E190, Kupang Indonesia, 2015
E190, Kupang Indonesia, 2015
On 21 December 2015, an Embraer 195 crew continued a significantly unstable approach which included prolonged repetition of 'High Speed' and a series of EGPWS Alerts which were both ignored and which culminated in a high speed late touchdown which ended in a 200 metre overrun. The Investigation attributed the event to poor flight management and noted the systemic lack of any effective oversight of pilot operating standards compounded in the investigated event by the effects of a steep flight deck authority gradient and the failure to detect anomalies in the normal operating behaviour of both the pilots involved.
On 21 December 2015, an Embraer 195 (PK-KDC) being operated by Kalstar Aviation on a scheduled domestic passenger flight from Ende to Kupang as KD 676 overran the landing runway at destination by approximately 200 metres after continuing an unstable approach in day VMC. The aircraft sustained significant damage and one runway light was destroyed. There were no injuries to the 130 occupants who used the evacuation slide at the main door to leave the aircraft.
An Investigation was carried out by the Indonesian National Transportation Safety Committee (NTSC/KNKT). Data from the two identical Digital Voice Data Recorders (DVDR) were successfully downloaded and 25 hours of flight data with 900 recorded parameters and hours of voice data were available to assist the Investigation.
It was found that the 46 year-old Captain had a total of 9,800 flying hours which included 598 hours on type. He had joined Kalstar as a qualified Boeing 737-300/400/500 pilot and after about a year had part of the first group of pilots to be trained on the newly acquired Embraer 195 and it was planned that he would become a TRI for this type. He had passed his final line check on 23 March 2015 and a subsequent recurrent simulator check three months later in both cases recording a "good performance". The 26 year-old First Officer had a total of 2,997 flying hours which included 557 hours on type.
It was established that prior to the accident sector, the same aircraft and crew had operated a flight from Bali to Ende and the plan was for them to operate a third flight from Kupang to Makassar. The departure from Bali was delayed by over an hour because the aircraft was late arriving. This meant that the opening hours of Ende had to be extended beyond the normal closing time to allow the departure to Kupang. For the flight from Ende to Kupang, the First Office was PF. No departure briefing was performed. The weather at Kupang was poor but likely to improve by the time the flight reached there sufficiently for the visibility to reach or exceed the minimum for an approach of 3,900 metres. As the aircraft got nearer to Kupang, ATC advised that the visibility was now 4,000 metres and issued clearance to make an RNAV approach to runway 07. This approach was "discussed by both pilots" and a landing configuration for the 2,500 metre runway of flaps 5 and auto brake low was decided on. The fact that the runway was wet was not considered.
ATC was advised that the runway was in sight when passing 2,500 feet and TWR advised that the wind was Calm and issued a landing clearance. The AP was disengaged and the Captain 'suggested to the PF' that they should "calm down" and that it was "OK with high speed". An aural alert 'HIGH SPEED' began at this point and was then more or less continuous until a few seconds before touchdown. The PF reported having "noticed that all the PAPI lights indicted a white colour" and so with 7 nm to touchdown, the crew selected the landing gear down to flaps 1 and then flaps 2, a reversal of the normal procedure in order to slow down, configure as briefed and reach the correct vertical profile. However, as the aircraft passed 500 feet agl, DVDR data showed that it was descending at "more than 2000 fpm" and as it passed 200 feet, the same data recorded a rate of descent of "more than 1,500 fpm" and although the correct vertical profile had been reached, the high speed had prevented selection of the intended landing flap since the maximum speed at which flap 3 could be deployed was 200 KIAS and the speed was still at or above that when the aircraft passed 50 feet agl. This speed was 62 knots above the applicable target for that point. The Captain's instruction to the PF to delay touchdown and the excessive speed resulted in a touchdown occurring around the mid point of the wet runway. The thrust reversers were deployed normally but with auto brake set to low, less than 1,000 psi brake pressure was applied compared to the maximum available pressure of 3,000 psi. When it became clear that the aircraft would not stop before the end of the runway, the Captain took over control and steered the aircraft slightly to the right to avoid obstructions. On leaving the paved surface, the aircraft continued for approximately 200 metres before it stopped. The forward left door was opened and passengers exited using the evacuation slide.
The ground track of the aircraft as it neared and passed the end of the runway [reproduced from the Official Report]
DVDR data showed that in addition to the succession of 'HIGH SPEED' warnings, successive activations of 'SINK RATE' (three times), 'PULL UP' (twice) and 'TOO LOW TERRAIN' (twice) had occurred, beginning after 17 HIGH SPEED annunciations over a period of 22 seconds. Also, the voice recording showed that there had been no standard crew approach briefing which it was concluded was likely to have resulted in an absence of any shared understanding of the approach which was going to be flown with misunderstanding subsequently "overcome by the Captain's suggestions and commands". It was considered clear from the voice data that "most of the instruction and coordination of the way the approach was flown indicated that the Captain was dominant to the extent that a steep authority gradient was indicated with the First Officer "reluctant to express concerns, question decisions, or even simply clarify instructions".
The Investigation decided to use the available DVDR data to examine the conduct of the previous approach into Ende and discovered that on this approach, for which the First Officer had also been PF, the Captain had instructed him to delay speed reduction and shorten the approach track miles as well as reminding him whilst on short final "not to fly too high". It was noted that there was no response from the First Officer to any of these instructions/assistance, which had already been found to be a feature of the subsequent unstable approach at Kupang.
All the available evidence was considered to demonstrate that a significant authority gradient had existed on the flight deck. In this situation, it was concluded that "the decision to land was based on the Captain's belief that the aircraft could be controlled and a safe landing achieved in the prevailing conditions without proper risk assessment" and that this confidence had developed because of past experience of successful landings on short runways.
In respect of management oversight of the operation, the Investigation found that "other than the (mandatory) training and proficiency checks in the simulator, the operator had not established a system to monitor the pilot performance and compliance with company procedures especially during routine operations". Voice data examined "did not record any Checklists being read or crew briefing performed" during either the accident approach or the prior approach to Ende. This data for the accident flight "revealed that some company procedures had been neglected" and they "did not show a satisfactory standard of pilot performance".
Overall, the Conclusion of the Investigation was that neither the accident approach to Kupang nor the earlier approach to Ende were conducted in accordance with the published visual approach (Ende) or the published instrument approach (Kupang) charts. Also, in neither case did the approaches meet the minimum requirements of the company stabilised approach criteria so that both should have resulted in the initiation of a go-around and "anomalies in the normal operating behaviour of both pilots had gone undetected". It was also found that there were no controller Standard Operating Procedures (SOPs) covering the use of visibility charts or for determining what wind direction should be communicated based on the three available sources or for the dissemination of significant runway condition information to enable pilots of aircraft on approach to determine likely braking action.
Three Contributory Factors which contributed to the accident were formally identified as follows:
- The steep authority gradient resulted in a lack of synergy that contributed to a failure to correct the improper condition.
- Improper flight management on the approach resulted in the aircraft being not fully configured for landing and a prolonged high speed touchdown combined with a low brake pressure application resulted in insufficient runway for deceleration.
- The deviation of pilot performance was undetected by the Operator's management oversight system.
Safety Action taken as a result of this event and known to the Investigation included the following:
- Kalstar Aviation are developing a Flight Operations Quality Assurance (FOQA) System as a means of routinely monitoring pilot performance.
- Kalstar Aviation have enhanced pilot recurrent CRM training, reorganised the management of pilot training, emphasised the importance of flying stabilised approaches and going around from unstabilised ones and reminded all pilots of the required responses to EGPWS annunciations.
- The AirNav Indonesia District Office Kupang updated Controller SOPs covering the use of the visibility chart and the determination of wind information.
Seven Safety Recommendations were made at the conclusion of the Investigation as follows:
- that Kalstar Aviation should ensure that their pilots perform the (prescribed) operational procedures such as checklist reading and crew briefings properly and consistently. [04.O-2016-73.2]
- that Kalstar Aviation should develop an oversight system that will ensure that procedures are implemented properly and to monitor pilot performance. [04.O-2016-78.1]
- that Kalstar Aviation should ensure that a comprehensive Safety Management System is implemented correctly. [04.O-2016-79.1]
- that the Kupang Airport Operator should develop a reporting system which will enable dissemination of significant information of runway condition to Air Traffic Service (ATS) unit and for those units to provide the necessary information to arriving and departing aircraft without delay as required in ICAO Annex 14 Chapter 2.9.1. [04.B-2016-25.2]
- that the Directorate General of Civil Aviation (DGCA) should emphasise to all aircraft operators that they must comply with stabilised approach criteria. [04.B-2016-1.3]
- that the Directorate General of Civil Aviation (DGCA) should ensure that all aircraft operators implement a safety management system as required by Indonesian Civil Aviation Safety Regulations. [04.B-2016-81.1]
- that the Directorate General of Civil Aviation (DGCA) should include in the Indonesian Regulations a procedure for reporting from the airport operator to the ATS unit which enables the dissemination of significant information on runway condition to arriving and departing aircraft without delay as required in ICAO Annex 14 Chapter 2.9.1. [04.B-2016-25.3]
Furthermore, four additional Safety Recommendations actioned by their respective recipients prior to the completion of the Investigation and not repeated in the Final Report were contained in a Preliminary Report issued on 29 February 2016 as follows:
- that Kalstar Aviation should ensure that their pilots perform appropriate crew briefings. [04.O-2016-16.1]
- that Kalstar Aviation should emphasise to their pilots the importance of complying with the existing stabilised approach criteria. [04.O-2016-17.1]
- that Kalstar Aviation should emphasise to their pilots that response to any EGPWS activations must be appropriate. [04.O-2016-18.1]
- that the AirNav District Office at Kupang should include in ATC Standard Operating Procedures the procedure for the utilisation of the visibility chart and the determination of the wind information. [04.A-2016-19.1]
The Final Report was released on 27 September 2016.