AAIB Bulletin: 2/2020

This intersection was referred to as ‘PSNUTMP’ (temporary position U)1 in the Electronic Flight Bag (EFB). The crew subsequently re-planned for a departure from Taxiway S1, but in recalculating the performance they both selected ‘PSNSTMP’ (temporary position S). This position was the intersection of Taxiway S4 with the runway ie the full length of the runway. The crew did not cross check the TORA from PSNSTMP against the TORA from Taxiway S1, so the error was not identified by the crew before takeoff. The aircraft departed from the Taxiway U5 intersection (TORA 2,410 m) at 2034 hrs using an engine thrust setting based on performance figures calculated for the full length of the runway (TORA 3,805 m) (Figure 2). With the reduced power setting, the commander commented subsequently that the takeoff “felt wrong”, but Takeoff/Go-around (TOGA) thrust was not selected. The aircraft passed the upwind threshold of the runway at a height of approximately 100 ft. During the flight, the crew realised what had happened and reported it to the operator after landing. Figure 1 Plan of Lisbon airport showing Taxiways S1, U5 and S4 (chart not orientated north-up) Footnote 1 See later sections, Airfield information and Electronic flight bag nomenclature. 5 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-EZTD EW/C2019/04/03 EFB calculated takeoff start point Start of takeoff roll Aircraft airborne Figure 2 Image of Lisbon Airport showing the calculated and actual takeoff points Recorded information Data from the FDR and digital access recorder (DAR) were downloaded from the aircraft by the operator on arrival at Luton and copies were subsequently provided to the AAIB. The DAR is used to provide data for the operator’s Flight Data Monitoring (FDM) programme. The 2-hour duration CVR recording was sent to the AAIB for download and analysis, but the duration of the flight from Lisbon meant that the takeoff portion of the flight had been overwritten by the time the aircraft landed. Analysis of the FDR data for the event showed that the takeoff roll was about 1,860 m long, with the aircraft becoming airborne 400 m before the upwind runway threshold, which it overflew at 100 ft climbing at about 2,700 ft/min. The airspeed at lift off was 170 KIAS. Airfield information Lisbon Airport has two runways which are orientated 03/21 and 17/35 as shown in Figures 1 and 3. Runway 03/21 is the preferred runway for both takeoffs and landings, and the prevailing winds mean that Runway 03 is more commonly used. At the time of publication, Runway 17/35 was expected to close and become a taxiway. For reasons described as “historic”, runway takeoff points are referred to as ‘Positions’ in the Aeronautical Information Publication2 (AIP) entry for Lisbon Airport. It is typical for airports elsewhere to use the intersection of taxiways with a runway to describe takeoff points. Commercial chart companies use information from the AIP to generate their publications and takeoff performance data, and they therefore refer to Positions at Lisbon Airport. However, Positions are not generally used by Lisbon ATC when issuing clearances. When Runway 21 is in use, the preferred takeoff point for all aircraft except heavy jets is ‘Position U’, which is the intersection of the runway with Taxiway U5. Pilots must advise ATC Footnote 2 Aeronautical Information Publication (AIP) is a publication issued by or with the authority of a State and containing aeronautical information of a lasting character essential to air navigation. (ICAO Annex 15 - Aeronautical Information Services.) 6 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-EZTD EW/C2019/04/03 on start-up if they require the full length of the runway for takeoff. Full-length departures are from Holding Point S4, which is known as ‘Position S’. Taxiway S begins abeam Runway 17, before crossing Runway 21 at Taxiway S1, and then turning north-east to run parallel to Runway 21 (marked on Figure 1 in blue). The taxiway ends at the threshold of Runway 21. There are therefore two points on Runway 21 where Taxiway S intersects the runway. Figure 3 Lisbon aerodrome ground chart © LIDO Operational procedures The operator uses an EFB to calculate the weight and balance of the aircraft as well as takeoff performance. Both pilots have a tablet computer on which they complete the required calculations. 7 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-EZTD EW/C2019/04/03 Electronic flight bag nomenclature Data for the EFB performance software is supplied to the operator by a third party. Within the software the crew must initially select the runway for departure and then a point on that runway from where the takeoff will begin. Some runways may have multiple intersections available for departure and, in the case of Lisbon Runway 21, two positions are available, Position U and Position S. These are named in the software as PSNU and PSNS. At the time of the incident, there was a NOTAM affecting the takeoff performance calculation (referring to an obstacle in the climb-out zone). This meant that the data supplier had inserted two further temporary selections for the two takeoff positions for Runway 21, which were labelled PSNUTMP and PSNSTMP as shown in Figure 4. Figure 4 EFB dropdown menu showing the all the intersections available The crew initially selected PSNUTMP for the performance calculation, ie intersection U5, but in discussing the likely takeoff point, they decided that they could use the S1 intersection if necessary, from which there was a lower TORA than from U5. They then performed the calculation from what they thought was the S1 intersection in the EFB selection: PSNSTMP. Operator’s procedures The operator has detailed standard operating procedures (SOPs) for calculating performance information for takeoff, and each pilot must make the calculation independently before push back. Before completing the performance calculation, the pilots must agree which intersection they will use for the calculation, using the one most likely to be used for takeoff. Should the aircraft depart from a less limiting intersection, no further performance calculation is required. The length of the runway selected is shown on the EFB calculation as shown in Figure 5. 8 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-EZTD EW/C2019/04/03 Figure 5 Performance calculation from the temporary Position S showing the distance display Both pilots are required to cross-check the runway distances available from the chosen intersection against the lengths displayed on the aerodrome ground chart, as shown circled in yellow in Figure 3 for Lisbon. Further event The operator subsequently reported an identical event which occurred with another company aircraft 14 days later. This event involved A320-214, registration OE-IJL, which departed Lisbon at 1906 hrs on 7 May 2019 for a flight to Paris Charles de Gaulle Airport. In this event, the aircraft lifted off 350 m before the upwind runway threshold which it crossed at about 75 ft aal. Further information The AAIB has investigated numerous serious incidents where aircraft have taken off using performance information calculated from a different start point. Worldwide, similar events present a significant hazard to civil aviation despite SOPs containing measures designed to prevent them, such as cross-checks and independent calculations. Pilots performing cross-checks often fail to notice errors or differences when the figures are unexpected. Humans are poorly adapted physiologically to discriminate between slightly-different acceleration rates, and many years of training have made pilots reluctant to move the throttles once takeoff power is set3. In recognition of this, the AAIB has previously Footnote 3 AAIB report into a serious incident in Belfast Aldergrove Airport, Boeing 737, C-FWGH, took off with insufficient thrust for the environmental conditions and struck an obstacle after lift-off https://www.gov.uk/aaib-reports/aircraft-accident-report- aar-2-2018-c-fwgh-21july-2017 [accessed December 2019] 9 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-EZTD EW/C2019/04/03 recommended that a technical barrier should be developed to capture the effects of an incorrect takeoff performance calculation when it occurs. The CAA has been working closely with EASA, operators, manufacturers and the AAIB to drive forward developments in mitigation strategies for takeoff performance errors. The strategies include increasing awareness in crews and operators about the criticality of takeoff performance data, development of flight data monitoring flags to detect takeoff performance errors, and the possibility of technological barriers to trap the effects of errors that are made. A copy of a letter on takeoff performance safety sent by the CAA to the CAT industry in December 2018 is at Appendix A. The aircraft manufacturer has developed a system aimed at protecting against incorrectly-calculated takeoff performance information for other types of aircraft within its fleet. This system performs a lift-off distance check and an aircraft position check before the aircraft begins its takeoff roll, and the manufacturer is in the process of extending the availability of this system to the A320 series of aircraft. The aircraft manufacturer indicated that the trial system would not have warned the crews of G-EZTD or OE-IJL against taking off because, at the start of the takeoff roll, the system-calculated value for runway remaining exceeded the forecast lift-off distance. Analysis During pre-flight preparation, both flight crew selected PSNSTMP in the EFB as the reference point for the takeoff performance calculation believing it to be where Taxiway S1 crossed Runway 21 whereas it was actually the reference point for the full length of the runway. The use of takeoff Positions gave rise to the situation where two points on Runway 21 could be construed by the crew as being ‘Position S’ within the EFB performance software. The operator’s SOPs required the crew to crosscheck the takeoff distance shown in the EFB against the equivalent distance shown on the aerodrome ground chart, but this crosscheck did not capture the error. Consequently, a lower thrust setting than required was used for the takeoff from S1 because it had been calculated for the full length of the runway (which had an additional 1,395 m available). After lifting off, the aircraft passed the upwind end of the runway at 100 ft aal. Another aircraft from the same operator, although operating under a different AOC, had an identical serious incident 14 days later. In both cases the pilots were confused by the EFB intersection selections because they did not refer to taxiway names, and the selection PSNSTMP could be confused between two runway intersections, S1 or S4. Therefore, the following Safety Recommendation is made: Safety Recommendation 2020-003 It is recommended that ANA Aeroportos de Portugal discontinue the use of takeoff ‘Positions’ at Lisbon Airport to minimise confusion in relation to takeoff points. 10 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-EZTD EW/C2019/04/03 Conclusion Both aircraft took off using incorrect performance data for the intersection used. In each case, a selection error was made in the EFB which led the crew to believe that they had calculated performance information for a departure from S1 when in fact they had selected the full length of the runway. In both cases, the procedural barrier of cross-checking the runway distance against the aerodrome ground chart failed to prevent to error. Human performance limitations mean it is difficult for pilots to recognise and react to reduced performance (acceleration) once the takeoff has begun, so robust adherence to procedures is a key defence against such incidents occurring. Safety action As a result of these serious incidents the following safety action was taken: ● The aircraft operator issued a notice to its flight crew clarifying the takeoff positions available on Runway 21 at Lisbon Airport. ● A NOTAM was issued highlighting ‘confusing runway holding point naming’ and reminding crews that ‘Position S’ referred to the full length of Runway 21 (Figure 6). ● The aircraft operator issued a description of the events and their causes to its flight crew to raise awareness of the risks of using the wrong intersection and distance for takeoff. ● The aircraft operator engaged with the aircraft manufacturer to review future developments that could offer extra protections against events such as those covered in this report. ● The airport authority undertook to rename taxiways so that Taxiway S intersected the runway at only one point; S4 (full length). Figure 6 Crew NOTAM 11 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-EZTD EW/C2019/04/03 Appendix A Letter from the CAA to the Commercial Air Transport industry 12 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-EZTD EW/C2019/04/03 Published: 16 January 2020. 13 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-SAJK and G-CDMH EW/C2019/08/03 SERIOUS INCIDENT Aircraft Type and Registration: 1) EMB-145EP, G-SAJK 2) Cessna P210N Pressurized Centurion, G-CDMH No & Type of Engines: 1) 2 Allison AE 3007/A1/1 turbofan engines 2) 1 Continental Motors TSIO-520-P piston engine Year of Manufacture: 1) 1999 (Serial no: 145153) 2) 1978 (Serial no: P210-00131) Date & Time (UTC): 7 August 2019 at 1740 hrs Location: London Southend Airport Type of Flight: 1) Commercial Air Transport (Passenger) 2) Private Persons on Board: 1) Crew - 3 Passengers - 35 2) Crew - 1 Passengers - None Injuries: 1) Crew - None Passengers - None 2) Crew - None Passengers - N/A Nature of Damage: 1) None reported 2) None reported Commander’s Licence: 1) Airline Transport Pilot’s Licence 2) Private Pilot’s Licence Commander’s Age: 1) 39 years 2) 51 years Commander’s Flying Experience: 1) 7,453 hours (of which 1,075 were on type) Last 90 days - 110 hours Last 28 days - 74 hours 2) 1,930 hours (of which 169 were on type) Last 90 days - 23 hours Last 28 days - 10 hours Information Source: AAIB Field Investigation Synopsis An Embraer 145 landing at London Southend Airport ran over a general aviation towbar which had been dropped on the runway. No damage was caused to the aircraft. The investigation found that the towbar had fallen from a Cessna 210 which departed Southend Airport 30 minutes before. The Cessna pilot had likely been distracted during his pre-flight checks by an earlier road traffic incident in which he was involved, and had inadvertently left the towbar attached. One Safety Recommendation has been made to the CAA to improve the visibility of general aviation ground equipment. 14 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-SAJK and G-CDMH EW/C2019/08/03 History of the flight G-SAJK – Embraer 145 G-SAJK was operating a scheduled service from Aberdeen to Southend, and made a normal approach to Runway 23 at Southend Airport, landing at 1815 hrs. The weather conditions were CAVOK. On landing, as the commander applied the brakes, he saw an object on the right of the centreline approximately 8 – 10 m in front of the aircraft. He estimated that the aircraft was travelling at between 105 and 110 kt at this stage. He applied slight left rudder as the object disappeared out of view and felt a small bump through the rudder pedals but was not sure if this was caused by the aircraft clipping the object or running over the centreline. The aircraft stopped, backtracked and vacated the runway normally. The commander reported the sighting to ATC who requested a runway inspection, which found a general aviation towbar (Figure 1) on the runway. There were no indications of any damage to the aircraft, so the commander continued to taxi the aircraft to stand. After the passengers had disembarked, both pilots inspected the aircraft but found no damage. The commander recalled that the object had been very difficult to see against the dark asphalt runway. He only saw it because part of the towbar was lying across the white centreline markings. He recalled that it was located just after Taxiway B. Shortly after the Embraer landed, Southend ATC received a call from Farnborough Radar. They had been notified by the pilot of a Cessna P210 (G-CDMH), which had departed Southend Airport at 1747 hrs, that he thought he may have departed with the towbar still attached. G-CDMH - Cessna 210 The pilot flew to Southend Airport regularly as he had an office nearby. He travelled between his office and the airport by motorcycle. On the day of the incident, whilst riding to the airport, a cyclist pulled out in front of him. He was able to miss the cyclist, and no one was injured, but the pilot described it as “a fright and a close shave.” He continued to the airport, pulled the aircraft out of the hangar and completed the pre-flight checks. The start-up was uneventful but, as he taxied to the runway, he noticed a slight tendency for the aircraft to track to the left. However, he considered it minor and made a mental note to check the tyre pressures on landing. The pilot reported that the takeoff from Runway 23 seemed entirely uneventful. Approximately 30 minutes into the flight he was thinking about the tracking issue and it occurred to him that he could not positively remember removing and stowing the towbar. He knew the towbar was no longer attached to the aircraft because the landing gear had successfully retracted. He immediately reported his concern to Farnborough Radar and asked for a message to be passed to Southend. The pilot continued his planned flight and landed without further incident. He inspected the aircraft after landing and did not find any damage. 15 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-SAJK and G-CDMH EW/C2019/08/03 The pilot discovered that he had also left his bags behind in the hangar at Southend. He reflected that he was distracted by the earlier motorcycle incident and that this was “on his mind” whilst completing the pre-flight checks. He reported that “the towbar is a small stowable unit that does not extend outwards much more than the tip of the spinner, but it is quite obvious and I cannot believe that I missed it.” ATC report After the landing Embraer reported seeing the towbar, ATC asked an operations vehicle to inspect the runway. They recovered the towbar from the runway approximately 50 m to the west of Taxiway B (Figure 1). There were scuff marks on the handle of the towbar. Between G-CDMH departing and G-SAJK landing two other aircraft had used the runway and a runway inspection had been carried out. Table 1 gives a summary of the timeline provided by ATC. TIME EVENT 1739 G-CDMH has cleared to taxi to holding point D where the aircraft carried out power checks. 1743 G-CDMH reported ready for departure. 1746 G-CDMH was cleared for takeoff and was recorded as airborne at 1747 hrs. G-CDMH was subsequently handed-over to Southend Radar. 1749 A PA-28 on a local flight was cleared to land on Runway 23. 1752 The PA-28 landed. 1753 An Operations vehicle was cleared to enter the Runway at C1 to complete an inspection. The vehicle entered the Runway initially in a south-westerly direction and on return vacated at Taxiway D at 1757 hrs. 1759 A Britten-Norman Islander was cleared to taxi to A1 and was then cleared to takeoff at 1801 hrs. The aircraft was recorded as airborne at 1803 hrs. 1811 G-SAJK was cleared to land. 1815 G-SAJK landed and the commander reported the aircraft had colliding with a general aviation towbar on the runway. 1817 An operations vehicle recovered the towbar adjacent to Taxiway B. 1825 ATC received a call from Farnborough Radar reporting the G-CDMH may not have removed their towbar prior to departure. Table 1 Summary of the timeline of the incident ATC reported that it was not possible to see the towbar from the control tower due to its size, shape and colour. They also reported that at the time of the event the location of the evening sun made it harder to see objects on the runway. 16 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-SAJK and G-CDMH EW/C2019/08/03 Figure 1 Towbar recovered from the runway Runway inspection report At 1751 hrs a fire officer driving an airport operations vehicle requested permission to enter the runway for a wildlife inspection. Permission was granted at 1753 hrs. He entered the runway at C1 and preceded south-west toward the Runway 05 threshold. He reported that when he entered the runway there was no wildlife activity, so he carried out a surface inspection. From the Runway 05 threshold he recalled that he drove to the Runway 23 threshold then back to Taxiway D where he vacated. He did not see the towbar. He vacated the runway at 1757 hrs. It was reported that because his initial intention was to complete a wildlife inspection his attention may have been directed towards the sky rather than on the surface which may be why he did not see the towbar. 17 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-SAJK and G-CDMH EW/C2019/08/03 Airfield information London Southend Airport has a single asphalt runway orientated 05/23 (Figure 2). Towbar location Figure 2 London Southend Airport Chart showing approximate location the towbar was found. The airport does not have any electronic means of detecting objects on the runway, taxiways or apron1 so relies on them being seen during runway inspections or being reported by other airfield users. London Southend Airport procedures specify a minimum of one runway inspection every four hours and two complete movement area inspections daily. On the day of the incident, airport records show that the last full runway inspection was completed at 1515 hrs. The inspection which occurred between the Cessna departing and the Embraer landing was intended to be a wildlife inspection and not considered a full runway inspection. Personnel The pilot of the Cessna 210 (G-CDMH) held a Private Pilot’s Licence with valid Single Engine Piston, Multi Engine Piston and Instrument (IR) ratings. He had a total of 1,930 flying hours with 169 hours on the Cessna 210. He had flown 10 hours in the last 28 days and 23 hours in the last 90 days. Footnote 1 Some larger airports have ‘foreign object detection radar’ which can detect objects on the runway and issue alerts to ATC. 18 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-SAJK and G-CDMH EW/C2019/08/03 Pre-flight checklist During the airport’s investigation into this incident it was noted that the pre-flight checklist used by the pilot did not include any reference to ensuring the towbar, or other ground equipment, is removed before flight. Following the incident, the pilot amended his checklist to add a reminder to remove and stow the towbar. Additionally, he added a visual reminder in the cockpit of G-CDMH. Other human factors The pilot was involved in a motorcycle incident during his drive to the airport. He reported that this incident was “on his mind” during his pre-flight checks and probably contributed to him forgetting to remove the towbar. Additionally, he left two of his bags behind at the airport which also suggests he was distracted. A stressful or traumatic event can be distracting and difficult to put out of mind. It may be tempting to continue with a planned operation and not realise the effect of such an event on subsequent performance. The CAA Skyway Code2 highlights the importance of pilots assessing their fitness to fly before any flight. The code suggests using the ‘IM SAFE’ mnemonic for self-assessing fitness for flight (Figure 3). In this incident ‘stress’ from the earlier motorcycle incident was probably a key factor. Figure 3 ‘IM SAFE’ mnemonic highlighted in the CAA Skyway Code Ground equipment markings The towbar was painted in a dark blue paint (Figure 1) which made it difficult to see against the dark asphalt runway surface. Footnote 2 CAA skyway Code is available at https://www.caa.co.uk/General-aviation/Safety-information/The-Skyway- Code (accessed 16 October 2019) 19 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-SAJK and G-CDMH EW/C2019/08/03 The International Air Transport Association (IATA) Airport Handling Manual (AHM)3 provides recommendations for aircraft ground support equipment. AMH 913 section 14 lists the following recommendations for non-motorised ground support equipment: ‘14.2.1 Non-motorised Ground Support Equipment should be visible to the operator(s) of any approaching Ground Support Equipment within the safety braking distance and under any angle of approach. 14.2.2 Colour schemes for markings of reflective material should be in compliance with the marking and illumination standards established by the local regulatory authorities. The reflective material shall be resistant to wear and tear. 14.2.3 Non-motorised Ground Support Equipment should have its presence accentuated by application of reflective material on all sides of the equipment inclusive of the tow-bar, outriggers or any other deployable devices. 14.2.4 Non-motorised Ground Support Equipment should have a minimum of 2 markings of reflective material on each side of the equipment with a maximum separation distance of 1.5 m (60 in) between each marking. Each marking should not be less than 100 cm2 (15.49 in2). 14.2.5 Markings of reflective material should also be applied on all comers of the equipment.’ Analysis The pilot of a Cessna 210 departing from Southend Airport inadvertently left the aircraft’s towbar attached to the nosewheel. As the aircraft took off the towbar fell off and landed on the runway. The towbar remained on the runway for approximately 30 minutes during which time another aircraft landed, another took off and an operations vehicle completed an inspection. A landing Embraer 145 ran over the towbar during its landing roll. The pilot of the Cessna reported that he was distracted by an early road traffic incident and this is probably why he forgot to remove the towbar. The incident highlights how stress from events unrelated to flying can cause a significant distraction and the importance of pilots honestly assessing their fitness for flight prior to every flight. The towbar was not seen on the runway by two other aircraft that used the runway nor by a fire officer conducting an inspection. It is not known exactly which part of the runway the two aircraft used, so it is possible they did not pass the towbar. Alternatively, their attention may have been on flying their aircraft. The inspection was initially intended to be a wildlife inspection, so it is possible that the driver’s attention was focused towards the sky rather than the runway surface. However, the towbar was painted in dark colours so it did not stand out against the runway surface. The towbar might have been seen sooner if it had reflective or other high visibility markings. Footnote 3 IATA Airport Handling Manual is available at https://www.iata.org/publications/store/Pages/airport-handling- manual.aspx (accessed 16 October 2019) 20 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-SAJK and G-CDMH EW/C2019/08/03 Most airports do not currently have automatic means of detecting objects on the manoeuvring area, so they rely on them being seen during inspections or being seen by other airport users. Therefore, it is important that any equipment that could be left on a manoeuvring area is highly visible. The IATA AHM provides recommendations for ground handling equipment to ensure it is clearly visible. However, these are not widely applied across general aviation ground equipment. Making ground equipment more visible would reduce the likelihood of it being left attached to the aircraft and increase the chance of it being seen quickly if it is left on a runway or manoeuvring area. During this incident no damage was caused to the landing aircraft. However, objects on the manoeuvring area have the potential to cause serious harm to aircraft. Therefore, the following Safety Recommendation is made: Safety Recommendation 2020-004 It is recommended that the Civil Aviation Authority communicate to the general aviation community the importance of increasing the visibility of ground equipment. Conclusion A general aviation towbar was inadvertently left attached to an aircraft because the pilot had been distracted by an earlier stressful event during his journey to the airport. The towbar dropped onto the runway during the departure and remained there for approximately 30 minutes, during which two other aircraft used the runway and a runway inspection was completed. A landing aircraft then ran over it. The towbar was inconspicuous because it did not have any reflective or other high visibility markings. Safety action The CAA has stated that if, during the general aviation-specific audits and inspections it conducts, it observes ground equipment that due to its colour is not sufficiently visible, it will bring this to the attention of the relevant operator. Published: 23 January 2020. 21 © Crown copyright 2020 All times are UTC AAIB Correspondence Reports These are reports on accidents and incidents which were not subject to a Field Investigation. They are wholly, or largely, based on information provided by the aircraft commander in an Aircraft Accident Report Form (AARF) and in some cases additional information from other sources. The accuracy of the information provided cannot be assured. AAIB Bulletin: 2/2020 23 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 EI-DEO EW/G2019/08/14 SERIOUS INCIDENT Aircraft Type and Registration: Airbus A320, EI-DEO No & Type of Engines: 2 CFMI CFM56-5B4/PS turbofan engines Year of Manufacture: 2005 Date & Time (UTC): 12 August 2019 at 2100 hrs Location: En route from Dublin Airport to London Heathrow Airport Type of Flight: Commercial Air Transport (Passenger) Persons on Board: Crew - 6 Passengers - 173 Injuries: Crew - None Passengers - None Nature of Damage: None Commander’s Licence: Airline Transport Pilot’s Licence Commander’s Age: 42 years Commander’s Flying Experience: 8,000 hours (of which 5,000 were on type) Last 90 days - 16 hours Last 28 days - 6 hours Information Source: Aircraft Accident Report Form submitted by the pilot After completing a cross-bleed start1, both flight crewmembers experienced a strong fuel/ oil smell which they assumed was as a result of the start procedure. During taxi, they experienced further fuel/oil fumes and thought it was due to the exhaust of an aircraft taxiing ahead. Further intermittent occurrences of fumes were noticed during departure and climb which they discussed. Whilst the flight crew were conducting their approach brief, they discussed that they both felt they were not operating to their normal standard and agreed they would maintain a heightened awareness. They did not think it necessary to don oxygen masks. The approach and landing were without incident. After landing the flight crew opened the cockpit direct vision windows and taxied onto stand. During a post flight review, unsure of their medical condition, both flight crew donned the oxygen masks for a short while. A basic medical examination of the flight crew showed that all their vital signs were normal. The aircraft was subjected to extensive fault finding by engineering staff, no cause for the fumes was identified and the aircraft has returned to operation with no further reports of fumes. Footnote 1 The aircraft was operating with its APU inoperative. As a result, the first engine had to be started on stand using a ground air supply. The second engine was then started after push-back using bleed air from the first engine using a cross-bleed procedure. 24 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 EI-DEO EW/G2019/08/14 The operator completed an internal investigation and is reviewing smoke and fumes guidance material provided to its flight crew. 25 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-VFIT EW/G2019/08/30 ACCIDENT Aircraft Type and Registration: Airbus A340-642, G-VFIT No & Type of Engines: 4 Rolls-Royce RB211 Trent 556-61 turbofan engines Year of Manufacture: 2006 (Serial no: 753) Date & Time (UTC): 21 August 2019 at 1230 hrs Location: En route from New Delhi Airport, India to London Heathrow Airport Type of Flight: Commercial Air Transport (Passenger) Persons on Board: Crew - 15 Passengers - 305 Injuries: Crew - 1 (Minor) Passengers - 1 (Serious) Nature of Damage: None Commander’s Licence: Airline Transport Pilot’s Licence Commander’s Age: 54 years Commander’s Flying Experience: 18,062 hours (of which 5,722 were on type) Last 90 days - 162 hours Last 28 days - 62 hours Information Source: Aircraft Accident Report Form submitted by the pilot and further enquires by the AAIB Synopsis Whilst passing through northern Turkey at FL360 the aircraft encountered moderate to severe turbulence during which a passenger was severely injured. The turbulence was not forecast and there were no indications, visually or on the aircraft’s weather radar, to suggest the aircraft was approaching an area of turbulence. History of the flight G-VFIT, an Airbus A340, was flying from New Delhi Airport, India to London Heathrow Airport (Heathrow). At approximately 1230 hrs, five and a half hours into the flight, the aircraft was over northern Turkey, at FL360, routing toward the border with Bulgaria (Figure 1). The commander had just returned from his rest period. The relief co-pilot was in the right seat and was the pilot flying. The weather forecast for the area was for isolated embedded cumulonimbus clouds with tops at FL320. The commander reported that as they approached waypoint OLUPO (Figure 1) he saw a single “non-threatening” cloud on their track which was not showing on the aircraft’s weather radar. There were cumulonimbus clouds on the horizon but no other clouds in their immediate vicinity. The aircraft was then cleared from OLUPO direct to waypoint ODERO on the Turkey/Bulgaria border, which took the aircraft clear of the single cloud. However, as the aircraft passed abeam the cloud, they encountered moderate to severe turbulence. The 26 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-VFIT EW/G2019/08/30 co-pilot saw the speed trend arrow indicating +40 kt and the speed increasing rapidly. To prevent an overspeed he selected a lower Mach number and extended the speed brakes. The commander turned the seatbelt signs on. The commander recalled seeing the wind shifting from a 40 kt tailwind to a 20 kt headwind in 2-3 seconds. The worst of the turbulence lasted for approximately 5 seconds. The aircraft was then subject to light turbulence until able to climb to FL380 at 1237 hrs. Approximate location of turbulence encounter Figure 1 G-VFIT route through Turkey showing location of turbulence event When the turbulence occurred the senior cabin crew member on duty made an announcement to reassure the passengers and to instruct the passengers and cabin crew to take their seats and fasten their seat belts. After the aircraft was clear of the turbulence the commander was informed that a passenger in the rear galley had fallen over in the turbulence and had injured his ankle. The passenger was moved to a more comfortable seat and medical advice was obtained from a ground- based service provider1. The aircraft continued to Heathrow without further incident. A medical emergency was declared on arrival in the London FIR and a priority approach was requested and granted. On arrival, the aircraft was met by paramedics who treated the passenger. He was taken to hospital with a suspected broken ankle. He had surgery on his ankle the next day and was released from hospital seven days later. One cabin crew member reported that he injured his shoulder when he fell against a bulkhead during the turbulence, but no medical treatment was required. Footnote 1 Flight crew can contact the service in-flight, via satellite phone, to receive emergency medical advice from a doctor. 27 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-VFIT EW/G2019/08/30 Passenger report The injured passenger reported that he was waiting to use the toilet in the rear galley when the turbulence occurred. The seatbelt signs were off, and he recalled that he suddenly found himself lying on the floor not knowing what had happened. His right ankle was extremely painful and swelled up within a few seconds. A cabin crew member tried to get him to move back to his seat, but he was unable to get up. His wife and son helped him back to his seat and the cabin crew brought him an ice pack. Later the crew moved him to a seat with more space and gave him painkillers. However, his ankle continued to be extremely painful. After landing he was taken to hospital, where it was found that his ankle was dislocated and broken in two places. Recorded data The operator reviewed the available flight data which showed that at 1230 hrs the aircraft was at FL360 and Mach 0.82. The turbulence occurred at 1231:08 hrs with vertical acceleration varying between +0.64 g and +1.56 g, the speed increasing to Mach 0.86 and the aircraft rolled +/- 5°. The seatbelt signs were turned on at 1232:57 hrs. Meteorology Forecast information available to the flight crew Parts of the significant weather and wind charts issued to the flight crew prior to the flight are shown in Figures 2 and 3. The significant weather chart shows an area of isolated embedded cumulonimbus clouds with tops at FL320 in the area where the turbulence occurred. There was no clear air turbulence forecast in this area. The wind chart shows the wind circulating round a low pressure centred in north-west Turkey. Figure 2 Segment of the significant weather chart valid at 1200 hrs on 21 August 2019 28 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-VFIT EW/G2019/08/30 The significant weather chart included a note that cumulonimbus cloud implies moderate or severe turbulence. However, these clouds were below the cruise level of G-VFIT and the commander reported that the aircraft was clear of cloud when the turbulence occurred. Ankara FIR had issued six SIGMETs2, valid at the time of the turbulence event, which reported that thunderstorms had been observed in the area. There was no forecast or report of turbulence. Flight plans issued to the flight crew provide forecast wind and temperature on the route which can give an indication of likely turbulence. The forecast wind at OLUPO was 202° at 63 kt; over the next 300 nm the wind was forecast to gradually change to an easterly wind. The temperature in this region at FL360 was forecast to be steadily decreasing from -43°C to -46°C. The plan also provided a forecast vertical shear rate in knots per 200 ft, which can indicate an area of turbulence. The value at OLUPO was one, increasing to four over the next 300 nm. The commander reported that shear values in single digits do not normally indicate significant turbulence. Figure 3 Segment of the wind chart valid at 1151 hrs on 21 August 2019 The flight crew used a charting application on their tablet devices which showed en route weather layers, forecast winds and areas of turbulence. The commander reported that Footnote 2 A SIGMET (SIGnificant METrological information) is a notification of an en route weather phenomena which may affect the safety of aircraft operations. 29 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-VFIT EW/G2019/08/30 the application did not show any turbulence in this area. At the time of the accident the operator only approved flight crew to upload data to the tablet on the ground, so it only contained the information available before the aircraft departed. Since the accident the operator has approved the use of the onboard WiFi system to update the application in- flight and to provide the most up-to-date information to the flight crew. Aftercast weather Figure 4 shows a derived satellite image showing cloud top heights at the time of the incident, produced by the Met Office after the accident. The image indicates there was a large area of cloud with tops between 30,000 ft and 35,000 ft across the area in question. Figure 4 Derived Satellite Image showing Cloud Top Heights valid at 1230 hrs on 21 August 2019 Seatbelt signs The seatbelt signs were selected off when the turbulence occurred. Flight crew will normally turn the seatbelt signs on if they anticipate turbulence, but on this occasion there was no indication that turbulence was likely to occur. In general, flight crew face a difficult balance in deciding when to turn the seatbelt signs on. If there are clear indications or reports of turbulence ahead, the signs can be turned on in advance. However, if the flight crew put the signs on every time there is any chance of turbulence, it is likely the signs would be on for extended periods in smooth conditions, which can be frustrating for passengers. So, flight crew must balance the need to ensure the signs are on prior to any significant turbulence with the need to ensure the are not on unnecessarily. 30 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-VFIT EW/G2019/08/30 Analysis The aircraft encountered turbulence over northern Turkey during which a passenger was seriously injured. The turbulence was caused by a rapid wind shift. There was no information available to the flight crew, either in the pre-flight paperwork, from the weather radar or visually to forewarn the flight crew that they were approaching an area of turbulence. Therefore, the seatbelt signs were off when they encountered the turbulence. 31 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 EI-DPK EW/G2019/02/24 INCIDENT Aircraft Type and Registration: Boeing 737-8AS, EI-DPK No & Type of Engines: 2 CFM56-7B27 turbofan engines Year of Manufacture: 2007 (Serial no: 33610) Date & Time (UTC): 26 February 2019 at approximately 1105 hrs Location: London Stansted Airport Type of Flight: Commercial Air Transport (Passenger) Persons on Board: Crew - 6 Passengers - 169 Injuries: Crew - None Passengers - 1 (Minor) Nature of Damage: None reported Commander’s Licence: Airline Transport Pilot’s Licence Commander’s Age: 39 years Commander’s Flying Experience: 5,000 hours (of which 1,800 were on type) Last 90 days - 236 hours Last 28 days - 85 hours Information Source: Aircraft Accident Report Form submitted by the operator and additional enquiries made by the AAIB Synopsis As passengers disembarked the aircraft using the forward integral airstairs, a child fell to the ground through the gap between the handrails. The child continued its journey without treatment after being assessed by medical personnel. Boeing 737 forward airstairs Some Boeing 737 series aircraft are fitted with a set of retractable airstairs at the forward left cabin door, to allow the boarding and disembarkation of passengers without the need for additional ground support equipment. The airstairs include an integral two-rung handrail on either side. These rise into position during deployment of the stairs, but due to the geometric restrictions imposed by the retraction mechanism design, they do not extend to the fuselage side. In order to bridge this gap between the top of the handrails and the fuselage, a manually extendable handrail is fitted to each of the integral rails. After deployment of the airstairs, these are extended and secured to points in the entry door frame (Figure 1). Each extendable rail is supported by a strut extending from the side rail of the airstairs. 32 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 EI-DPK EW/G2019/02/24 Figure 1 Boeing 737 forward airstairs History of the flight and reporting of the event The flight originated in Oslo and the passengers were disembarking on arrival in Stansted. It was daytime and weather conditions were reported to be dry. As a child left the aircraft with his family, he stumbled and fell sideways to the ground from the top two or three steps of the integral airstairs. The exact sequence of events leading to the fall was unclear, but the father reported that the child was walking while holding his hand. He felt the child pull from his hand and then fall sideways from the stairs. After assessment by airport medical personnel, the child was allowed to continue their journey without treatment. The flight crew did not know that the child had fallen and, therefore, did not submit an Air Safety Report (ASR). The operator stated that their Customer Services department became aware of the fall in March 2019 when they received external correspondence relating to the event. In accordance with their internal procedures, they obtained reports from the ground crew and cabin crew, with the latter being reminded of the operator’s procedures, which require that all inflight events are reported to the aircraft captain. The operator did not inform the AAIB because the child’s injuries did not constitute an accident or serious incident as defined in ICAO Annex 13 or Regulation (EU) 996/2010. Irrespective of this, however, the event should have been reported to the Irish Aviation Authority (IAA) under Regulation (EU) 376/2014. This regulation requires the mandatory reporting of occurrences related to injury. 33 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 EI-DPK EW/G2019/02/24 The AAIB became aware of the event in April 2019 and decided to investigate the circumstances under Article 5 of Regulation (EU) 996/2010. This decision was made in the knowledge of a previous similar event on EI-DLJ (AAIB Bulletin 8/2010) and the potential for a more serious outcome. The AAIB contacted the Air Accident Investigation Unit (AAIU) in Ireland, who subsequently contacted the operator’s Safety Services Office. On becoming aware of the event, the Safety Services Office collated the cabin crew reports, processed an ASR, and submitted a Mandatory Occurrence Report to the IAA. Investigation The scope of the investigation was limited because the exact circumstances of the fall were unclear. There were four cabin crew on the flight, but none of them saw the child fall. The airport operator reviewed the CCTV footage, but the forward left door of the aircraft was outside the field of view and the fall was not recorded. Previous events within Europe The AAIB identified eight previously reported events in Europe since 2009 and six of these involved children of various ages. Typically, the airstair equipped Boeing 737 fleet completes over 780,000 flights per year, carrying over 120 million passengers across all age groups. The aircraft manufacturer confirmed that these events had been reported to them and reviewed in accordance with the Boeing / Federal Aviation Administration Continued Operational Safety Process. They considered that the existing safety actions provided adequate mitigation (see previous safety actions, below). Certification of integral airstairs Federal Aviation Regulation 25 and its European equivalent, Certification Standard 25, define the airworthiness standards for transport category aeroplanes. The specifications do not contain any requirements relating to integral airstairs. EASA advised that the Boeing 737-800 approval within Europe was issued by the Joint Aviation Authorities (JAA) following a validation activity in 1998. The integral airstair design is understood to have been accepted on the basis of previous Boeing 737 models. Previous safety actions Anti-skid material and warning placards In September 2007, the Federal Aviation Administration issued a Special Airworthiness Information Bulletin (SAIB) after four reports of injuries resulting from small children falling through or over the airstair handrails1. The bulletin recommended the introduction of anti-skid material and warning placards to advise people accompanying children to hold the child’s hand whilst on the stairs. Footnote 1 Special Airworthiness Information Bulletin NM-07-47 http://rgl.faa.gov/Regulatory_and_Guidance_Library/ rgSAIB.nsf/dc7bd4f27e5f107486257221005f069d/cab005ca55f1abd78625734e006eb6b7/$FILE/NM-07- 47.pdf (accessed 25 September 2019) 34 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 EI-DPK EW/G2019/02/24 The operator of EI-DPK confirmed that warning placards and anti-slip material were in place on the aircraft when the child fell. In addition to these recommendations, after the previous AAIB investigation, the operator installed a retractable roller-tensioned tape and detachable airstair rails to further reduce the likelihood of somebody falling. Boarding and disembarking announcements The previous AAIB investigation resulted in Safety Recommendation 2010-018. This was made applicable to all UK operators of 737 aircraft with integral airstairs in April 2011 when the UK Civil Aviation Authority issued a Safety Notice (SN-2011/02) entitled ‘Safe Use of Airstairs.’ The safety notice required operators to review their boarding and disembarkation procedures so that special assistance is made available to passengers accompanied by small children, or those with special needs. The notice stated that operators: ‘should also review announcements made by staff at the boarding gate and before disembarkation to ensure that passengers’ attention is drawn to the need to exercise caution when boarding and disembarking using airstairs. Passengers in these circumstances should, in particular, be advised to keep small children under close supervision throughout the boarding and disembarkation process.’ This Safety Notice was cancelled in 2018. The operator’s ground operations manual contains the pre-boarding announcement, which includes the words: ‘adults with young children must hold their hands whilst walking to the aircraft and on the aircraft steps. Use the handrails provided.’ The Safety Equipment Procedures manual contains the before disembarking announcement, which contains the words: ‘All passengers should use the handrail provided when walking down the stairs. For passengers travelling with children please hold their hands as you walk down the stairs and until you are inside the terminal building.’ Discussion The child continued his journey without medical treatment, so the event did not meet the criteria of an accident or serious incident defined in ICAO Annex 13 or Regulation (EU) 996/2010. However, the investigation was instituted because safety lessons were expected to be drawn from it. The flight crew were unaware of the fall, which occurred within the defined period of a flight. The operator reminded their cabin crew that their incident reporting procedures requires them to notify the aircraft captain of events that occur in flight. If appropriate, the flight crew will submit an ASR and the regulator can be notified if necessary. In the case of this event, 35 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 EI-DPK EW/G2019/02/24 there was a mandatory requirement to notify the IAA under Regulation (EU) 376/2014. In this case there was a delay of approximately two months because the flight crew were unaware of the fall. There are no certification requirements for integral airstairs and in the last 10 years, there have been nine reports (within Europe) where people, both adults and children, have fallen whilst using them. The manufacturer is aware of these events and considers the recommendations in the FAA SAIB to be adequate mitigation. When considering the arising rate, the operator reported that in 2018, they completed over 787,000 flights and carried over 126 million passengers across all age groups. 36 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-STBB EW/G2019/08/29 INCIDENT Aircraft Type and Registration: Boeing 777-36, G-STBB No & Type of Engines: 2 General Electric Co GE90-115B turbofan engines Year of Manufacture: 2010 (Serial no: 38286) Date & Time (UTC): 24 August 2019 at 1740 hrs Location: En route from London Heathrow Airport to JFK Airport, New York Type of Flight: Commercial Air Transport (Passenger) Persons on Board: Crew - 17 Passengers - 302 Injuries: Crew - None Passengers - None Nature of Damage: Burnt out oven motor in aft galley Commander’s Licence: Airline Transport Pilot’s Licence Commander’s Age: 53 years Commander’s Flying Experience: 13,899 hours (of which 4,827 were on type) Last 90 days - 173 hours Last 28 days - 39 hours Information Source: Aircraft Accident Report Form submitted by the pilot Whilst in the initial part of the cruise, the captain was called by the senior cabin crew member informing him that there was a suspected fire in the aft galley. Shortly afterwards the cabin crew member confirmed that blue smoke was emanating from the oven unit. The cabin crew carried out their fire drill and the flight crew completed the ‘smoke, fire or fumes’ checklist. The flight crew then informed ATC of the event. After electrically isolating the aft galley the rate of smoke production reduced; however, smoke remained in the area. The captain made the decision to divert to Shannon, where an uneventful overweight landing and disembarkation took place. There were no injuries to any of the passengers. Examination of the aft galley oven by the operator’s engineers found that an electrical fault in an oven fan unit had caused the smoke. The fan unit was replaced and the aircraft returned to service. No further investigation of the oven fan unit was conducted by the operator. 37 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-DDAY EW/G2019/10/18 ACCIDENT Aircraft Type and Registration: Piper PA-28R-201T Turbo Cherokee Arrow III, G-DDAY No & Type of Engines: 1 Continental Motors Corp TSIO-360-F piston engine Year of Manufacture: 1977 (Serial no: 28R-7703112) Date & Time (UTC): 22 October 2019 at 1400 hrs Location: Nottingham Airport Type of Flight: Private Persons on Board: Crew - 1 Passengers - 1 Injuries: Crew - None Passengers - 1 (Minor) Nature of Damage: Damaged beyond economic repair Commander’s Licence: Private Pilot’s Licence Commander’s Age: 69 years Commander’s Flying Experience: 240 hours (of which 91 were on type) Last 90 days - 0 hours Last 28 days - 0 hours Information Source: Aircraft Accident Report Form submitted by the pilot The pilot, with a passenger, was conducting a 90-day, three takeoffs and landings revalidation and had completed two of these without incident. The weather was good, and the wind was from 230° at 10 kt. On the third landing, the aircraft touched down on Runway 21 ‘on the numbers’; the pilot reduced flap to two stages and selected full throttle. Directional control was lost, and the aircraft veered off the runway to the left and became airborne just as the propeller contacted rough ground. The aircraft became difficult to control and the pilot flew a low downwind leg to return to the airfield. Once within the airfield boundary, the aircraft became uncontrollable and it struck the ground just north of Runway 27 incurring substantial damage. Both pilot and passenger were wearing lap and diagonal harnesses and only the passenger suffered minor injuries. The reason for the loss of directional control is not known. 39 © Crown copyright 2020 All times are UTC AAIB Record-Only Investigations This section provides details of accidents and incidents which were not subject to a Field or full Correspondence Investigation. They are wholly, or largely, based on information provided by the aircraft commander at the time of reporting and in some cases additional information from other sources. The accuracy of the information provided cannot be assured. AAIB Bulletin: 2/2020 41 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 Record-only investigations reviewed November - December 2019 13-May-19 Pioneer 300 G-OWBA Oxenhope Airfield The aircraft floated in the flare and landed half-way along the runway following a higher than normal approach. The pilot was unable to stop the aircraft before reaching the end of the runway. 06-Jul-19 Robinson R22 Beta G-WINR Princes Risborough The student pilot accidentally shut down the engine by moving the mixture lever rather than the carburettor heat. Autorotation into a field was successful, but the helicopter rolled onto its side after touchdown due to uneven ground. 06-Aug-19 Rotorway Executive 90 G-BVOY Deenethorpe Airfield While taxiing out-of-wind the tail of the helicopter was lifted by a gust and the front of a skid caught the ground, causing the helicopter to roll over. 21-Aug-19 Piper PA-17 Vagabond G-ALIJ Shoreham Airport The aircraft veered to the left unexpectedly after touchdown. When the pilot attempted to brake and correct, the aircraft tipped onto its nose. 25-Aug-19 DH82A Tiger Moth G-AYDI Lodge Farm House, Wantage Oxfordshire During landing a crosswind gust caused the aircraft to veer to the right. The right wheel and wing contacted the ground and caused the aircraft to flip over damaging the propeller, rudder, fin and lower mainplane. 12-Sep-19 Jabiru SK G-BYBZ Private airstrip, Durham The aircraft landed off the centre line. The main landing gear caught the long grass and the aircraft left the runway at low speed. 03-Nov-19 Morane Saulnier Rallye 150ST G-BERA Redhill Aerodrome, Surrey The aircraft departed the runway edge on landing and hit a marker board causing minor damage to the trailing edge of the left flap. 18-Nov-19 Cessna 172M G-EGLA Bodmin Airfield The aircraft collided with a fence during a runway excursion on landing, suffering minor damage. It is reported to have landed long and fast. 20-Nov-19 Grob G115 G-CIMI Leicester Aerodrome The aircraft landed long on Runway 15 and ran off the end before stopping. Record-only investigations reviewed November - December 2019 42 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 Record-only investigations reviewed November - December 2019 Record-only investigations reviewed November - December 2019 cont 02-Dec-19 Europa G-MFHI Rochester Airport During taxi after landing the nose gear collapsed. The propeller and nose landing gear were damaged. There were no injuries. 04-Dec-19 Sherwood Kub (SSDR) G-TLEE Egerton Airfield, Kent During the takeoff run the nosewheel dug into the grass runway and broke causing the aircraft to flip over. 07-Dec-19 Cessna 150F G-ATHV Warrington Grass Airfield During taxi the aircraft slid on wet grass and the port main wheel and nosewheel slipped into a drainage ditch resulting in damage to the port wingtip, port tailplane and the propeller. 31-Dec-19 Savannah Jabiru G-CDLR Widdrington, Northumberland Whilst carrying out a practice forced landing the engine failed. During the subsequent forced landing in a field the nosewheel dug into soft ground and the aircraft flipped inverted. 43 © Crown copyright 2020 All times are UTC Miscellaneous This section contains Addenda, Corrections and a list of the ten most recent Aircraft Accident (‘Formal’) Reports published by the AAIB. The complete reports can be downloaded from the AAIB website (www.aaib.gov.uk). AAIB Bulletin: 2/2020 45 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-AYXU EWG2019/09/14 BULLETIN CORRECTION Aircraft Type and Registration: Champion Citabria, G-AYXU Date & Time (UTC): 15 September 2019 Location: Gloucester Airport Information Source: Record only investigation AAIB Bulletin No 12/2019, page 92 refers The original entry was incorrect in that it stated ‘The nosewheel collapsed after landing’. The entry has since been corrected to read: On landing directional control was lost resulting in a ground loop, which collapsed the left hand undercarriage resulting in substantial damage to the wing and propeller. The aircraft has been written off. The online version of the report was corrected on 9 January 2020. 46 © Crown copyright 2020 All times are UTC BULLETIN CORRECTION Aircraft Type and Registration: DH82A Tiger Moth, G-ANDP Date & Time (UTC): 9 September 2019 Location: Ballymagreehan, Newtownards, County Down Information Source: Record only investigation AAIB Bulletin No 12/2019, page 92 refers The original entry was incorrect in that it stated that the aircraft landed back on the strip but struck a boundary fence. This should have read and struck a boundary fence. The entry has been amended to read: The aircraft failed to achieve climb performance after takeoff but landed back on the strip and struck a boundary fence. The online version of the report was corrected on 9 January 2020. AAIB Bulletin: 2/2020 G-ANDP EW/G2019/09/08 47 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-CHWJ EW/G2019/05/03 BULLETIN CORRECTION Aircraft Type and Registration: Guimbal Cabri G2 G-CHWJ Date & Time (UTC): 7 May 2019 Location: Cotswold Airport Information Source: Record only investigation AAIB Bulletin No 12/2019, page 89 refers The original entry was incorrect in that it stated that the incident occurred on 4 July 2019, when in fact it was the 7 May 2019. The online version of the report was corrected on 9 January 2020. 48 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BYTU EW/G2019/10/17 BULLETIN CORRECTION Aircraft Type and Registration: Mainair Blade 912, G-BYTU Date & Time (UTC): 18 October 2019 Location: Chirk Airfield, Wrexham Information Source: Record only investigation AAIB Bulletin No 12/2019, page 93 refers The report incorrectly states that the accident occurred at Clacton Airfield. The accident occurred at Chirk Airfield, Wrexham. The pilot has also reported that, during the deceleration on wet uneven ground, the aircraft yawed with increasing severity to the right which caused the right wing to strike the ground. The online version of the report was corrected on 9 January 2020. 49 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 SUPPLEMENT Registered Owner and Operator Canfield Hunter Ltd Aircraft Type Hawker Hunter T7 Nationality British Registration G-BXFI Place of accident A27, Shoreham Bypass, at the junction with Old Shoreham Road, North of Shoreham Airport Date and Time: 22 August 2015 at 1222 hrs (Times in this report are UTC1 unless stated otherwise) Summary In June 2019 the AAIB was asked to consider additional information related to the accident involving Hawker Hunter G-BXFI at Shoreham in 2015. As part of its review of this material the AAIB considered further aeromedical opinion and produced more detailed estimates of acceleration experienced by the pilot in the manoeuvres preceding the accident. The review concluded that the findings of the AAIB investigation published in Aircraft Accident Report 1/2017 remain valid. Introduction On 3 March 2017 the AAIB published the final report2 of its investigation of the accident involving Hawker Hunter G-BXFI near Shoreham Airport on 22 August 2015 (the AAIB investigation). In the period between June and October 2019 the AAIB was asked to consider additional information. This included witness statements, several analyses3 of the pilot’s actions and a video of a practice display at Duxford. The purpose of this review was to determine if these documents contained new and significant evidence of cognitive impairment. For the purpose of this review, the AAIB defined cognitive impairment as a physiological state in which an individual cannot think as well as usual, so is less able to do a task reliably and the probability of error is increased. The review considered whether the material was: ● evidence – containing facts relevant to the accident, distinct from analysis of those facts, Footnote 1 Co-ordinated Universal Time. 2 Aircraft Accident Report 1/2017: report on the accident to Hawker Hunter T7, G-BXFI near Shoreham Airport on 22 August 2015 available at https://www.gov.uk/aaib-reports/aircraft-accident-report-aar-1-2017-g-bxfi- 22-august-2015 [accessed December 2019] 3 These analyses were prepared for a purpose other than safety investigation and do not necessarily represent the complete opinion of their authors. 50 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 ● new – not previously considered by the AAIB, and ● significant – having a different effect from evidence that the AAIB had already considered. If all three criteria were met, the AAIB was required to reopen its investigation.4 The review involved Inspectors who were not part of the AAIB investigation described in Aircraft Accident Report 1/2017. The review team included an Inspector (Human Factors), an Inspector (Recorded Data) with expertise in aircraft performance, and two Inspectors (Operations) who were formerly fast jet pilots with experience in instruction and display flying. As part of the process of considering the significance of the material presented, the review applied additional modelling techniques to determine the aircraft’s flight path, and to provide more detailed estimates of +Gz (the “head to foot” acceleration experienced by the pilot, normal to the flightpath), in the manoeuvres preceding the accident. These are described in Appendix 1. It also considered additional aeromedical opinion. Summary of Gz analysis The additional modelling indicated that +Gz during the positioning turn was briefly about 3.8 g, four seconds after the start of the positioning turn, falling within three seconds to approximately 2.2 g, then rising slightly to a level predominantly around 3 g before falling again to around 1 g over the final six seconds of the turn. The review also estimated +Gz exposure in the first part of the loop. This indicated that a maximum +Gz of 4 g occurred about five seconds after the start of the manoeuvre and remained above 3 g for about four seconds. The +Gz load then reduced in a linear manner to a value of approximately 1.6 g some four seconds later. Conclusions of the AAIB investigation regarding pilot impairment The issue of possible cognitive impairment due to +Gz exposure was considered in the AAIB investigation and was discussed in the AAIB final report5. The pilot’s behaviour as captured by a cockpit action camera6 was assessed by pilot expert advisors to the investigation, human factors experts and an aeromedical expert. The AAIB report stated: ‘As far as could be determined from cockpit image recordings the pilot appeared alert and active throughout the flight.’ Footnote ⁴ Regulation 18(1) of the Civil Aviation (Investigation of Air Accidents and Incidents) Regulations 2018. ⁵ See sections 1.18.10.1 page 124 and 2.2.3 page 166 of AAIB Aircraft Accident Report 1/2017. ⁶ This camera captured a partial view of the pilot from behind, a portion of the instrument panel and a portion of the view through the canopy and windscreen. See section 1.11.3 page 40 of AAIB Aircraft Accident Report 1/2017. 51 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 The aeromedical expert who advised the AAIB investigation stated: ‘Although some day to day variation in G tolerance occurs in every individual, it would be very unusual for an individual to suffer G related impairment at less than +3Gz while wearing a G suit. The video evidence reviewed herein shows no evidence of the classically described G-LOC or A-LOC syndromes, and the G levels in the accident video are similar to previously experienced levels flown without incident (eg in the Duxford video). 7 Therefore, I can find no evidence of G related impairment in the material available for review.’ The pilot had G currency8 having flown nine displays in the two weeks prior to the accident. Although not documented in detail in the AAIB final report, other factors that can affect G tolerance9 were considered to the extent possible using information from the accident flight and previous flights conducted by this pilot, medical information and pilot interviews. There was no evidence that these factors differed significantly from previous occasions. The AAIB report provided the following analysis of the factual information it presented: ‘There was no evidence of any g-related impairment of the pilot during the aerobatic sequence flown. If the pilot was unwell before the accident, it was not established in what way he was unwell or when the onset of any condition was first experienced.’ On this topic, the report concluded that: ‘The g experienced by the pilot during the manoeuvre was probably not a factor in the accident.’ Analysis of pilot actions - AAIB investigation The pilot’s actions were considered during the AAIB investigation. Evidence was limited because the pilot did not recall the accident and his plan for the display was not documented in detail. It was not possible to draw firm conclusions about what influenced the pilot’s performance on the day. The AAIB’s analysis focused on two decision points where the pilot may have been able to recover from any deviations in the planned manoeuvres that had occurred and prevent the situation from progressing into an accident: 1. The entry to the accident manoeuvre 2. The apex of the accident manoeuvre Footnote ⁷ Cockpit image recording of a display at Duxford. ⁸ Recent experience of flying with +Gz exposure. G currency increases G tolerance. ⁹ Factors reducing G tolerance include hypoxia, hyperventilation, infection, drugs, alcohol, heat stress, dehydration, fatigue, hunger and poor physical fitness among others. Factors increasing G tolerance include G currency, use of anti-G clothing and anti-G straining manoeuvre among others. 52 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 The AAIB determined that these were the areas where the greatest safety learning could be obtained to improve display pilots’ ability to recover when manoeuvres do not progress as planned. The Royal Air Force Centre for Aviation Medicine (RAFCAM) conducted a human factors analysis10 of these decision points. It used a range of evidence sources such as the cockpit action camera footage, results from flight trials and notes from interviews with the pilot. The RAFCAM applied recognised systematic human factors analysis techniques.11 The RAFCAM analysis identified the credible errors and performance shaping factors12 that could have been present at the entry to the loop manoeuvre, during the climb and at the apex of the manoeuvre. Additional analyses of the pilot’s actions The material presented to the AAIB in 2019 included analyses by six different authors of the pilot’s actions preceding the accident. Among them the authors had experience and qualifications in display flying, aviation medicine and human factors. None of these documents were ‘evidence’. They were all opinions regarding the pilot’s actions based on the authors’ interpretations of facts that were already known to the AAIB. These analyses of the pilot’s actions were reviewed to determine if they contained or referenced new and significant evidence, or if they offered new insights. Figure 1 summarises how the analyses considered that the pilot’s performance diverged from what was required. Several of the authors asserted that such a pattern of behaviour by this pilot could only be explained by some form of cognitive impairment. One of the authors argued that the only source of such impairment possible in this accident was +Gz. Referenced research papers One of the documents referred to several published papers on the topic of human response to +Gz. These papers were evidence that was potentially relevant and had not been considered previously by the AAIB (Appendix 2 provides summaries). Therefore, they were considered as part of this review to assess whether they were significant. Only those papers referenced within the documents presented to the AAIB have been considered specifically. Footnote 10 Royal Airforce Centre for Aviation Medicine. Aircraft Accident Human Factors Report. Hawker Hunter G-BXFI. Shoreham Airshow, 22 August 2014. Referred to in Appendix M, pages 404 – 420 of the AAIB Aircraft Accident Report 1/2017. 11 Task analysis and two methods of human error analysis, Systematic Human Error Reduction and Prediction Approach (SHERPA) and an adapted version of the Australian Transportation Safety Board Human Factors analysis of see and avoid. 12 Performance shaping factors are characteristics of an individual, group, task, environment or organisation that influence human performance. Some performance shaping factors enhance performance, such as motivation, training or practice and some reduce human performance such as high workload, unclear procedures or badly designed equipment. 53 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 Figure 1 Summary of pilot actions according to material provided to the AAIB There are other published academic papers that also cover this topic. It was outside the scope of this review to search the wider literature, but additional aeromedical opinions were invited so that current scientific consensus could be considered. General aeromedical information regarding +Gz tolerance As part of this review, the AAIB requested further aeromedical opinion from two external experts13, who considered the more detailed Gz profile and cockpit action camera footage. Both experts were qualified aviation medical practitioners with a specialism in G-related physiology and the effects on human performance of G exposure. The AAIB also consulted standard texts14 regarding the effects of G on human performance. The following definitions are relevant: ● Greyout is the partial loss of vision. Sometimes referred to as partial light loss. ● Blackout is the complete loss of vision with preserved consciousness. Sometimes referred to as light loss. Footnote 13 The aeromedical experts were from different organisations, one in the UK and one in another State. 14 Gradwell, D.P. and Rainford, D.J. (2016). Ernsting’s Aviation and Space Medicine. 5th edition. CRC press. Green, N., Gaydos, S., Hutchison, E. and Nicol, E. (2019). Handbook of Aviation and Space Medicine. CRC press Reinhart, R.O. (2008). Basic Flight Physiology. 3rd edition. McGraw-Hill. 54 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 ● A-LOC is ‘almost loss of consciousness’. It is G-related incapacitation without overt loss of consciousness. There is inconsistency in the literature about the performance effects, but symptoms reported following A-LOC include sensory abnormalities, amnesia, confusion and a disconnection between cognition and the ability to act. ● G-LOC is ‘G induced loss of consciousness’. It features a period of absolute incapacitation followed by a period of impaired consciousness while the crew member is recovering. ● G tolerance is the level at which individuals experience greyout or G-LOC. It varies between individuals and there is variation within individuals on different days. The figures stated for average levels of Gz tolerance varied slightly in the literature depending on what research source was used. Figure 2 summarises results from a compilation of Gz tolerance studies involving participants who were not wearing any anti-G clothing or performing any techniques to increase their G tolerance. Each point on the graph shows the result of a research study in terms of when the end point of greyout, blackout or unconsciousness was experienced under different levels of Gz and different durations of exposure. The line on the graph summarises the general finding seen across studies that high levels of Gz of short duration may not result in symptoms, but symptoms progress quickly if the onset of Gz is rapid. Lower levels can be tolerated for longer, especially if the onset is slow. It shows that it is rare to find effects within four seconds of exposure or below +3 g. A typical G-suit could provide up to 1.5 g additional tolerance if working correctly. Wearing anti-G clothing even if not functional is said to add approximately 0.4 g additional tolerance.15 Greyout, blackout, A-LOC and G-LOC are the only performance effects of +Gz that are widely acknowledged by the aeromedical community. None of the standard texts mentioned a possibility of cognitive impairment at low levels of +Gz exposure insufficient to induce A-LOC or G-LOC. The two aeromedical experts were specifically asked about performance effects of low +Gz exposure. They provided opinion supported by published literature, a bibliography of which is included at the end of this supplement. One expert stated: ‘Some limited research has been conducted looking at cognitive impairment under +Gz loads that do not result in A-LOC or G-LOC. The findings of these studies are somewhat inconsistent, and in many cases contradictory. As such, it is not possible to make definitive conclusions from these limited experimental studies, which by and large have no practical implications for the flying task under +Gz.’ Footnote 15 Parkhurst, MJ, Leverett SD Jr, Shubrooks JR. Human Tolerance to High, Sustained +Gz Acceleration. Aerospace Med. 1972; 43(7):708712. 55 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 The other expert cited three studies that found no change in performance under low levels of +Gz and two studies that found some minor changes in performance during or after +Gz exposure. In conclusion this expert stated: ‘Based on the limited evidence that does exist, and international medical and flying experience, the probability that pilots can become cognitively impaired during exposure to low levels of +Gz acceleration is extremely low.’ Figure 2 Relaxed participants’ tolerance to +Gz acceleration based on a compilation of research studies (reproduced from Gradwell, D.P. and Rainford, D.J., 2016) The AAIB reviewed the full original text of all the publications cited by the aeromedical experts. It also asked the opinion of the RAFCAM, which advised the AAIB that: ‘the existence of low Gz induced cognitive impairment is not supported by decades of flight experience or flying training under high G loads requiring completion of complex cognitive tasks.’ The RAFCAM consulted the relevant NATO panel, air forces in the European Air Group, the Five Eyes Air Force Interoperability Council, and the United States Air Force. None of these authorities recognised the existence of low Gz-induced cognitive impairment, nor do they train their respective military pilots to avoid this condition. 56 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 Re-assessment of video footage During this review, videos of the accident flight and previous displays and practises by the pilot at Shoreham in 2014, Duxford 2014, Bray 2015, Shuttleworth 2015 and Eastbourne in 2014 and 2015 were re-examined by two AAIB Inspectors of Air Accidents (Operations). Their experience included flying, displaying and instructing in various aircraft types including the Jet Provost, Hawk, Buccaneer, Hunter, Jaguar and T33 Shooting Star. They were not part of the AAIB investigation team. These Inspectors, familiar with observing students while sat behind them in tandem cockpits, concluded that the pilot’s head and body movements were consistent with what they would expect from someone flying a loop manoeuvre. They did not observe any significant differences in behaviour between the accident flight and previous displays. They could not identify a point at which the pilot’s behaviour changed in an observable way that would indicate impairment. One of the aeromedical experts reviewed the cockpit action camera footage of the accident flight and the Shoreham display in 2014 for evidence of +Gz induced impairment. He stated: ‘the accident clip does not demonstrate any of the typical head movements associated with impairment due to high +Gz.’ Figure 3 and Figure 4 illustrate one of the behaviours he described as an example of optokinetic cervical reflex.16,17,18,19 This is a well-documented phenomenon in low-level flight in visual meteorological conditions, in which a pilot orients their head with respect to the visible horizon rather than the aircraft’s attitude. This aeromedical expert noted that: ‘the presence of these typical flight-related head movements is entirely consistent with normal, routine flight operations.’ He concluded that: ‘there was no discernible significant difference between the head movements of the pilot in either of the two flights. What movements were seen were entirely consistent with a pilot attempting to maintain an adequate lookout during low-level aerobatic manoeuvring.’ Footnote 16 Beer J, Freeman, D. Flight display dynamics and compensatory head movements in pilots. Aviat Space Environ Med 2007; 78(6):579-87. 17 Gallimore J BN. Effects of FOV and Aircraft Bank on Pilot Head Movement and Reversal Errors During Simulated Flight. Aviat Space Environ Med 1999; 70:1152-60. 18 Gallimore J P, F, Brannon, N, Nalepka, J. The Opto-Kinetic Cervical Reflex During Formation Flight. Aviat Space Environ Med 2000; 71(8):812-21. 19 Merryman R, Cacioppo, A. The Optokinetic Cervical Reflex in Pilots of High-Performance Aircraft. Aviat Space Environ Med 1997; 68(6):479-87. 57 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 Figure 3 Shoreham 2014: Head movement example of optokinetic cervical reflex Figure 4 Accident flight: Head movement example of optokinetic cervical reflex Analysis The AAIB investigation considered the possibility of G-related pilot impairment. The final report included an approximate calculation of +Gz in the turn preceding the accident manoeuvre, analysis of the pilot behaviour from the cockpit action camera and expert aeromedical opinion. Given the relatively low levels of +Gz and the absence of any signs of impairment, the investigation concluded that ‘The g experienced by the pilot during the manoeuvre was probably not a factor in the accident’20, though it could not be ruled out. Footnote 20 Finding 31, page 197 of AAIB Aircraft Accident Report 1/2017. 58 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 The issue was raised again by additional information provided to the AAIB, which included analyses by other authors that differed from analysis presented in the AAIB final report. None of the information was assessed to be new and significant evidence but the review provided an opportunity to consider the possibility of +Gz impairment in more depth in case further safety learning could be obtained. Exposure to Gz As part of the review, the cockpit video recordings and other evidence were analysed further21 to model +Gz exposure during the positioning turn. This provided additional detail for the manoeuvres preceding the accident. Calculated +Gz for the positioning turn was briefly about 3.8 g, four seconds after the start of the positioning turn, reducing within 3 seconds to approximately 2.2 g, then rising slightly to a level predominantly around 3 g before falling again to around 1 g over the final six seconds of the turn. The review also estimated +Gz exposure in the first part of the loop. This indicated that a maximum +Gz of 4 g occurred about 5 seconds after the start of the manoeuvre. The G-onset rate peaked at 0.8 g/s with +Gz remaining above 3 g for about 4 seconds. The +Gz load then reduced in a linear manner to a value of approximately 1.6 g some four seconds later. These calculations show how +Gz developed over time. When compared to aeromedical information about human G tolerance it shows that the forces experienced were outside the range usually considered a hazard to human performance. The literature cited in the additional material provided to the AAIB (summarised in Appendix 2) was consistent with this aeromedical consensus. It generally related to +Gz levels greater than those experienced in the accident or was based solely on pilot recollections without empirical measurement. The forces experienced on the accident flight were unlikely to have affected the pilot considering his aerobatic currency and the protection of the G-suit. The more detailed analysis of the flight path and Gz in this review adds weight to the AAIB investigation’s finding that G-related impairment was probably not a factor. Cognitive impairment by +Gz The cockpit action camera footage showed that the pilot was active throughout the flight. He appeared to be controlling the aircraft and using a variety of cues as would be expected for the manoeuvres flown. Some of the documents submitted to the AAIB asserted it is possible the pilot suffered cognitive impairment so subtle as to be not observable in the video footage or conduct of the task. Subtle cognitive impairment by +Gz has not been considered an issue within aviation even though G-related visual symptoms were first recorded in 192022. It is not recognised by the aeromedical community in general Footnote 21 See Appendix 1. 22 Head, H. (1920). The Sense and Stability of Balance in the Air. In Medical Research Council Report into the Medical Problems of Flying. London HM Stationary Office 1920: 214 - 56 cited in Gradwell, D.P. and Rainford, D.J. (2016). Ernsting’s Aviation and Space Medicine. 5th edition. CRC press. 59 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 and major military authorities around the world do not consider it to be an issue. Both aeromedical experts consulted during this review considered it unlikely, basing their view on the balance of published evidence of which they were aware. Where performance effects were documented in these published studies, their direction, nature and magnitude were variable, and some studies found no change in the performance variables they measured. There were not enough studies of low +Gz exposure overall to draw conclusions about any particular performance effect. It was outside the scope of this review to perform a comprehensive search of the literature and there may be additional relevant material that has not been considered. The UK Civil Aviation Authority (CAA) stated that it is conducting a medical review of the potential risk of cognitive impairment in civilian pilots due to G forces. It will not re-examine the specific circumstances of the 2015 Shoreham accident. Following this review and if appropriate, the CAA will consider taking regulatory action to improve flight safety. Performance shaping factors Some of the analyses provided to the AAIB offered an account of where in the accident flight the authors felt the pilot’s actions differed from what was intended or appropriate. Not all these actions were discussed in the AAIB final report due to limited evidence about the pilot’s intentions and a focus on areas where safety could be improved. Some of the authors of the additional material considered in this review asserted that, looking at the pilot’s overall pattern of behaviour, impairment was the only explanation. However, there are alternative explanations that do not involve cognitive impairment. It is possible that an individual doing any task will do it incorrectly. For an easy task or a highly trained person doing a task they are very familiar with, the likelihood of this is low. For a difficult task, or a task performed by someone not sufficiently familiar with it, the likelihood is high. The probability that someone will do a task incorrectly is influenced by performance shaping factors. Impairment was one possible performance shaping factor in the G-BXFI accident. The AAIB investigation found no evidence of impairment but, if present, it did not affect the pilot’s observable behaviour and the source of any impairment was unknown. Other performance shaping factors were more likely than impairment to have contributed to this accident. During the AAIB investigation, the RAFCAM conducted a human factors analysis.23 It identified credible performance shaping factors, including the possibility of glare affecting the pilot’s ability to read critical information from the instruments, and the possibility that the pilot did not correctly recall the required speeds and heights due to the differences between the various aircraft he regularly flew. Footnote 23 Royal Airforce Centre for Aviation Medicine. Aircraft Accident Human Factors Report. Hawker Hunter G-BXFI. Shoreham Airshow, 22 August 2014. Referred to in Appendix M, pages 404 – 420 of the AAIB Aircraft Accident Report 1/2017. 60 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 Pilot’s overall pattern of behaviour The overall pattern of behaviour described by the analyses provided to the AAIB can be explained in other ways that do not require impairment as a common factor. The pitch oscillations, the pull-up technique and the decision not to eject were proposed as significant in some of the analyses but the AAIB did not consider them to be significant actions in the sequence. The pitch oscillations were assessed to be minor deviations probably due to the aircraft being slightly out of trim in pitch. The pull-up technique was consistent with the pilot’s technique seen in other display videos. A decision to eject would not be taken in preference to flying an escape manoeuvre if the pilot realised this was required. By the time the situation was unrecoverable, the aircraft was outside the limits from which an ejection would be likely to be survivable. The reduction in thrust24 prior to the start of the loop, if commanded by the pilot25, may have been an appropriate action if he intended to commence the loop at less than 350 kt, or if he mis-recalled the target speed. He may have considered the speed to be correct. Any analysis of the pilot’s position and pull-up point relies on assumptions about his plan. There was no evidence documenting the exact track he planned to fly and therefore none with which to compare the actual flight path. The track at the end of the positioning turn was only marginally different to what the pilot described when interviewed. Had he rolled out of the positioning turn earlier he would have flown over Lancing College; which was prohibited. On the actual final heading achieved, a later pull-up would have risked infringing the display line. If the pilot had deviated from his planned positioning, he may have become preoccupied with this, causing him to pay less attention to flying the manoeuvre accurately. This could account for the early roll and perhaps the insufficient thrust. For the pilot to decide to fly an escape manoeuvre he would need to: realise he was too low, believe that the escape manoeuvre was likely to be successful, and be capable of flying it. The RAFCAM human factors analysis considered the apex of the loop in depth.26 It described several credible mechanisms by which the pilot could be unaware he was too low. For example, the AAIB investigation highlighted a design issue that would have increased the likelihood of misreading the altimeter due to the obscuration of the digit showing thousands of feet.27 The investigation found that the pilot may not have known an escape manoeuvre could be successful from the height and speed achieved at the apex of the final manoeuvre. If the pilot realised he was too low, the time available may Footnote 24 Analysis of cockpit audio recordings from the accident flight enabled thrust levels to be assessed, but throttle movements were not visible in the cockpit image recordings. Section 1.11.3 on page 41 of Aircraft Accident Report 1/2017. 25 An uncommanded reduction in thrust during the accident manoeuvre could not be ruled out. Finding 13 on page 196 of Aircraft Accident Report 1/2017 26 See paragraphs 19 – 21, pages 415 - 418 of AAIB Aircraft Accident Report 1/2017. 27 See Figure 28, page 129 of AAIB Aircraft Accident Report 1/2017. 61 © Crown copyright 2020 All times are UTC AAIB Bulletin: 2/2020 G-BXFI EW/C2015/08/04 only have been sufficient to execute a rule-based decision28 of a practiced action. By the time it would have been obvious to the pilot he could not complete the loop safely it is unlikely it would have been possible to avoid the road. Hence, the actions may have been linked together and do not require cognitive impairment as an explanation. Whatev