Description

On 31 December 2022, an Airbus A320 (OE-INE) being operated by easyJet Europe on a scheduled international passenger flight from London Gatwick to Bordeaux as EJU49QH was on an ILS approach to runway 23 at destination in day VMC when it received an instruction unexpected by the crew at a recorded 232 feet agl to go around and complied. The go around instruction was only issued after the pilot of a light aircraft who had been instructed to line up and wait on the same runway and was aware of the landing clearance given but unable to see how far out the aircraft on approach was had reminded the controller of their presence on the runway which the controller had overlooked.

Investigation

An Investigation was carried out by the French Civil Aviation Accident Investigation Agency, the BEA, assisted by QAR data from the A320, ATC recorded communication data and a series of pilot and controller statements. It was noted with concern that the Tower Supervisor “did not immediately notify the specified addressees, which included the BEA”.

It was noted that both the A320 pilots were based at Bordeaux and that the Italian Captain had approximately 9,000 hours flying experience and his French First Officer had approximately 2,100 hours flying experience. The DR400 pilot, who had approximately 210 hours flying experience, was using an aircraft from the locally based flying club to make a pleasure flight with his nine year-old son as a passenger. The two controllers on duty for a full Saturday morning shift were a 46 year-old TWR Supervisor who had held a controller licence for over 20 years and 53 year-old TWR controller who had held a controller licence for almost 30 years supported by one TWR Assistant. The in-position controller was responsible for the combined APP, TWR, GND and INFO positions which meant that communications to traffic on any of these frequencies could be heard on all the others as well.

What happened

The inbound A320 was cleared to descend to 3,000 feet QNH and then, four minutes later, as it descended through FL104, was cleared for the ILS approach to the runway in use, 23. At this time, the TWR assistant was taking a break and his place was taken by the TWR Supervisor. A third controller was present and on duty and had been in position earlier but was on a break. Communications with the A320 took place in English whereas all other communications were in French.

Almost five minutes after the A320 had been cleared for the ILS, the DR400 pilot requested taxi clearance from his parking position and was cleared to taxi to holding point Alpha for runway 23. For almost the whole of the next minute, the controller was kept busy replying to a VFR flight and in the absence of any instructions from the controller, the DR400 pilot gave way to an Air France A321 which was also taxiing to runway 23 and on a potentially conflicting path with him.

Six minutes after being cleared for the ILS approach, the inbound A320 was established on ILS 23 at 13 nm at an altitude of approximately 4,300 ft and at a speed of 223 KCAS. The Air France A321 then departed and a few seconds later, the DR400 pilot called to say he was approaching the holding point to which he had been cleared and would hold there but the controller responded with a clearance to line up and wait and asked him to change to the TWR frequency. The controller did not then transfer the strip for the DR400 to the runway area of the strip board.

After a further 30 seconds, with the A320 8 nm from the runway threshold at approximately 2,800 feet QNH and a speed of 170 KCAS, the controller told its crew to continue the approach. Thereafter the controller “was in constant communication with the pilots of several aeroplanes”: 

• the pilot of a VFR flight who had filed a flight plan for Courchevel, and who asked for his flight plan to be activated in the air, a request which was followed by exchanges regarding the desired flight level;
• the crew of the just departed A321 who continued their climb to FL 140;
• two VFR flights in transit through the Bordeaux CTR, one of which required a message to be repeated;
• various other VFR flights requesting traffic information.

Almost three minutes later a VFR flight called but the controller did not reply and instead cleared the inbound A320 to land (see position 5 on the illustration below). At this time it was 1.7 nm from the runway threshold and passing 766 feet QNH. Immediately after the A320 clearance readback, the controller cleared the recently departed A321 crew to an en-route waypoint. The DR400 pilot, aware that he had been lined up on the runway for more than three minutes and that the A320 had been cleared to land on the same runway he was lined up on had no idea how far away the A320 was and had no ability to look behind. He was also aware that the controller was busy and as he considered "that he was not a priority as he was on the ground, he decided to wait before contacting him”.

However, when nothing else was said about his takeoff and aware of the need for only a two minute gap between the departure of the A321 and the departure of his light aircraft which had easily been exceeded, he called “La tour du Fox Yankee?” and received the response “Fox Yankee?”. He then stated (in French) that “he was lined up on runway 23 and that he thought that the two minutes had expired”. The controller responded almost immediately by instructing the A320, which was just 1,000 metres from the runway threshold at 232 feet agl, to “go around immediately aircraft on the runway” (see position 6).

The missed approach was initiated by the crew at 133 feet agl and 525 metres before the runway threshold but only having reached a minimum of 103 feet agl at around 260 metres from the threshold (see position 7) did a climb commence. Eighteen seconds later, the A320 flew over the DR400 at a height of 178 feet (see position 8). The DR400 pilot saw the A320 fly low overhead his aircraft and reported seeing the landing gear being retracted.

The controller asked the DR400 pilot to wait on the runway and indicated that he would call him back and the A320 then called on a standard missed approach climbing to 4,000 feet. The TWR Supervisor decided to relieve the controller and took over with a second controller who was on a break but still in the TWR cabin taking the TWR Assistant position. The DR400 was subsequently cleared for takeoff and the A320 was positioned for a second ILS approach to runway 23 which led to a landing without further event.

The A320 Captain subsequently stated that “he did not notice the aeroplane on the runway as he was not looking at the threshold but at the wheel touchdown zone” and only when they were instructed to go around had “he realised that something serious had happened”. The A320 First Officer subsequently stated that at around 500 feet, and still not cleared to land, he had checked to see whether the runway was occupied but “did not see anything”. The Captain then asked him to interrupt the controller to obtain the landing clearance which was received. Thirty seconds later the controller instructed them to go around as there was an aeroplane on the runway and “he looked for the aeroplane but could not see anything”.

Data from the A320 QAR confirmed that a TCAS TA had been activated at 302 feet agl by the presence of the DR400 ahead but this was only indicated by an amber circle around its position on the pilots’ navigation displays (ND) since TCAS aural warnings are, by design, inhibited below 400 feet agl.

The final stages of the A320 approach/go around track/vertical profile derived from QAR data. [Reproduced from the Official Report]

Why it happened

Had the DR400 pilot not spontaneously reminded the controller of his presence on the runway, it is quite possible that given the fact that neither of the A320 pilots had acquired the DR400 visually or noticed the corresponding TCAS TA on their NAV displays, an accident could have occurred. However, the Investigation was concerned to examine the arrangements for ATC manning given that the context for the controller error involved his combination of responsibilities. This had resulted in him being sufficiently busy to have been potentially ‘overloaded’ without an opportunity for relief given the actual shift manning level.

The roster for the daytime duty period on the day of the investigated event showed the presence of six controllers. However, in accordance with formalised but unofficial unit practice, the TWR Supervisor as shift leader had the day before, in agreement with his team, reduced the programmed staff required to attend so that there were only three controllers present at the time the conflict under investigation had occurred. This was assessed by the Investigation to have been as “insufficient” and to have led not only to the combined positions which had contributed to the occurrence but also to “the unsafe immediate relief of the controller involved and the continuation of a latent risk to safe operations until the end of the shift”.

It was noted that this flexibility in controllers’ actual attendance for rostered duty had been made possible by the latitude implicitly given to Tower Supervisors’ to manage staffing levels without complying with the duty roster and without there being any means of external verification.

It was noted that “a social consensus, which has been in place for many years at the (State ANSP) DSNA, has allowed a situation to persist in which the teams of controllers organise, outside of any legal framework, for a number of staff to be present that is generally lower than the number theoretically determined as necessary”. It was further noted that “this situation, which is outside of any legal framework but known of and implicitly tolerated, is such as to bar any official collection of information that would lead to the identification of these differences, even in the context of the analysis of safety events” and that (despite) “the analysis being carried out within the scope of a just culture, unintentional errors and deviations are tolerated, but repetitive and deliberate deviations are not”.

The Investigation found more generally that “the subject of a reduced effective staffing level compared with that programmed on the duty roster and its possible impact in terms of safety is never addressed during the analysis of a safety event by the DSNA, either at local or national level (yet) access to reliable information about the presence of controllers in position and at their place of work is an essential element in air traffic safety”.

It was noted as an explicit matter of concern to the Investigation that the occurrence of the investigated event had not given rise to a modification to the organisational practices which had created the circumstances that made it more likely to occur.

No comment was made on the fact that neither A320 pilot had noticed in daylight VMC that the runway they had been cleared to land on was occupied beyond the fact that the contrast between the light grey runway threshold and the “predominantly white” DR400 was poor. No comment was made either on the pilots’ degraded situational awareness attributable to the fact that the DR400 line up and wait clearance was given in French whereas the A320 pilots were neither French nationals nor French speakers. The controller’s failure to advise the A320 that their landing clearance would be delayed due to departing traffic on the runway, which would conventionally be regarded as good practice was noted as was the fact that the strip for the DR400 had not been placed in the runway section of the TWR strip board when it was cleared to enter the runway.

The following Potential Contributory Factors were identified in respect of particular concerns:

The controller giving a clearance to land on an occupied runway: 

• The reduction in the number of staff present, which was notably lower than the number programmed on the duty roster, leading to the combination of positions of a very different nature (GND and TWR compared to APP and INFO) not provided for in the OM and thus a high workload for the controller who was simultaneously in charge of these four positions; this situation may have contributed to the strip for the DR400 not being positioned on the "runway" band of the strip board.
• The poor contrast between the DR400 and the runway surface at the runway threshold, making it more difficult for both the controller in the tower and the A320 crew to see the DR400.
• The absence of supervision of the activities within the control tower by the Tower Supervisor, who was himself carrying out the TWR assistant duties at the time.
• A probable lack of awareness of the risk generated by having an insufficient number of controllers present at their place of work, particularly in the case of an unforeseen increase in traffic or a controller having to be relieved in the case of a safety event.

The combination of positions of a very different nature not provided for in the Operations Manual and to the insufficient manning of control positions: 

• The practice of Tower Supervisors’ reducing the number of staff actually present compared with the staff programmed on the duty roster, a practice which the management know of and implicitly tolerate. 
• The absence of a tool to record, reliably and automatically and in real time, information on the manning of control positions and the number of controllers present at a workplace and to analyse the potential contribution of these two factors in the case of a safety event. Such analyses would enable Tower Supervisors and Unit Management to assess the real needs in terms of the number of staff to be present and the manning of control positions according to the levels and types of air traffic activity forecast or observed on an objective basis.

A high risk level being maintained after the serious incident: 

• The Tower Supervisor planning for a lower number of staff than that programmed on the duty roster, which did not allow for relief or reinforcements in the short term.
• The failure to question the planned organisation which had been in place despite the serious incident which had occurred.

One Safety Recommendation was made as a result of the Findings of the Investigation on the basis that “on the day of the Serious Incident, an insufficient number of controllers present at their place of work and the consequent insufficient manning of the control positions, had led to a situation of combined positions" which not only contributed to the occurrence of the investigated Serious Incident but also to “the unsafe immediate relief of the controller involved (and) the maintaining of a latent risk in operations until the end of the shift”. This Recommendation was therefore made to address:

• insufficient manning of control positions and an insufficient number of controllers present at their place of work; 
• the flexibility at control units which implicitly gives Tower Supervisors the option to reduce the number of controllers actually present during shifts compared with the number programmed on the duty roster, outside any legal framework;
• the lack of visibility for management as to the number of staff actually present;
•  the widespread extent of these practices at a national level;
• the fact that these practices, which have been in place for many years and are implicitly tolerated in the quest for social peace, mean that it is not possible to rely on a declaration system to reliably determine the manning of the control positions and the controller’s presence at work;
• the fact that the DSNA (as ANSP) currently has no means of reliably and objectively knowing the control positions manned and the presence of controllers at their place of work;
• the fact that the manning of control positions and the number of controllers present are essential components in the level of safety of the service provided;
• the fact that this absence of reliable information prevents the DSNA from analysing the potential impact of these two factors in terms of safety, at both local and national level; 
• the fact that access to reliable information on the number of staff present at the workplace and in position would enable a more relevant analysis of safety events;
• the fact that access to reliable information on the number of staff present at the workplace and in position would enable tower supervisors and their teams to be trained and made more aware of the risks involved in making position combining/splitting decisions;
• the fact that access to reliable information on the number of staff present at the workplace and in position is an essential element for checking that the number of staff programmed by the duty rosters matches the traffic actually encountered;
   
• that the DSNA equip the control centres with an automatic and nominative system to record the presence of controllers in position and at the workplace, and ensure that this information can be used by the DSNA services, in particular to ensure the appropriateness of staffing levels and to enable the analysis of safety events. [FRAN-2023-023]
  
  The Final Report was published in the definitive French version on 19 December 2023 and in English translation on 5 April 2024.

Related Articles

• ATM Shift Management
• Controller Workload
• Situational Awareness
• Multi-language ATC Operations
• Memory in ATC
• Controller Detection of Manoeuvring Area Conflicts
• See and Avoid
• Missed Approach