startling effect” of the aircraft swinging off the centreline immediately after passing V1.
In respect of the failure to consider the QRH pointer to the possible advisability of remote parking when taxiing in after the RTO, the Investigation took the view that “preoccupation with what had just happened and concerns for the SFOs wellbeing” may have represented a concurrent distraction.
- The takeoff was commenced following the receipt of a ‘spot wind’, which was just within the applicable crosswind limit.
- The mean wind velocity in the current TAF was equivalent to a 29 knot crosswind component and the maximum gust given there was equivalent to a 41 knot crosswind component. The METAR observation timed two minutes prior to takeoff (and which the crew would not have been aware of) replicated exactly the same crosswind components for mean speed and gust as those in the TAF. These components were materially in excess of those implied by the spot wind on which the decision to commence takeoff relied.
The formally documented Conclusion of the Investigation was as follows:
The Captain had a high workload managing the departure and, to give himself time, decided that the co-pilot should fly the take off. In very gusty wind conditions, the aircraft encountered a windshear event near V1 which caused a 13 knot reduction in airspeed. Additionally, the strength of the crosswind during takeoff caused the aircraft to veer right. Concerned that the aircraft might leave the runway, and considering the situation to be unsafe, the commander initiated an RTO five seconds after V1. The crew had not called V1 although the automatic callout had sounded. The SOP is to continue a takeoff when V1 has been reached because, as the manufacturer commented, successful outcomes are more likely when a takeoff is continued rather than rejected.
The Final Report was published on 10 December 2020. No Safety Recommendations were made.