Description

On 31 July 2015, an Embraer EMB-505 Phenom 300 (HZ-IBN) being operated by a single pilot and on a private flight from Milan Malpensa to Blackbushe overran the landing runway at destination after a long and fast touchdown following a day VMC visual approach. After hitting a small embankment just beyond the end of the runway, it collided with unoccupied parked cars and was destroyed by a combination of impact damages and an intense fire in which the pilot and three passengers were killed.

Investigation

A Field Investigation was carried out by the UK AAIB. Recorded data was successfully recovered from the aircraft CVFDR and from Air Traffic radar and voice communications, CCTV and video. CVFDR data ended as the aircraft overran the end of the runway - it was equipped with an impact switch designed to stop it in the event of an accident. The air traffic recordings included TCAS downlink data for an incident on approach from the accident aircraft and altitude information for a microlight during which the accident pilot observed and avoided the microlight in the Blackbushe circuit. This data was separately available to a resolution of 25 feet. This information, in conjunction with an examination of the accident site and the runway, enabled a comprehensive reconstruction of the final stages of the accident flight.

It was found that the 57 year-old pilot had 11,000 total flying hours including 1,181 on the accident aircraft type. He had been employed by the aircraft operator as a Phenom 300 pilot since 2011. Prior to this employment, he was found to have worked as a flying instructor and as a Cessna 560XLS pilot and, for one year only, as a pilot on an airline-operated Airbus A320. On the morning of the accident, he had travelled from his hotel to Blackbushe in order to position the aircraft to Milan, collect the three passengers and convey them to Blackbushe. It was noted that according to his family, he had been "in good health both mentally and physically" before leaving the hotel that morning.

The Saudi Arabian-based company which employed the pilot operated the accident aircraft privately for its owner and had a second Embraer Phenom 300 in its fleet. Between them, these two aircraft were found to have visited Blackbushe on 35 occasions in the year prior to the accident with the accident pilot flying on almost half of these visits. However, it was noted that "most of the destinations to which the Phenom 300 aircraft flew were large international airports in Europe and the Middle East" and "analysis of FDR data for both Phenoms showed that visual circuit arrivals were rare and seldom occurred other than at Blackbushe", an aerodrome which was outside controlled airspace, without any instrument approach procedures and with air traffic service provided by Flight Information Service (FIS) rather than ATC. It was further noted that although the operator was not subject to any regulatory requirement to operate under the same level of documentation as a commercial operator, it had "several features of a commercial transport undertaking, including an Operations Manual and Standard Operating Procedures".

A reconstruction of the final stages of the accident flight established that in benign weather conditions, the aircraft had joined the visual circuit for runway 25 at Blackbushe on the crosswind leg before entering the downwind leg outside a much slower microlight aircraft expecting to extend downwind to allow it to land first. Subsequently, after disconnecting the AP and with the awareness of the other pilot and the Blackbushe AFISO, the microlight had been overtaken by climbing slightly to pass ahead of and above it. However, at approximately 1,000 feet aal as this climb began, a TCAS 'DESCEND' RA had been triggered by proximity to that aircraft. This was ignored and separation was maintained visually by continuing the climb. The RA then changed to 'Maintain' and then 'Adjust' Vertical Speed, possibly because of another light aircraft which was tracking southwards to the east of Blackbushe. The relative positions of the three aircraft at the time of these two TCAS RAs are depicted on the annotated diagram of aircraft tracks below. Following this climb over the microlight, the accident aircraft then flew a curved base leg onto final whilst descending at up to 3,000 fpm. As it passed 1,200 feet aal and the TCAS 'Clear of Conflict' was annunciated, the aircraft was 1.1 nm from the landing runway threshold at 146 KIAS with the landing gear down and flap 3 selected.

The tracks of the accident aircraft and the two proximate aircraft annotated with their relative positions when the accident aircraft's second and third TCAS RAs were triggered [Reproduced from the Official Report]

From there it continued its approach to 500 feet aal at an average of around 150 KIAS and an average of around 3,000 fpm. Passing 1,125 feet aal, the first of six TAWS 'PULL UP' Warnings were annunciated. The speed brakes were selected out at 675 feet aal but this had no effect since their deployment is inhibited when the flaps are extended. Passing 500 feet aal the descent rate was still 2,500 fpm and although the aircraft crossed the Runway 25 threshold at approximately 50 ft aal, this occurred at 151 KIAS compared to the calculated applicable target threshold speed of 108 KIAS. The aircraft floated and finally touched down 710 metres beyond the displaced runway threshold at 134 KIAS from which position, it was estimated that the landing ground roll required to stop the aircraft would have been at least 616 metres when only 349 metres of the promulgated landing distance remained.

Deployment of the ground spoilers and application of the brakes occurred almost immediately. When the aircraft reached the end of the runway paved surface, its groundspeed was 83 knots. In the second before the end of the CVFDR data, "both thrust levers were recorded as having been advanced from the idle position and the spoilers had started to close".

After departing the paved surface at the end of the runway slightly to the left of the extended centreline, the aircraft collided with a low earth embankment which led to the loss of the nose landing gear and doors and the aircraft briefly becoming airborne again before colliding with a group of unoccupied parked cars in a right wing low attitude before finally stopping about 70 metres beyond the earth embankment. CCTV evidence of the impact sequence showed that the separation of the entire left wing from the fuselage had led to it rolling through almost 350°before coming to rest on top of the detached wing, still largely intact. The first signs of fire could be seen almost immediately and the fire quickly became intense, consuming most of the aircraft.

The AFISO had initiated a full emergency as the aircraft touched down, because he felt that “it was clear at this time that the aircraft was not going to stop”. However, the arrival of the AFS at the crash site was delayed by access difficulties and by the time they arrived the post crash fire was intense and any attempt at life-saving intervention was impossible. No evidence was found during the post mortems carried out on the four occupants of any injuries not related to the fire that could have caused or contributed to their deaths.

The Investigation noted that the unstabilised approach had followed a change of plan in respect of the order of approach relative to the microlight aircraft which was already in the circuit when the accident aircraft arrived crosswind. There were clear indications of the initial situational awareness of the accident aircraft's pilot who had initially envisaged extending the downwind leg of his circuit to allow the microlight to land to prevent a conflict and satisfy the relevant rules. By joining the circuit overhead the runway, he had "maximised the time available to configure the aircraft, reduce speed and height, and manoeuvre towards the final approach within the airspace constraints". It was only after the microlight pilot's proposal to extend downwind and let the jet in first in response to an AFISO transmission saying “I’m not sure how this is going to work", that the accident aircraft pilot accepted the offer to position number 1 "after the contrary plan had been proposed and agreed".

It was considered that "if an air traffic control service had been in place at Blackbushe, a suitably-qualified controller could have issued appropriate instructions to resolve the potential conflict, instead of relying on the suggestion of one of the pilots involved". Thereafter, as the consequences of this change of plan unfolded, the workload of the accident aircraft pilot rose which would have required him to "make rapid changes to his mental model". Multiple aural inputs and the need to make avoidance manoeuvres created a high-workload situation "in the course of what was already a more than usually demanding approach". It was considered possible that in these circumstances the pilot had "fixated on his initial strategy (landing) and lacked the mental capacity to recognise that the approach had become unstable and should be discontinued". It was also considered that the absence of a response to the AFISO transmission informing the pilot that he could land was "consistent with the pilot’s workload having reached a point at which he could no longer process his aural environment or perceive critical information".

Other aspects of the accident circumstances identified included:

• The fact that the continuation of the approach to an attempt at landing had breached the Operator's stabilised approach policy which stated that "if a stabilised approach cannot be achieved before descending below the minimum stabilised approach heights, immediate action will be taken to execute a missed approach or go-around".
• In respect of the earth bank at the end of the paved surface, the accident runway was not long enough for the regulatory requirements for the presence of a RESA to apply, so its presence was permissible.
• Although since the accident flight had been private and there was therefore no regulatory requirement for it to use a licensed aerodrome or for rescue and fire fighting cover to be available, Blackbushe aerodrome was licensed and provided Category 2 RFFS cover which would have been sufficient for a commercial transport flight by an aircraft such as the one involved.
• The delay to AFS arrival at the accident site despite the proactive alert given was attributable to a locked gate which was not mentioned on related briefing. This omission was contrary to national Licensing of Aerodromes provisions which include a requirement to fully detail the procedures for an emergency response within 1000 metres of each runway threshold.

The formally documented Conclusion of the Investigation was as follows:

"The pilot was appropriately licensed and experienced, and had operated into Blackbushe Aerodrome on 15 previous occasions. He was reported to be physically and mentally well. The aircraft was certified for single-pilot operations and the pilot was qualified to conduct them.

The engineering investigation of the accident aircraft did not find evidence of any preexisting technical defect that caused or contributed to the accident. The meteorological conditions were suitable for the approach and landing and, at the actual landing weight and appropriate speed, a successful landing at Blackbushe was possible.

HZ-IBN joined the circuit at a speed and height which would have been consistent with the pilot’s stated plan to extend downwind in order that the microlight could land first. The subsequent positioning of HZ-IBN and the microlight involved HZ-IBN manoeuvring across the microlight’s path, in the course of which the first of several TCAS warnings was generated.

After manoeuvring to cross the microlight’s path, HZ-IBN arrived on the final approach significantly above the normal profile but appropriately configured for landing. In the ensuing steep descent, the pilot selected the speedbrakes out but they remained stowed because they are inhibited when the flaps are deployed. The aircraft’s speed increased and it crossed the threshold at the appropriate height, but 43 KIAS above the applicable target threshold speed. The excessive speed contributed to a touchdown 710 metres beyond the threshold, with only 438 metres of paved surface remaining. From touchdown, at 134 KIAS, it was no longer possible for the aircraft to stop within the remaining runway length.

The brakes were applied almost immediately after touchdown and the aircraft’s subsequent deceleration slightly exceeded the value used in the aircraft manufacturer’s landing performance model. The aircraft departed the paved surface at the end of Runway 25 at a groundspeed of 83 knots.

The aircraft collided with an earth bank and cars in a car park beyond it, causing the wing to separate and a fire to start. Although the aircraft occupants survived these impacts, they died from the effects of fire.

Towards the end of the flight, a number of factors came together to create a very high workload situation for the pilot, to the extent that his mental capacity could have become saturated. His ability to take on new and critical information, and adapt his situational awareness, would have been impeded. In conjunction with audio overload and the mental stressors this can invoke, this may have lead him to become fixated on continuing the approach towards a short runway."

Safety Action taken by the Aircraft Operator consequent upon the accident and noted by the Investigation included the following:

• the Phenom 300 is now operated with two pilots, although a risk analysis process may be developed which can assess the risk of each flight in order to determine if a single-pilot operation can be justified for the purposes of operational flexibility such as on short flights in relatively low-workload environments.
• The Stabilised Approach 'gate' for VMC approaches has been raised from 200 feet agl to 500 feet agl.
• Pilot recurrent training requirements have been enhanced to better address both short runway operations and time-critical decision making.
• Plans have been made to implement OFDM.

A Special Bulletin was published on 6 August 2015 to report the initial factual findings of the Investigation and the Final Report of the Investigation was published on 8 December 2016. No Safety Recommendations were made.

Related Articles

• Stabilised Approach
• Energy Management during Approach
• Landing Distances
• Rejected Landings
• Terrain Avoidance and Warning System (TAWS)
• Pilot Workload
• Decision Making
• Continuation Bias
• Airborne Collision Avoidance System (ACAS)
• Flight Information Service
• Flight Data Monitoring (FDM)