B734, Timbuktu Mali, 2017
B734, Timbuktu Mali, 2017
On 5 May 2017, a Boeing 737-400 made a visual approach to Timbuktu and slightly overran the end of the 2,170 metre-long runway into soft ground causing one of the engines to ingest significant quantities of damaging debris. The Investigation found that the landing had been made with a significantly greater than permitted tailwind component but that nevertheless had the maximum braking briefed been used, the unfactored landing distance required would have been well within that available. The preceding approach was found to have been comprehensively unstable throughout with no call for or intent to make a go around.
On 5 May 2017, a Boeing 737-400 (JY-JAP) being operated by Jordan Aviation on a scheduled domestic passenger flight from Bamako to Timbuktu as JAV 7843 departed the end of the destination runway after a fast and slightly late touchdown off a visual daylight approach and insufficient deceleration thereafter to exit as intended at the end of the runway. None of the occupants were injured but the aircraft’s left engine sustained internal damage to all its fan blades due to ingestion of multiple pieces of small gravel from the soft ground area where it ended up.
An Investigation was carried out by the Civil Aviation Regulatory Commission of the State of the Operator, Jordan. Data from the SSFDR was downloaded and used to inform the Investigation but it was found that relevant data from the SSCVR had been erased. It was not determined whether this erasure was intentional or not.
It was found that the 63 year-old Captain, who was PF for the incident flight, had a total of 9,000 flying hours on type and previous ratings on the Boeing 727 and Boeing 747 and the 30 year-old First Officer had 3,300 hours on type but no previous ratings. No total flying hours figure was recorded for either pilot. The crew also included a travelling loadmaster who was tasked with preparing each load and trim sheet for acceptance by the Captain.
Since no evidence of any airworthiness issue which could have contributed to the event was found, the Investigation focused on the conduct of the approach and landing involved. It was established that a weather report for Timbuktu received prior to the top of descent had included a surface wind from 250° at 4 knots, a temperature of 34°C and a visibility of 5 km. The performance of a visual approach was found to have been the only option for the crew after the withdrawal of the runway 07 ILS three months prior to the investigated event since the aircraft was not equipped to fly the only remaining RNAV(GNSS) instrument approach procedures. However, the runway was equipped with DMEPAPI and limited approach lighting and the DME located on the aerodrome remained in service.
Having checked in with Timbuktu TWR, the aircraft was subsequently cleared to land on runway 07 after reporting that the aircraft was on final approach and the runway was in sight at a range of 7 nm. At that point, the aircraft was fully configured for landing with flap 30 set and the Captain had briefed that he would fly the final part of the approach manually and use manual braking and maximum reverse thrust after touchdown. The ELW was 55,000 kg. At approximately 5 nm from the runway, the First Officer stated that he had noticed a tail wind component which significantly exceeded the 10 knot limit for landing on both the EHSI and the FMC and had advised the Captain accordingly and received an acknowledgement. The Captain did not recall this exchange.
Data from the FDR indicated that the approach thereafter had been unstable throughout with thrust largely set to flight idle, an abnormally high rate of descent and a speed consistently more than 10 knots above the selected Vapp (141 KIAS). The whole approach was conducted in the continuing presence of a tailwind component which soon settled at 16 knots and remained there until touchdown. At around 4 nm, the AP was disconnected followed by the A/T. At the applicable Company Stabilised Approach gate for VMC at 500 feet aal, FDR data showed a rate of descent of 1540 fpm and 165 KIAS, the latter 24 knots above the selected Vapp. No go around was called or flown despite the fact that the criteria for continuing the approach were comprehensively not met. The aircraft crossed the threshold at the correct ‘screen’ height but with the airspeed 15kts higher than Vapp and with the tail wind component still 16 knots.
Main gear touchdown on the 2,170 metre-long 30 metre-wide runway occurred at 146 KIAS and 350 metres beyond the runway threshold. Deployment of the thrust reversers to idle was made immediately but achieved with slight asymmetry, the left engine reverser deployment lagging 3 seconds behind the right resulting in a small drift to the right of 2° to which the response was a 5 second left-side-only brake application. Full thrust reversers were deployed and manual braking began once the nose landing gear was down but this did not occur until 900 metres past the threshold. However, this left only 1,270 metres of runway ahead and the initially applied brake pressure represented no more than low to medium rather than maximum. As it became clear that the aircraft would not have slowed down sufficiently by the time the end of the runway where the exit taxiway was located was reached, brake pressure was increased to maximum. As the end of the runway paved surface was crossed at a slow speed onto soft ground, the aircraft was steered to the right and stopped parallel to and 10 metres to the left of the exit taxiway (see the illustration below). There were no indications of fire and the Captain did not order an emergency evacuation. External steps were brought to the right front door of the aircraft and the occupants were disembarked. The aircraft was then removed from the soft ground by towing it backwards using cables attached to both main landing gears as no suitable tow bar was available at the airport and once back on the paved surface, the undamaged engine No 2 was started and it was taxied to the parking area.
In understanding why the unstabilised approach had been continued, the Investigation noted that the Captain had experienced delays on turnround at Timbuktu on previous visits which he attributed to non-availability of ground services and, in particular, very significant delays of up to 2 hours if refuelling was requested since fuel had to be obtained from a nearby air base. Delays of this duration would be liable to affect crew duty hours. The fact that the First Officer had failed to meaningfully monitor the Captain’s performance as PF and call exceedances and for a go around was also considered to be significant and categorised as “a lack of assertiveness” although without associated reference to the implicit challenge posed by a large difference in age and experience between the two pilots. Data from the FDR was used to examine some of the previous flights made by the same crew pairing and it became apparent that the carrying out of the mandated go around if an unstabilised approach developed “was not sufficiently ingrained by them” although these unstabilised approaches were all found to have continued to an uneventful landing.
Landing Performance was examined and it was found that at the prevailing 55,000 kg ELW and 34° surface temperature, the required unfactored flap 30 landing distance on a dry runway with maximum manual braking and thrust reversers set to the normal detent 2 position was, even after allowing for the prevailing (out of limits) tailwind component, 1583 metres which is well within the LDA of 2170 metres. It was assessed that use of the autobrake set to maximum would have produced a similar result. It could therefore be concluded that the failure of the Captain to maintain sustained maximum braking was the immediate cause of the overrun.
Given that the causes of the event were evidently to be found in the way the crew had conducted the approach and landing, the Investigation considered that it could have been avoided had it taken place in a “pressure-free environment” given that the crew had “several options available to correct the chain of events” which were summarised as follows:
- Although the notified runway in use was 07, there were no obstacles around the aerodrome that limited their use of runway 25.
- Once lined up on finals for 07, the approach was obviously unstable well before reaching 500 feet aal and Company procedures clearly stated that the pilot monitoring must require a go-around if the approach is not stabilised.
- Had consistent maximum brake pressure been applied either manually with brake pressure modulation avoided or by use of the autobrake set appropriately, the overrun could have been avoided.
- If the CRM process taught in groundschool had been applied in the flight deck, a decision to go for any of the above mentioned options would have been more likely.
The formally-stated Cause of the excursion was “a high energy unstabilised approach followed by a landing with excessive speed in the presence of a tailwind component in excess of the operational limit exacerbated by inefficient use of the wheel brakes”.
A Contributory Factor was identified as “a combination of deficiencies involving aspects of crew resource management and human factors which involved the nature of the Captain’s leadership, flight crew team work and the extent of the First Officer’s assertiveness”.
Six Safety Recommendations were made as follows:
- that Jordan Aviation should ensure that their Training and Flight Operations departments review and enhance the CRM training of the crew to raise their awareness of the importance of CRM skills. [REC 1]
- that Jordan Aviation should ensure that their Flight Operations department increases its monitoring and evaluation of crew’s abilities in flight by conducting more frequent flight inspections and including the human behaviour, CRM and Human Factors findings of the inspection in the report and then forwarding this information to the Training Department. [REC 2]
- that Jordan Aviation should ensure that its Flight Operations Department stresses the applicability of the unstabilised approach policy and in particular, the requirement to go around when the approach does not meet the stability criteria. [REC 3]
- that Jordan Aviation should stress on the importance of preserving flight data after occurrences and to set a procedure for consulting the management of the Company whenever an occurrence takes place to establish whether retention of these data is required. This procedure should include administrative guidance to technical staff who may be requested to follow the data preservation and protection procedures and record actions in the Technical Log and to sign against these actions. [REC 4]
- that Jordan Aviation and All Jordanian Operators should emphasise the importance of stabilised approach criterion and monitor the available flight data to ensure crew compliance with such procedures. [REC 5]
- that Jordan Aviation and All Jordanian Operators should conduct safety risk assessments of their operations, especially at airports where frequent unstabilised approaches are disclosed by flight data monitoring systems. [REC 6]
The Final Report of the Investigation was released on 24 April 2018.