IB8254 - Incident: Mel AT72 at Melilla on Mar 31st 2024

A Mel Air Avions de Transport Regional ATR-72-212A on behalf of Iberia, registration 9H-NFR performing flight IB-8254 from Almeria,SP to Melilla,SP (Spain) with 68 passengers and 4 crew, landed on Melilla's runway 15 at 08:18L (06:18Z) but veered right off the runway temporarily before returning onto the center line of the runway. The aircraft subsequently taxied to the end of the runway and backtracked the runway to the apron.

The aircraft remained on the ground in Melilla until Apr 7th 2024, then positioned to Valencia,SP (Spain) and is still on the ground in Valencia on Apr 9th 2024.

Spain's CIAIAC opened an investigation.

On Apr 12th 2024 the CIAIAC reported: "During the landing run, it was unable to maintain its trajectory, leaving the right edge of the runway, rolling momentarily along the dirt area and returning to the runway until it stopped. The occupants were unharmed and the aircraft and airport facilities were undamaged."

On Feb 23rd 2026 the CIAIAC released their final report concluding the probable causes of the incident were:

The investigation of the incident has determined that the main cause of the runway excursion to the right of the axis was the momentary loss of directional control of the aircraft during the landing run, caused by inadequate management of the flight controls in the presence of a crosswind with variable gusts.

Insufficient corrective action to recover the runway axis to prevent the excursion onto the unpaved area is considered a contributing factor to the loss of aircraft control.

The CIAIAC analysed:

The crew flew visual directly to mile 5 at the end of runway 15. Just before reaching mile 5, they began to configure the aircraft. Once aligned with the runway, configured with flaps 30 and landing gear deployed, they disengaged the autopilot and commenced descent from 2500 ft approximately 4.5 NM from the displaced runway 15 threshold. According to the records, during the descent the velocity oscillated significantly, probably due to the turbulent environment itself. The approach was steeper than usual (which is usually 3°), as the runway PAPI is 4.38°, with descent rates around 1000 ft/min. According to the captain, the final approach was normal except for a small deviation of the PAPI which he communicated to the PF who corrected appropriately and continued with a soft landing.

From 2100 ft to 600 ft above the airport, according to the records, the IAS averaged around 125 kt and the heading stabilised around 160° (QFU 148°). The left crosswind was around 30 kt, which induced a drift angle of around -15°. The flight continued from the CM1 position,26 the descent rate decreased to 1200 ft/min and the roll angle varied between -5 and 5°.

From 600 ft to 450 ft above the airport, the aircraft performed successively 8° of roll to the left and 5° of roll to the right. The heading moved left to 150° (QFU 148).

It is worth noting the increase in speed in the last 500 ft before landing, when a speed of 140 kt was reached, averaging 133 kt, seven knots above the upper limit contemplated in the MO to meet the stabilised approach criterion (VREF+20 kt, which would be equivalent to 126 kt). The approach was destabilised in speed, which was not identified as such in any operator's report.

Neither pilot was aware of this fact, for which the MO indicates aborting the approach.

The reference speed (VMBH30) corresponding to the weight of the aircraft and for the landing was 106 kt and according to the FDR data, the crew selected an approach speed (VAPP) of 116 kt (IAS 115.625 kt).

According to the records, they took as maximum IAS for stabilised approach 136 kt, where 116 kt was the VAPP plus the 20 kt, when the maximum value should have been 126 kt (106+20). The speed during the final approach was, therefore, above the upper speed limit (VREF+20) contemplated in the MO, for a stabilised approach. At no time was it identified in the records that they performed the stabilised/unstabilised callout at 1000ft or 500ft, nor was there any conversation about it by the crew.

The aircraft made contact with the runway at 06:18:30 UTC at an IAS speed of 112 kt with reverse applied. Although the captain, according to his testimony, recalled that the landing was soft and at the right runway contact point. According to the records, touchdown occurred approximately 320 m from the displaced runway 15 threshold, slightly past the last touchdown marks, probably as a result of excessive speed during the approach. According to the MO, in this case, where the landing is not to be made in the established touchdown zone, the aborted approach manoeuvre should have been executed. They therefore took a long landing which the operator in his report did not identify as such. The crew was presumably unaware of the excess speed and did not consider aborting the landing at any time.

After touchdown, the flight controls were positioned in accordance with the manufacturer's recommended technique (operator should ensure compliance with procedure ATR FCOM.PRO.NOP.ANOR.8.2)., with "horns to the wind" (roll to the right), maintaining the runway axis with the rudder pedals. However, after three seconds, the first officer operated the right rudder pedal and subsequently changed the control wheel correction (roll to the left); after five seconds after contact with the runway, the aircraft began to drift rapidly to the right, resulting in the momentary excursion with the landing gear at a speed, according to the records, of 87 kt.

The PF depressed the right rudder pedal, but the aircraft kept turning right, changing heading from 145° to 166°, and with a lateral acceleration to the right reaching 0.39 g, The steering rudder pedal force was then increased to the left, trying to control the direction of the aircraft and compensate for the deviation, reaching at 06:18:38 UTC the maximum lateral acceleration of -0.4 g. The steering rudder deflected fully to the left, the torque was 12% and the IAS was 67 kt, already showing a significant reduction in speed and managing at 06:18:46 UTC to return to the runway axis at a speed of 41 kt. According to available records, it was found that the right landing gear rolled off the paved surface for approximately 156 m and the left landing gear for approximately 50 m.

According to the aircraft's FCOM, the technique to be used during a crosswind landing by the PF is to use the rudder pedals to keep the aircraft on the runway axis so that any deviation in heading should be gently corrected especially upwind. Aircraft alignment requires less effort on the rudder pedals upwind than downwind. In strong crosswinds, it may be necessary to keep the controls crossed after landing to prevent the upwind wing from lifting and to counteract the weathervane effect. In the case of the event, the correction with the pedals was inadequate, producing a weathervane effect, with the aircraft heading into the wind, to the right, favouring a momentary departure on the far right side of the runway. The corrections downwind would have been effective had they been made smoothly according to the aircraft manufacturer's instructions, but this was not the case.

From the available records it can be deduced that the runway excursion was made without abruptness and with the aircraft stabilised, returning to the runway axis after 11" and continuing its taxi to the authorised parking position.

Neither the cabin crew nor the passengers identified any incident as having occurred; no one had noticed any significant events or sudden movements, excessive vibrations or impact noises from the cabin.

The first officer, both in the cockpit voice recorder recordings and in the interview afterwards, explained that he felt the wind changing direction, so he changed the correction.

It was the eighth time the two pilots had flown together as a crew to Melilla airport that week. During all flights, the first officer had shown no difficulties in controlling the aircraft in crosswind conditions and the LTC was confident in his skills in handling the aircraft. The day before the incident, the first officer had landed on Melilla's runway 15 with a south-south-westerly wind without incident. During the approach briefing, the pilots pointed out the particularities of the approach that day, including the crosswind and the high terrain around the airport. The first officer's training documentation provided showed no record of training on the day of the event, he had already completed the planned sectors and had already been recommended for the line-check for promotion to captain. From a training point of view, it is considered that the first officer was able to perform crosswind landings.

Analysis of weather conditions

Melilla airport is an airport with special meteorological conditions due to its orography and geographical location. These conditions were known to the crew, with previous flight experience to prove it. The crew were aware according to their testimonies that, in south-westerly wind conditions, the approach was likely to be turbulent.

Although during the briefing it was planned to land on runway 33, which is the airport's preferred runway, during the descent it was changed to runway 15 after being informed by ATC of the wind and runway conditions in service. The change of runway was done in time for the crew to do the briefing.

The PF, having noted that the wind was coming directly from the Gurugú mountain as a precaution against possible wind shear, decided to take the maximum gust value as a reference for the performance calculation and CL 100% OVRD for the landing. The crew commented on the weather information from the active SIGMET which reported a "severe mountain wave" in the Casablanca FIR affecting Melilla and based on all this information the captain, as instructor pilot, accepted the decision as appropriate, as it showed adequate situational awareness and good practice.

As the aircraft descended towards the runway, the crosswind component decreased to 15 kt.

At the time of landing, both airport windsocks indicated a south-westerly wind as did the ten-minute records. On the other hand, the first officer, according to the CVR recordings and his own statements, does not seem to have perceived the wind shift reported by the LTC, and the FDR records have also failed to establish whether or not the wind shift occurred.

The first officer, as indicated above, explained that he felt that the wind was changing direction, which is why he changed the correction.

The fact is that the airport has wind measurement units and wind gauges at both runway heads, with both pieces of measuring equipment approximately 1000 m apart. According to the ten-minute records of both runway heads, during the landing run, there was an indication of a slight difference between the two, of 14-18º in direction and from 2 kt to 6 kt in intensity, but it cannot be assured that in the middle of the runway there were no other different values.

Melilla is an autonmous city of Spain in North Africa facing the Mediterranean Sea and bordering Morocco.

Related Flight: IB8254, Iberia News