What happened

On 27 February 2015, the pilot of a Liberty XL-2 aircraft registered VH-CZT, conducted circuits at Moorabbin Airport, Victoria. At the completion of the first circuit, the pilot flared the aircraft to land on runway 17 Left (17L). The main landing gear contacted the runway and the aircraft bounced into the air. The pilot immediately applied full power and conducted a go-around.

The pilot then conducted a second circuit and established the aircraft on final approach to the runway about 3-4 kt slower than the previous approach. The aircraft touched down normally and the pilot then performed a touch-and-go and continued the circuit.

On the third approach to runway 17L, as the pilot flared the aircraft for the landing, it collided with the runway. The aircraft then bounced into the air and the pilot initiated a go-around by applying full power and slight forward pressure on the control stick. Instead of climbing away, the aircraft struck the runway again and veered to the left. The pilot applied right rudder in an attempt to counteract the left yaw and the aircraft again became airborne before colliding with the runway and skidding towards the grass to the left of the runway.

During the accident sequence, the main landing gear collapsed and the propeller and left wingtip struck the runway, resulting in substantial damage. The pilot was not injured.

This incident highlights how different aircraft types and configurations can affect the approach profile and landing characteristics. The visual cues and the control inputs required to conduct safe landings vary depending on many factors including aircraft design and performance, weight and balance, and environmental conditions including wind strength and direction.

Aviation Short Investigations Bulletin - Issue 40

What happened

On 20 February 2015, at about 1130 Western Standard Time (WST), a Diamond DA20 aircraft, registered VH-YNB (YNB), taxied at Jandakot Airport, Western Australia, for a flight to the training area south of the airport, with a trainee instructor and a flight instructor on board. The instructor was seated in the left seat and acting as a student pilot, with the trainee instructor in the right seat. The sequence to be practiced was climbing and descending, with emphasis on keeping a good lookout and maintaining awareness of other traffic at all times.

At about the same time, the pilot of a Mooney M20 aircraft, registered VH-SJT (SJT), taxied for a private flight to Busselton Airport with three passengers on board. The instructor of YNB observed the Mooney taxiing. During the taxi, the pilot of SJT demonstrated to the front seat passenger, how to close the aircraft door. The pilot then elected to leave the door open to improve comfort in the aircraft as it was a warm day, and planned to have the passenger close the door prior to take-off.

At about 1136, the aerodrome controller (ADC) cleared YNB for take-off from runway 24 Right (24R) and advised the pilot of helicopter traffic ahead, which the trainee instructor reported in sight. About 16 seconds later, the pilot of SJT reported ready at the holding point for runway 24R. The ADC cleared SJT for take-off about 23 seconds later, at 1137, and did not advise of any traffic. The pilot of SJT then commenced the take-off run and directed the front seat passenger to close the door. The front seat passenger was unable to fully close the door, resulting in air flowing in on the passengers seated in the back seat and articles being blown around inside the cabin. The pilot continued with the take-off and asked the passengers to stay calm and silent, and the door was left partially open. Both aircraft departed Jandakot via Yangebup Lake at about 1,000 ft (Figure 1).

Figure 1: Perth Visual Terminal Chart with relevant points overlaid

Source: Airservices Australia annotated by the ATSB

After passing Yangebup Lake, the trainee instructor of YNB switched the radio from Jandakot Tower to Perth Centre frequency and changed the transponder code from 3000 to 1200.^[1] The pilot of SJT did likewise, and when changing the transponder code, realised that he had departed with the transponder selected to ‘Standby’ rather than ‘Alt’. He observed a high-wing aircraft about 2 NM away in his 2 o’clock position^[2] and assumed it was the aircraft that had departed Jandakot ahead of him and assessed that his track to Lake Thomson would be well clear of that aircraft. He then turned onto a heading of 192° and established the aircraft in a climb to 1,500 ft.

When overhead Lake Thomson, the trainee instructor of YNB conducted a climb to 1,500 ft. At 1,500 ft, he lowered the aircraft nose and levelled off to check the area ahead was clear of traffic and both pilots scanned from right to left and did not see any aircraft. The instructor then observed SJT pass diagonally from behind and left to right about 20 ft above YNB. The trainee instructor sighted SJT as it appeared from overhead to pass YNB.

At about 1141, the instructor of YNB called the pilot of SJT on Perth Centre frequency and advised that SJT had just passed straight over the top of them. The pilot of SJT responded and looked behind but did not see YNB. He assumed that it was the pilot of the high-wing aircraft, some distance away, who had contacted him and did not believe there was any risk of collision. SJT continued to Busselton with the door partially open.

At about 1144, the instructor of YNB asked the Tower controller whether SJT had been given YNB as traffic and as the ADC had handed over to another controller, was told they would find out but did not subsequently provide a response.

Pilot comments

The pilot of SJT reported that in future he plans to absolutely identify the type, and maintain visual contact with, aircraft in the control zone. He had misidentified the aircraft he sighted and should have realised it was not the Diamond that had taken off in front of him. In future, he would close the door himself prior to commencing the take-off run. Despite the distraction from the passengers due to the open door, he maintained his focus on the take-off, initial climb and after take-off checks. The distraction may have resulted in his misidentifying the aircraft ahead.

Airservices Australia investigation

Airservices Australia conducted an internal investigation into the incident and found the following:

Both aircraft were operating under the visual flight rules (VFR) and the take-off clearance was issued to SJT about 60 seconds after the clearance was issued to YNB.

The Manual of Air Traffic Services (MATS) Version 30 section 9.1.5 Traffic Information, paragraph 9.1.5.1 stated that in Class D airspace VFR aircraft will be provided traffic information on other VFR aircraft. Section 9.1.6 Traffic information assessment and content, paragraph 9.1.6.2 stated to ‘Pass traffic information to qualifying aircraft when data assessment indicates the possibility of conflict’.

No traffic information was passed to SJT on the preceding departure of YNB. The controller assessed that based on the existing separation at departure and the expected speed differential of the aircraft, no possibility of conflict existed inside Class D airspace between YNB and SJT.

The information available suggested that the aircraft came into conflict 2 NM beyond the Jandakot control zone boundary. The absence of secondary surveillance radar data for SJT was consistent with the pilot report that their transponder remained in standby mode until approximately the time of the conflict with YNB. As both aircraft were then outside controlled airspace they were not subject to a separation service. Due to the absence of surveillance data for SJT, no opportunity existed for a controller to identify the conflict.

Safety action

Whether or not the ATSB identifies safety issues in the course of an investigation, relevant organisations may proactively initiate safety action in order to reduce their safety risk. The ATSB has been advised of the following proactive safety action in response to this occurrence.

Operator of VH-YNB

As a result of this occurrence, the operator of YNB has advised the ATSB that they have reminded all students and company pilots to remain vigilant in looking outside for other aircraft.

Safety message

This incident highlights the importance of being aware of other aircraft operating in the area, particularly around airport departure points. While operating in Class D airspace, pilots and air traffic control have a dual responsibility to maintain situational awareness of other traffic. When departing into Class G airspace it is important for pilots to continue their awareness of other aircraft and to keep a good lookout at all times.

Distractions such as an open door can adversely affect the safety of a flight. The ATSB research report Dangerous Distraction: An examination of accidents and incident involving pilot distraction in Australia between 1997 and 2004, stated that the most serious source of pilot distraction occurred as a result of an unexpected equipment malfunction.

Two similar incidents were investigated by the ATSB where an open door resulted in pilot distraction. The reports are available at the following links:

www.atsb.gov.au/media/4532960/ao-2013-191_final.pdf
www.atsb.gov.au/media/4082078/ao-2012-151_final.pdf

Aviation Short Investigations Bulletin - Issue 41

About this report Decisions regarding whether to conduct an investigation, and the scope of an investigation, are based on many factors, including the level of safety benefit likely to be obtained from an investigation. For this occurrence, a limited-scope, fact-gathering investigation was conducted in order to produce a short summary report, and allow for greater industry awareness of potential safety issues and possible safety actions. Publishing information  Released in accordance with section 25 of the Transport Safety Investigation Act 2003 Published by: Australian Transport Safety Bureau © Commonwealth of Australia 2015 Ownership of intellectual property rights in this publication Unless otherwise noted, copyright (and any other intellectual property rights, if any) in this report publication is owned by the Commonwealth of Australia. Creative Commons licence With the exception of the Coat of Arms, ATSB logo, and photos and graphics in which a third party holds copyright, this publication is licensed under a Creative Commons Attribution 3.0 Australia licence. Creative Commons Attribution 3.0 Australia Licence is a standard form licence agreement that allows you to copy, distribute, transmit and adapt this publication provided that you attribute the work. The ATSB’s preference is that you attribute this publication (and any material sourced from it) using the following wording: Source: Australian Transport Safety Bureau Copyright in material obtained from other agencies, private individuals or organisations, belongs to those agencies, individuals or organisations. Where you wish to use their material, you will need to contact them directly.

__________

1. 3000 is the generic code used for civilian flights in class D airspace and 1200 is the code used for VFR flights in class G (or E) airspace.
2. The clock code is used to denote the direction of an aircraft or surface feature relative to the current heading of the observer’s aircraft, expressed in terms of position on an analogue clock face. Twelve o’clock is ahead while an aircraft observed abeam to the left would be said to be at 9 o’clock.

The Civil Aviation Safety Authority (CASA) released Airworthiness Directive AD/GEN/87 Primary Flight Control Cable Assembly Retirement in February 2015 which required the retirement of control cable assemblies after 15 years in service. Following the release of the AD, CASA received a service difficulty report where a cable terminal fitting exhibited cracking after less than 10 years in service. As the cable exhibited cracking prior to the retirement life stated in the AD, it was requested that the Australian Transport Safety Bureau (ATSB) examine the terminal to determine the mode of failure, and if it was the same mechanism as that described in the AD.

The submitted terminal exhibited four longitudinal cracks extending approximately 5mm from the end of the terminal. Small indentations were identified on either side of two of the cracks on opposite sides of the terminal. These were considered to be an artefact of the swaging process. The location and orientation of the cracking was different to those of the terminals examined previously as part of investigation AE-2012-028. There was no evidence of the highly branched cracking that was seen in other examples and which is normally associated with stress corrosion cracking.

The cracking on the submitted terminal likely initiated as a result of the mechanical damage that occurred during the manufacturing process. The terminal has been returned to CASA, who are continuing their investigations into the origin of the terminal and the mechanical damage.

Figure 1: Terminal submitted to the ATSB, crack location arrowed

Source: ATSB

Figure 2: Magnified view of terminal, showing the mechanical damage either side of the crack

Source: ATSB

______________________

The information contained in this web update is released in accordance with section 25 of the Transport Safety Investigation Act 2003.

Safety summary

What happened

At approximately 0510^[1] on 15 February 2015, the leading bogie on the leading wagon of empty Pacific National (PN) coal train NB901 derailed after climbing the railhead at 366.811^[2] km while descending the grade between Ardglen and Kankool. The bogie travelled in a derailed state for approximately 1,950 metres before it collided with No. 51 catchpoints at Kankool. The impact caused the first wagon to uncouple from the rear of the third locomotive and veer towards a cutting wall. The derailed wagon and the following 18 wagons came to rest concertinaed in various positions within an 80-metre earthen cutting. The three locomotives continued until an automatic brake application, triggered when the train parted, brought them to a stand at 369.450 km in the Kankool to Chilcotts Creek section.

Of the 19 derailed wagons, 18 sustained severe damage and were decommissioned and the third locomotive received minor damage. The derailment damaged approximately 1,900 metres of track and 2,377 concrete sleepers. This included the total destruction of 115 metres of track, three sets of points and 1393 sleepers. The train crew did not report any injuries.

What the ATSB found

The ATSB found that train management procedures, designed to keep the train stretched and reduce in-train forces on descending grades, were not adhered to as the train descended between Ardglen and Kankool. This induced sufficient compressive forces in the train to affect the first wagon. As a result, the flange tip of the left hand side wheel of No.3 axle on the leading wagon gripped and climbed the railhead of the outside (down side^[3]) rail and derailed as it transitioned a right hand curve in the section at 366.811 km.

What's been done as a result

PN revised its train-handling guidelines for Electronically Controlled Pneumatic (ECP)-braked trains and train crew monitoring program to ensure conformance to the train management requirements.

The Australian Rail Track Corporation (ARTC) replaced the damaged infrastructure and installed an additional flange lubrication system in the Ardglen to Kankool section to minimise friction in the wheel/rail interface.

Safety message

Operators are to maintain constant vigil to ensure the correct application of train management procedures. Any departure from or non-conformance to the procedures, whether major or minor, can result in catastrophic damage to both track and rolling stock.

Third locomotive of NB901

Source: ATSB

__________

1. The 24-hour clock is used in this report. Local time was Australian Eastern Summer Time (AEST).
2. Distance in kilometres from a track reference point at Sydney Central Station.
3. The down side and down side rail are on the left-hand side of any (down) train travelling away from Sydney.

The occurrence

Pacific National (PN) service NB901 was an empty coal train en route from Port Waratah (Newcastle) for loading at a mine at Turrawan, near Narrabri. It consisted of three locomotives and 82 wagons.

At approximately 0130 on 15 February 2015, a crew took charge of NB901 as it transited through PN’s maintenance facilities at Greta. During safety tests in preparation for departure, they observed a fault indication on a computer display screen in the leading locomotive. Maintenance staff identified that the fault related to the second (middle) locomotive (9206). However, As it did not affect the safety or operation of NB901, the fault was recorded for later repair.

At 0221, NB901 departed Greta for the mine approximately 40 minutes ahead of schedule. Maintenance staff conducted a roll by^[4] safety inspection as NB901 departed. The train was stopped briefly to replace a missing brake shoe key on a wagon part-way along.

The train crew described the journey after departing Greta as uneventful and NB901 did not incur any delays en route. NB901 passed through Murrurundi at 0449 where it commenced climbing the grade up and over the Liverpool Ranges. The grade crested inside a tunnel at Ardglen.

Exiting the tunnel, the driver set up dynamic braking^[5] to control the speed of NB901 down the grade towards Kankool. However, an additional 20% - 30% application of the Electronically Controlled Pneumatic (ECP) service braking system in combination with the dynamic braking, as required by PN’s Train Handling Guidelines for ECP Trains (dated 17 November 2014), was not made. The ECP brake application was required to keep the train stretched and control the in-train forces to prevent its sudden compression when the rear of the train crested the grade. Instead, the driver used only dynamic braking to control the speed and in-train forces during the descent.

Passing through Kankool at 0513, with dynamic braking still applied, the crew felt the train ‘surge’ followed by an alert of a 120% brake application that appeared on the driver’s computer screen. This alert indicated that a loss of air had occurred on the train causing an automatic brake application. Despite the automatic brake application, the locomotives continued until the driver brought them to a stand at 369.450 km^[6] at approximately 0515 by using the locomotive independent brake.

The crew noticed dust appear in the locomotive rear vision mirrors. As a result, the driver immediately reported the ‘loss of air’ on the train to the Australian Rail Track Corporation (ARTC) Network Control Centre North (NCCN) at Broadmeadow (Newcastle). At the same time, the second driver exited the locomotive cabin to investigate the situation and found that the train had parted from behind the third locomotive. Shortly afterwards, the second driver further reported that approximately 10 wagons had derailed and formed in a concertinaed pattern in a cutting.

The driver reported the derailment to the NCCN and PN’s Divisional Control Centre at Waratah before shutting down and securing the locomotives. The second driver continued further back along the track to secure the rear portion of the train.

With the arrival of first response and recovery personnel, both crew members were breath-tested onsite by supervising staff with negative results. Supervising staff subsequently drove both crew members to the PN depot at Werris Creek. Here, the crew underwent drug testing and both returned a negative result.

The Main North line remained closed for four days for recovery of the damaged rolling stock and track restoration.

__________

4. An visual inspection of a passing train conducted by maintenance staff in maintenance facilities or other train crews when in traffic.
5. See ‘Train information’ section
6. Distance in kilometres from a track reference point at Sydney Central Station.

Context

Location

Kankool is a locality situated on the Main North line, between Maitland and Werris Creek. (Figure 1) It is situated between Murrurundi, in the Upper Hunter region, and Willow Tree, in the Liverpool Plains region of NSW.

Figure 1: Location of Kankool

Source: Geoscience Australia, annotated by ATSB

Weather

Bureau of Meteorology records for the nearby township of Murrurundi indicated that weather conditions at the time of the derailment were cool, dry and clear. The minimum temperature was 14.3 °C and the maximum was 32.0 °C. The weather was found not to have contributed to the incident.

Train information

General

NB901 consisted of three locomotives (TT128, 9206, TT116), two RHDH type wagons and 80 NHDH type wagons. It had a total mass of 2,293 tonnes and measured 1,325 metres in length. Each wagon weighed 23 tonnes tare.

Train braking systems

Although the locomotives were of different classes, each was fitted with ECP (service) braking and dynamic (regenerative) braking systems that were fully compatible and controllable from the leading locomotive.

Traditional train pneumatic brakes rely on the driver’s brake controller producing a pneumatic signal relayed down a brake pipe along the train to activate a pneumatic valve and triggers air to be pathed to the brake cylinders. The force generated by the air compressing in the brake cylinders actuates a mechanical linkage and a force is applied to a friction material against the wheel tread face to brake the train. There is a time delay in braking between the front of the train and the rear of the train due to the propagation rate of the pneumatic signal along the brake pipe.

ECP braking is an electronically controlled pneumatic system that applies or releases the service brake on each wagon almost simultaneously at the driver’s command. The system connects from the locomotives to each wagon both electrically and pneumatically. Also incorporated in the system is a shadowing feature that automatically applies the brakes should the electrical or pneumatic connections between the wagons break, such as when the train parts and the train air supply inadvertently vents to atmosphere. With ECP, an electronic signal is sent to a car control device (CCD) mounted on each wagon and initiates the brakes to apply simultaneously along the length of the train. The CCD triggers an air flow to the brake cylinders and braking occurs in a similar fashion to the traditional pneumatic brake system.

The ECP system has the ability for the incremental application or release of the brakes to provide the driver with a greater control of the train braking. In downhill situations, the driver can apply dynamic braking and apply the ECP brakes in small proportioned applications to stabilise the train and improve train handling.

Dynamic braking is a function on locomotives designed to reduce wear and heat in the friction type braking equipment on the train. Being a supplementary system, it provides an additional means to control train speed however, it is not a substitute for the train air brakes (electronically controlled or otherwise). Normally, electrical power is supplied to rotate the traction motors that in turn drive the locomotive wheels. With dynamic braking, the function of the traction motors is reverse and the traction motors become electrical generators. When using dynamic braking, the current generated by the traction motors is converted to heat and dissipated through an electrical resistor bank on the locomotive. Increasing or decreasing the amount of electrical resistance varies the retardation or a braking effect on the rotating locomotive wheels and the train.

Maintenance history

Wagon maintenance history records indicated that the following scheduled inspections had been conducted on unit train 29 that formed NB901:

The records also indicated all safety critical defects located during the inspections had been actioned in accordance with maintenance standards.

Wagon couplings and in-train force control

The RHDH-type wagons were permanently coupled as a 2-pack set using a fixed drawbar coupling. The NHDH-type wagons were coupled as 4-pack sets using fixed drawbars. Automatic couplings connect the sets together and to the locomotives.

Fixed drawbar coupling assemblies are used to reduce play and to transmit longitudinal in-train forces, both tensile and compressive, generated between the wagons when the train is in motion. Resilient draft gear packages also form part of the assemblies for both drawbar and automatic couplings to dampen the in-train forces between wagons.

Train crew

General

Both crew members were attached to the PN depot at Werris Creek. The driver had a total of 14 years rail experience, 10 of which were driving trains. The assistant driver had three years rail experience as a trainee driver.

Both crew members held the required qualifications and competencies to operate the train over the route. There were no medical restrictions on either crew member.

Fatigue

An examination of the rosters for the driver and the assistant driver for the previous 14 days did not identify any evidence to support fatigue management issues in relation to the train crew involved in the incident. After two book-off days, they worked day shifts locally on Friday before crewing another train from Werris Creek to Port Waratah on Saturday. At Port Waratah, they undertook an eight-hour rest break before signing on duty at 0100 to work NB901 from Greta back to Werris Creek.

PN use a fatigue management program based on the Fatigue Audit InterDyne (FAID) system in accordance with its SMS and there was no evidence that fatigue contributed to this incident.

ECP train handling procedures

In 2011, PN and ARTC entered into a project to increase the length and capacity of coal trains operating between Narrabri and Port Waratah from 72 wagons and 25 tonne axle load (TAL) to 82 wagons with a 30 TAL. As part of the project, ARTC, as track owner, was required to upgrade the track and its maintenance procedures prior to the commencement of the 30 TAL project while PN, as rolling stock operator, was required to review the ECP train handling guidelines.

PN engaged independent contractors to conduct the engineering analysis for the 30 TAL. The analysis included the testing of fully loaded trains operating over the route on 22, 23 and 24 March 2011. The tests were used to measure the traction forces, head end dynamic braking forces and banking forces on the trains. The analysis identified and initiated mitigation strategies for any potential issues associated with the increased train lengths and axle loads, particularly when ascending and descending the steep grades over the Liverpool Ranges. The Train Handling Guidelines for ECP trains were developed as a result of these tests.

Significantly, the testing and analysis did not include unloaded trains; as was NB901. As a result, the differences in the handling characteristics between loaded and unloaded trains were not considered and not included in the guidelines. Despite this, the trial results and the Train Handling Guidelines for ECP Trains were accepted by ARTC as track owner.

The guidelines were issued to train crews as a Local Safety Notice (LSN) dated 17 November 2014. The LSN specified the train handling requirements for train crews operating ECP-braked trains in the Hunter Valley and North West Plains region.

The LSN required the following actions by the driver:

• dynamic brake be set up in preparation for the descent of grades
• an initial ECP (service) brake application of 20% to 30% (as per driver’s computer screen display) be applied (to keep the train stretched)
• the use of dynamic braking up to 250kN^[8] tractive effort
• management of the train braking to maintain constant tractive effort on the train. This included blending with ECP braking should the forces under dynamic braking exceed the prescribed 250kN.

Event recorder analysis

Analysis of the on-board event recorders from the three locomotives of NB901 for the five-kilometre section between Ardglen and Kankool indicated the following:

• Although dynamic braking was set up and engaged at the commencement of the descent at Ardglen, the LSN-prescribed 20% to 30% ECP brake application was not made
• The speed of the train exceeded the posted speed limit of 50 km/h by 3 km/h and 9 km/h on two occasions during the descent. Despite this, on both occasions, only increased dynamic braking was used to reduce the train speed
• Locomotive tractive effort peaked at 338kN on the approach to the point of the derailment (366.811 km). At this time, maximum dynamic braking was engaged with the coupling between the third locomotive and the leading wagon was most likely bearing the full weight of the trailing train
• The train was travelling at 32 km/h when it derailed at 0510.26, still with peak dynamic braking applied
• The first indication of the derailment registered at 0513.04 when the train air supply pressure fell, indicating an automatic (emergency) brake application on the train. The train was travelling at 54 km/h at the time with dynamic braking engaged and 111kN tractive effort. The automatic brake application coincided with the train parting and the train air supply venting to atmosphere at No.51 catchpoints at 368.587 km
• The locomotives continued until the driver brought them to a stand at 0514.25 at 369.450 km using the independent brake system on the locomotives.

In total, the leading wagon travelled approximately 1,950 metres in a derailed state before parting from the rear locomotive at the catchpoints and coming to rest. The locomotives continued for a further several hundred metres before coming to a stand.

Track information

General

NB901 was operating on part of the Main North line leased by the ARTC from the NSW Government. The single line, standard gauge track^[9] consisted of continuously welded 60 kg/m rail fastened to concrete sleepers by resilient clips. Crossing loops along the single-line track permitted train-crossing movements.

Track speed for freight trains descending the ruling 1:40 gradient from Ardglen to Kankool varies between 50 and 60 km/h. Track curvature in the section varies above 240 metres radius.(Figure 2)

The section between Ardglen and Kankool was not equipped with a track-mounted wheel flange lubrication system to reduce friction or wear between the wheel and the rail on curves.

Figure 2: Curve and Gradient Diagram

Source: ARTC with annotation by ATSB

Track maintenance

Track maintenance history records indicated that the following inspections had been conducted recently in the vicinity of the derailment site:

The records also indicated that, apart from the rail profile grinding, all other inspections were conducted within the required time intervals.

Train Control

ARTC’s NCCN at Broadmeadow manages the train movements between Port Waratah and Narrabri. The signalling system operates in accordance with ARTC Network Rule ANSY 500 Rail Vehicle Detection System. NB901 was operating under clear signals at the time with no proposed crossing movements before Werris Creek.

Post incident examinations

Track

The track was examined post-incident with the following observations:

• A point of mount (POM) on the down rail at 366.811 km, shortly after it had transitioned into a 240 metre radius curve, indicated that NB901 had derailed after climbing the rail head at that point. (Figure 3) The wheel mark continued for approximately one and a half metres to a point of drop (POD) where the derailed wheels had dropped from the railhead onto the sleepers and track bed.

Figure 3: Point of mount and point of drop on the down side rail

Source: ATSB

• Around the area of the POM, significant wear was observed to the gauge (inside) face of the down side rail. (Figure 4) Although the wear was still within prescribed ARTC limits, the worn gauge face was inducing full flange contact between the wheels and the railhead as trains traversed the curve. Similar rail wear conditions were also observed on the opposite up side rail on the reverse 240 metre curve prior.

Figure 4: Diagram of rail wear, contact point and forces at POM

Source: ARTC, with annotation by ATSB

• There were no rail lubrication systems fitted to this section of track or any evidence of recent rail lubrication leading up to the incident site. This was despite recommendations made in an ARTC report for higher axle loads in the Hunter Valley that the numbers and placement of lubricators be assessed to optimise lubrication effectiveness over the section between 357.000 to 374.000 kms (Pages River to Willow Tree).
• The concrete sleepers between the POM and No.51 catchpoints exhibited chipping on their approach side consistent with the derailed wheels running along the track. (Figure 5).

Figure 5: Sleeper damage caused by dragging bogie

Source: ATSB

A number of timber sleepers prior to No.51 catchpoints also exhibited flange marks on the top of the sleepers and corners of the base plates consistent with derailed wheels running over the sleepers

In the vicinity of No.51 catchpoints:

• the electric point machine and drive rods for the catchpoint suffered extensive damage
• the tip of the switch blade for the catchpoint was shattered
• the up rail had broken at a weld.

Most of the damage to No.51 catchpoints occurred when the derailed wheels tracked on the inside runoff rail of the catchpoints, breaking the up rail past the catchpoints and directing the following wagons off the track towards the cutting embankment. (Figure 6)

No.52 and No.53 points were destroyed.

Figure 6: Damage at No.51 Points

Source: ATSB

Rolling stock

NB901 was examined onsite on the day of the incident with the following noted:

• The 63 wagons remaining on track were in a serviceable condition with no obvious defective safety critical components
• A king pin, various side bearer components and a number of spring nests had been ejected from a derailed wagon and were located on the track in an area between 30 and 150 metres after the POM. (Figure 7) The presence of these components at this point indicated that a wagon body had lifted from a bogie during the derailment sequence.

Figure 7: Bogie springs on track

Source: ATSB

• Although dragged from the POM to the catchpoint, the derailed bogie tracked parallel to the track centre during its passage through the section. (Figure 8) Despite lifting at one point during the derailment sequence, the alignment of the flange marks also indicated that the derailed bogie had remained attached to the wagon body, most likely by the connecting brake lever (that was later noted still attached to the wagon body). Associated with this were wheel impact or wear marks on the external face of end slope panel of the leading end of the wagon. The wheel marks on the end slope panel were further evidence that the wagon body had lifted off the bogie during the derailment sequence.

Figure 8: Wheel impact damage on track

Source: ATSB

• While the derailed wagons concertinaed in the cutting approximately 300 metres south of the Glenyalla Road level crossing, the locomotives continued before coming to a stand several hundred metres beyond the main incident site. Examination indicated that they were in a serviceable condition with no obvious defective safety critical components. However, minor damage was observed on the rear of the third locomotive, TT116, and this most likely occurred during separation from the leading wagon, RHDH 99071M. The damage consisted of minor body panel damage, dislodgement of the connector cable for the ECP brake system, dislodgement of the receptacle box for the ECP brake system cable from its mounts and high force impact marks on the automatic coupler shank. (Figure 9)

Figure 9: Damaged electrical equipment on rear locomotive TT116

Source: ATSB

• One wagon wheel set from a bogie had detached and landed in grassy vegetation alongside the New England Highway close to passing road traffic. (Figure 10)

Figure 10: Position of wheelset near New England Highway

Source: ATSB

• The derailment of wagon RHDH 99071M initiated the derailment of a further 18 wagons which came to rest in a concertinaed pattern within an 80 metre cutting. However, due to their unsafe position, they were not examined on-site in detail. A protection order was subsequently placed on the wagons for their recovery to holding facilities at Werris Creek and later safe examination.

The derailed wagons were examined at Werris Creek on 24 February 2015 with the following observations:

• The damage to the bogie frames and wagon body of RHDH 99071M was consistent with that sustained in the derailment
• Significant wear and flattening had occurred on the flange tips of the left side wheels from the leading bogie of the leading wagon, RHDH 99071M. This damage was consistent with derailed wheels running on the ballast for a significant period. The condition of the wheels indicated that this was most likely the first wagon to derail in the incident. No wheel sets from the other derailed wagons exhibited such flange damage. (Figure 11)

Figure 11: Damage on L3 Wheel

Source: ATSB

• Measurement of the two wheelsets from the leading bogie of RHDH 99071M indicated that they were within prescribed tolerance. This included the flange thicknesses that were near full wheel profile. No abnormal or excessive wear patterns were observed on the wheels other than the flange wear and flattening. Measurements from the other wheelsets inspected also indicated they were within tolerance with no abnormal wear patterns or conditions.
• There were no abnormal wear patterns or conditions observed in the centre castings or side-bearer housings of the derailed wagon or its bogies
• There was no evidence of any contamination on the wheel flanges
• The presence of minor wheel impact or wear marks on the external face of the wagon body end slope panel indicated that the derailed bogie had remained tethered to the wagon, but did not dislodge, when dragged
• The damage to the remaining wagons was consistent with that sustained in the derailment. The safety critical components of these wagons were in a serviceable condition
• There was significant twisting and deformation observed on the rubber element plates in the draft gear packages on various wagons that dampen the longitudinal in-train forces between wagons. However, this damage did not appear to be recent or accident related. (Figures 12 and 13)

Figure 12: Typical damage to rubber element plates

Source: ATSB

Figure 13: Diagram of draft gear and fixed drawbar arrangement

Source: PN annotated by ATSB

Although controlling in-train forces, it could not be positively concluded that the condition of these rubber elements contributed any role in the derailment sequence. Despite this, PN commenced a change-out program for the draft gear package. Amongst other issues, PN also commenced a program of continued monitoring of the train handling techniques for drivers and introduced an updated inspection regime for the replacement draft gear packages.

__________

7. This inspection includes a full mechanical inspection of the train as well as a brake pipe leakage test, air brake inspection and test, brake retention test, and brake pipe continuity test.
8. Tractive effort per locomotive as displayed on the driver’s annunciator panel.
9. The terminology used for track with a width of 1435 mm between running rails.
10. This inspection covered all rail lubricators between Muswellbrook and Werris Creek.

Safety analysis

Introduction

The derailment occurred when the L3 wheel on the leading wagon of NB901, RHDH 99071M, climbed up and over the railhead. (Figure 14) With the two pack RHDH wagons immediately behind the locomotives, the derailed wagon, being first in the consist, bore considerable compressive forces on the first drawbar. These forces were the result of the downhill run in of the train, the curving resistance as it transitioned into a tight 240 metre radius curve in the track and the locomotives being dynamically braked.

Two main factors were considered likely for the wheel to climb. The first factor related to the train management of NB901 as it traversed the Ardglen to Kankool section and the non-conformance to the LSN. The second factor related to the wheel/rail interface.

Figure 14: Wheel configuration of first derailed wagon

Source: PN annotated by ATSB

Conformance to Pacific National Local Safety Notice

PN developed the LSN as part of a hazard mitigation strategy to address the hazard of in-train forces within its 82 wagon trains. The strategy also included the use of blended braking (ECP plus Dynamic) as the control measure for the hazard.

The locomotive event recorder data extracted from all three locomotives indicated that the required 20% to 30% ECP brake application had not been made in accordance with the LSN after the train crested the grade at Ardglen. Instead, only dynamic braking was engaged as NB901 commenced the descent towards Kankool. With the wagons only part way over the crest, the train remained in a ‘stretched’ condition.

As the train crested the peak of the grade, without any ECP braking applied to maintain the stretch and stabilise the train, the wagons compressed against the locomotives under dynamic braking. As a result, the compressive forces on the locomotive couplings, measured as tractive effort on the driver’s enunciator panel, increased markedly.

As the dynamic braking increased to maximum, the retardation effort further increased and eventually exceeding the LSN prescribed limit of 250kN. However, the driver did not make the required ECP brake application to stretch the train and reduce the tractive effort once it exceeded the LSN limits. The sole use of dynamic braking to manage train speed is common when driving older trains that are equipped with only conventional air brake systems.

Train crew records indicated that the driver of NB901 had long-term experience with conventional air-braked trains and limited experience with trains fitted with an ECP braking system, although deemed competent for both braking systems.

At interview, the driver stated that it was usual practice for unloaded trains to descend the grade from Ardglen to Kankool using dynamic braking only. Yet, despite an ongoing competency assessment program by PN, the driver recalled only ever undergoing one assessment on ECP trains descending steep grades. This assessment was conducted on a loaded train travelling from Werris Creek to Newcastle. However, the non-conformance to the LSN and the use of dynamic braking only on the steep declines was not picked up during the assessment.

Two over speed events of 3 km/h and 9 km/h over the posted 50 km/r speed board also occurred during NB901’s descent from Ardglen to Kankool. However, ECP braking was not used in conformance with the LSN to manage NB901’s speed or the retardation effort placed on the train.

The LSN made no distinction about any handling differences between loaded and empty trains. However, train crews considered that the LSN requirements were guidelines and related only to heavier, loaded trains, particularly as they required greater speed control when travelling towards Newcastle in the opposite direction of NB901.

Because of the non-conformances identified with the LSN, in March 2015, PN revised the LSN to clarify that it related to both loaded and unloaded trains. In June 2015, as well as reiterating the train management procedures for loaded and unloaded trains, PN further prescribed that the speed and tractive effort of trains operating in an empty condition between Ardglen and Chillcotts Creek be reduced to a maximum of 40 km/h and 230kN. In October 2015, PN reduced the area of the 40 km/h speed limit back towards Kankool.

Wheel/rail interface

Records indicated that the rails in the vicinity of the POM had been replaced as part of the 30 TAL project in July 2014. In August 2014, the profile of rails was ground so as to:

• control the contact points on the flanges of train wheels
• reduce friction between the wheel and the rail
• reduce wheel and rail wear.

The records also indicated that the rails were programmed for replacement due to railhead wear in December 2014. However, this had been deferred to ARTC’s 2015 – 2016 re-railing program. The rails were replaced post incident during track repairs.

Despite the refurbishment of the rails in August 2014, the down (outside) rail at the POM had worn heavily on the gauge face and railhead. Although still within wear limits, the wear pattern indicated that the wheels were making full flange contact on the gauge face while traversing the curve. Subsequent measurement of the railhead wear indicated a 27% material loss, this was within ARTC’s 32% maximum wear criterion. Further, the gauge face of this rail had worn to an angle of 16.5 degrees off vertical. Although within ARTC’s 26-degree criterion, the angle promoted wheel climbing due to the lack of clearance between the flange tip and the rail. Subsequently, when the train compressed, the combination of elevated lateral and longitudinal forces, a light wagon weight and the full flange contact on the outside rail all increased the likelihood for a wheel to climb. The likelihood for wheel climb is particularly greater on wheels with full thickness flanges where flatter profile angles and less flange/rail clearance promote climbing on the railhead in comparison to worn wheels. The flanges on the L3 wheel of RHDH 99071M were near full thickness.

At the time the wheel climbed the rail, the wagon was compressed between the locomotives and the other 81 wagons of the train. Without any ECP brake application made to keep the train stretched, the in-train forces compressed the wagons against the locomotives when dynamic braking was engaged. This was evident by the position of various bogie components on the ground in close proximity after the POM. While normally confined in the centre casting assembly of the bogie, their position relative to the POM suggests that the in-train forces were of such to lift the wagon body completely off the bogie. Despite the presence of light but continuous wheel marks on the railhead, the location of the expelled bogie components indicated that the wagon body had lifted after the wheel had derailed. (Figure 3)

Although the wagon body had lifted, the derailed bogie remained connected to the wagon body by the handbrake assembly remaining intact. Significantly, event recorder data indicated that the tractive effort on the locomotives was exceeding the 250kN LSN limit at the same time L3 wheel climbed the rail.

Both the running and gauge faces of the railhead were clean, dry and shiny; and without contamination from any nearby vegetation or sullage from passing trains. Further, there were no abnormal surface conditions on the running faces of the rail. As a result, all these conditions combined to cause a higher coefficient of friction between the rail and the wheel flange.

There were no abnormal conditions observed on the wheel tread or flange of L3 wheel with recent machining marks still evident. Further, despite the tight track curvature over the Liverpool Ranges, there was no evidence of any flange lubrication on the wheels of the derailed wagons. This indicated that no lubrication transfer had occurred when passing over prior flange lubrication systems. ARTC have subsequently installed additional flange lubricators in the section between Kankool and Ardglen.

Summary

The L3 wheel of the leading wagon RHDH 99071M climbed the down (outside) rail and derailed at 366.811 km when the lateral and longitudinal in-train forces of the train exceeded the vertical forces keeping the wagon on significantly worn track. Although the wear to the rails was within ARTC specifications and standards, compounding the likely causes of the derailment were a combination of full flange contact by the wheel, a lack of lubrication on the rail and dry weather conditions. Subsequently, the wagon body also lifted off the bogie. The in-train forces were not adequately managed because an ECP brake application, designed to keep the train stretched and control the in-train forces, was not made as prescribed by the LSN. The derailed wheel then continued undetected until it struck No.51 catch points where the following 18 wagons also derailed. The remainder of the train stayed on track.

Findings

From the evidence available, the following findings are made with respect to the derailment of Pacific National empty coal service NB901 at Kankool NSW on 15 February 2015. These findings should not be read as apportioning blame or liability to any particular organisation or individual.

Safety issues, or system problems, are highlighted in bold to emphasise their importance. A safety issue is an event or condition that increases safety risk and (a) can reasonably be regarded as having the potential to adversely affect the safety of future operations, and (b) is a characteristic of an organisation or a system, rather than a characteristic of a specific individual, or characteristic of an operating environment at a specific point in time.

Contributing factors

• The derailment of NB901 occurred when the heavy use of the dynamic brake without an accompanying ECP brake application generated excessive in-train forces causing L3 wheel of the leading wagon RHDH 99071M to climb the railhead on the down rail as it traversed a tight curve.
• The derailment sequence was also promoted by a lack of lubrication on a rail with a raised coefficient of friction.

Other factors that increased risk

• The Local Safety Notice, which prescribed the train handling requirements when descending steep grades, did not adequately highlight that the procedures also related to empty trains.
• Development of the Local Safety Notice was based solely on the analysis conducted on loaded trains and without reference to empty trains.

Other findings

• There were two speed infractions during NB901s descent of the grade but not at the point of derailment.

Safety issues and actions

Whether or not the ATSB identifies safety issues in the course of an investigation, relevant organisations may proactively initiate safety actions in order to reduce their safety risk. The ATSB has been advised of the following proactive safety actions taken in response to this occurrence:

Additional safety actions taken by Pacific National

Pacific National has advised of the proactive safety actions:

• the Local Safety Notice “Train Handling Guidelines for ECP Trains” was revised to also highlight the requirements for the working of empty Electronically Controlled Pneumatic braked trains on steep grades
• conformance to the requirements of the Local Safety Notice “Train Handling Guidelines for ECP Trains” were included as part of its train crew monitoring program.

Sources and submissions

Sources of information

The sources of information during the investigation included the:

• Pacific National
• Australian Rail Track Corporation
• Statements made by the crew members of NB901
• Pacific National Event Recorder Analysis Report
• Pacific National 30 TAL Project Analysis Report.

References

• Pacific National Local Safety Notice “Train Handling Guidelines for ECP Trains”
• ARTC Emergency Management procedure TA44
• RISSB Glossary of Railway Terminology Version 1.0.

Submissions

Under Part 4, Division 2 (Investigation Reports), Section 26 of the Transport Safety Investigation Act 2003 (the Act), the Australian Transport Safety Bureau (ATSB) may provide a draft report, on a confidential basis, to any person whom the ATSB considers appropriate. Section 26 (1) (a) of the Act allows a person receiving a draft report to make submissions to the ATSB about the draft report.

A draft of this report was provided to the following parties:

• Pacific National
• Australian Rail Track Corporation
• the Office of National Rail Safety Regulator
• the crew members of NB901.

Submissions were received from all parties, with the exception of the train crew of NB901. The submissions were reviewed and where considered appropriate, the text of the draft report was amended accordingly.

Purpose of safety investigations The objective of a safety investigation is to enhance transport safety. This is done through: identifying safety issues and facilitating safety action to address those issues providing information about occurrences and their associated safety factors to facilitate learning within the transport industry. It is not a function of the ATSB to apportion blame or provide a means for determining liability. At the same time, an investigation report must include factual material of sufficient weight to support the analysis and findings. At all times the ATSB endeavours to balance the use of material that could imply adverse comment with the need to properly explain what happened, and why, in a fair and unbiased manner. The ATSB does not investigate for the purpose of taking administrative, regulatory or criminal action. Terminology An explanation of terminology used in ATSB investigation reports is available here. This includes terms such as occurrence, contributing factor, other factor that increased risk, and safety issue. Publishing information  Released in accordance with section 25 of the Transport Safety Investigation Act 2003 Published by: Australian Transport Safety Bureau © Commonwealth of Australia 2018 Ownership of intellectual property rights in this publication Unless otherwise noted, copyright (and any other intellectual property rights, if any) in this report publication is owned by the Commonwealth of Australia. Creative Commons licence With the exception of the Coat of Arms, ATSB logo, and photos and graphics in which a third party holds copyright, this publication is licensed under a Creative Commons Attribution 3.0 Australia licence. Creative Commons Attribution 3.0 Australia Licence is a standard form licence agreement that allows you to copy, distribute, transmit and adapt this publication provided that you attribute the work. The ATSB’s preference is that you attribute this publication (and any material sourced from it) using the following wording: Source: Australian Transport Safety Bureau Copyright in material obtained from other agencies, private individuals or organisations, belongs to those agencies, individuals or organisations. Where you wish to use their material, you will need to contact them directly.

The pilot obtained and studied relevant meteorological forecasts and submitted to Townsville Air Traffic Control a flight plan for a flight Townsville-Julia Creek-Townsville, operating in accordance with the Visual Flight Rules (VFB). He did not hold an instrument rating and was therefore restricted to VFR operations. The planned route was via Thornton Gap and Wando Vale on each sector and the nominated cruising altitudes were 8000 feet outbound and 7000 feet on the return sector. The forecasts indicated that VFR flight in accordance with the flight plan would be possible but, over the eastern section of the route, scattered stratus cloud base 1000/2000 feet and visibility decreasing to 5000 metres was expected in rain showers. The forecast for Townsville indicated that, between 1200 and 1800 hours for periods not exceeding 30 minutes, there would be heavy rain showers with 5/8 stratus cloud base 1000 feet and visibility decreasing to 4000 metres.

The flight from Townsville to Julia Creek was completed without known incident and with only the pilot on board. Four passengers boarded at Julia Creek and the return flight to Townsville commenced.

Shortly after the aircraft departed Julia Creek, the pilot was advised of amended forecast cloud conditions for the eastern section of the route. The amended forecast included scattered stratus cloud base 1000/2000 feet, scattered cumulus cloud base 2000 feet to 4000 feet and scattered strato-cumulus. At 1439 hours the pilot reported his position as abeam of Richmond and that his amended cruising altitude was 9,000 feet.

At about 1445 hours weather conditions in the Townsville Control Zone began to deteriorate with rain and reduced visibility. A Special weather report issued at Townsville at 1500 hours indicated visibility 6000 metres in heavy showers with 2/8 stratus cloud base 600 feet. This report and subsequent special weather reports issued at 1510 hours and 1530 hours were not communicated to VH-ROC. At 1520 hours the pilot reported his position to Townsville Flight Service Unit (FSU) as Wando Vale, on descent to 5000 feet, and his estimated time of arrival at Townsville at 1600 hours. He was advised that the Townsville Control Zone was then closed to VFR operations, the Thornton Gap was closed and that Townsville Air Traffic Control considered his most suitable route to be via Charters Towers.

At 1528 hours the pilot asked if it seemed that Thornton Gap was "going to clear in the near future or not" and he was advised that advice from the Control Tower was "your best chance would be below two via the Charters Towers railway line. It doesn't look as though Thornton Gap will clear. We are open now VFR to the north through east to south but still closed to the west and south-west". The pilot replied that he would be tracking via Charters Towers.

By 1533 hours the Control Zone was opened to VFR operations but Thornton Gap, which is outside the Zone, appeared to be still in an area of adverse weather. At 1539 hours the pilot asked for an appraisal of the weather at Thornton Gap adding that "from my present position I can see the western side and it appears to be fairly good". He was informed that the tower controllers could not see much past Mount Bohle (some 7 km west of the airport) in that direction. The pilot then advised that "west of the ranges the cloud base is approximately three thousand and only broken cumulus".

At 1541 hours the pilot reported "from my position I can see most of Thornton Gap. I'd like a clearance to track via Thornton Gap". The aircraft was then operating outside controlled airspace and the pilot was instructed to call Townsville Approach Control approaching Thornton Gap. The pilot established communication with Townsville Approach Control at 1548 hours and advised "approaching Thorntons Gap at three thousand. Request clearance through Thorntons Gap. From the western side Thorntons Gap looks quite okay. The shower activity just between Thorntons Gap and the City". The aircraft was given a clearance to make a visual approach via Thornton Gap and was requested to "report one five DME". At 1550 hours, on request, the pilot advised he was 23 miles by DME from the airport and at 1554 hours, again on request, reported DME distance as "one six and we're approaching one thousand". He was instructed to continue a visual approach, given information on traffic in the circuit area and requested to report again when 10 miles by DME. The pilot did not acknowledge this instruction and information and there was no response to subsequent calls directed to the aircraft by Approach Control. The Alert Phase of Search and Rescue procedures was declared at 1605 hours and an Army helicopter operating in the area commenced search action.

The helicopter checked the eastern side of the Thornton Gap area and the pilot advised that the cloud was "right on the ground". The Distress Phase was declared at 1615 hours and, at this time, the helicopter pilot estimated the cloud base as 750 feet, some 300 feet below the level of the Gap. There was drizzling rain in the area. Soon after 1630 hours, the wreckage of VH-ROC was located on the south-eastern side of Mount Cataract, some 8 km north-east of Thornton Gap and 4 km north of the track bearing 062 magnetic from Thornton Gap to Townsville.

The aircraft had struck trees near the crest of the hill on the heading of about 015 in a slightly right wing low attitude, climbing on a 5° gradient. It was severely damaged by contact with the trees and dived steeply to the ground while rolling to the left. Detailed examination of the aircraft did not reveal any evidence of unserviceability or malfunction which might have contributed to the accident.

Evidence of persons on the ground in the vicinity of Thornton Gap and the accident site indicates that there was drizzle, rain periods, low cloud and reduced visibility in the area throughout the day.

The Occurrence

A limited-scope, fact-gathering investigation into this occurrence was conducted in order to produce this short summary report, and allow for greater industry awareness of potential safety issues and possible safety actions.

What happened

At 1455 (Western Standard Time) on 16 February 2015, Happy Buccaneer (cover) berthed at the Roy Hill Wharf, Port Hedland, which was under construction. On board were 47 modules for the wharf, with individual weights of up to 931 t. The larger modules could take up to 24 hours to lift off the ship and place into position on the wharf pilings.

At 1235 on 22 February, the fifth module had been placed in position. Later that afternoon, the ship was moved along the wharf, in preparation for a lift the next day and, at 1800, work for the day was stopped.

At 0600 on 23 February, the crew was tasked by the chief mate with clearing an area for stacking hatch covers in order to gain access to the next module. The work involved moving packs of load spreaders[1] from on top of the hatch covers to the ’tween deck.[2]  

Figure 1: Positions of the crew
 

Source: Australian Transport Safety Bureau (ATSB)

At 0630, after a tool box talk,[3] the crew started the work. Working with the second mate on deck were two able-bodied seamen (AB) tasked with hooking up the packs of load spreaders. The boatswain (bosun), the ship’s technician and two Abs were on the ’tween deck (Figure 1), to position the packs and remove the sling hooks. All crew involved in the moving of the load spreaders had performed this task before and were familiar with the operation.

The second mate was operating the crane’s[4] 25 t auxiliary hook using a wireless remote control unit. The remote allowed him to move around the area to maintain a good line of sight while lifting from the deck, and to move to the edge of the hatch opening when lowering the packs.

By about 0900, two packs of load spreaders had been moved and the third pack had been lowered to the ’tween deck. The two ABs removed their hooks and left the chain slings hanging against the side of the pack. They then moved clear, towards the centre of the hold.

The technician was having difficulty removing the sling hook on his corner so the bosun went to assist him. Once the sling hook had been freed, they began walking forward between the load spreaders and the side of the hold. As they walked, the bosun signalled to the second mate to raise the hook.

When the technician and the bosun were about half way along the side of load spreaders, the two ABs saw the topmost load spreader lift up and start to slide off the top of the pack. They shouted warnings to the second mate and the two men, however, there was not enough time for either of the men to move clear and they were both struck by the load spreader as it slid towards them.

On the bridge, the chief mate heard the loud noise of the load spreader falling. As he made his way to the ’tween deck he called the second mate on the radio and asked what had happened. When he got closer, he heard the cries of the injured men. He called the master, advised that there had been an accident and that medical assistance was required. The master asked the ship’s agent, who was on board at the time, to arrange for medical assistance from ashore.

When the chief mate arrived at the accident site, he saw the bosun and the technician lying in awkward positions on top of the load spreader (Figure 2). The technician’s right leg was pinned under the load spreader and he had fractured both legs. The bosun was not trapped but had suffered fractures to his lower left leg. The chief mate updated the master as to the severity of the situation and began arranging first aid.

Figure 2: Fallen load spreader
 

Source: AMSA

At about 0910, the first paramedics from McConnell Dowell, the company constructing the wharf, arrived. Shortly afterwards, other emergency service personnel from McConnell Dowell, the Fire and Emergency Service and St John Ambulance arrived.

The injured men were freed by the ship’s crew with the assistance of shore personnel. Both were then stabilised and prepared for evacuation. At about 1005, the bosun was lifted ashore and taken to Port Hedland hospital. About 50 minutes later, the technician was lifted ashore and taken to the hospital.

Treatment for the leg fractures of both men required extensive orthopaedic surgery. The technician’s right leg had to be amputated below the knee and he also had a fractured pelvis.

On 24 February, both men were transferred to a hospital in Perth. They continued to receive medical treatment there until they were well enough to travel home.

ATSB comment

The crew’s usual practice was to lift the load spreaders as a pack of four by attaching the sling hooks to the bottom spreader. Timber was placed between each load spreader and the pack was not lashed or secured by any means as a single unit.

After releasing the chain hooks from the spreaders, the slings were left against the side of the pack. When raised, the chain hooks were dragged up the pack’s side. In this instance, the hammerlock link (Figure 3), used to attach the sling’s hook to the chain, probably caught on the lip of the topmost load spreader and lifted it to the point where it was able to slide sideways off the pack.

Figure 3: Hammerlock joining link
 

Source: AMSA

The crane’s auxiliary hook winch was fitted with a soft start/stop feature to avoid shock loads. Therefore, when the remote control’s lever was moved to the neutral position to stop the hook, it would take a few moments for the hook to slow and then stop. This feature meant the hook continued to lift the load spreader after the lever was placed in the stop position.

When the load spreader started sliding sideways, the bosun and technician were not clear of the pack. Further, their position between the pack and the side of the cargo hold restricted their ability to move away from the falling spreader.

Safety action

Whether or not the ATSB identifies safety issues in the course of an investigation, relevant organisations may proactively initiate safety action in order to reduce their safety risk. The ATSB has been advised of the following proactive safety action in response to this occurrence.

Spliethoff's Bevrachtingskantoor

Spliethoff's Bevrachtingskantoor, Happy Buccaneer’s managers, advised that they have taken the following action to avoid a similar accident:

Additional (simulator) training and instruction will be given to crane operators with emphasis on:

• Hook speed during simultaneous crane movements, and that the crane shall not be moved until there has been clear communications by hand, radio or other method.
• Crane operators, banksmen (dogmen) and assisting crew should be aware of each other’s location and movement of suspended loads.
• Rigging gear shall be guided until it is free of obstacles and equipment should be made safe for lifting prior to the lift.
• Lifts, formed of multiple loose items, will be combined and secured to form a single unit prior to lifting.

Safety message

Other than unintended falling / dropping, there are numerous other hazards associated with suspended load operations. Some of those hazards continue posing a serious injury risk even after the load is landed. Risks can be particularly high when the load comprises a number of unsecured components that can move as individual units. Lifting and handling procedures and practices should anticipate the hazards (through risk assessment) and be planned to minimise the associated risks.

The ATSB SafetyWatch program highlights the broad safety concerns that come from investigation findings and from the occurrence data reported by industry. Marine work practices are one of those safety concerns.

Purpose of safety investigations The objective of a safety investigation is to enhance transport safety. This is done through: identifying safety issues and facilitating safety action to address those issues providing information about occurrences and their associated safety factors to facilitate learning within the transport industry. It is not a function of the ATSB to apportion blame or provide a means for determining liability. At the same time, an investigation report must include factual material of sufficient weight to support the analysis and findings. At all times the ATSB endeavours to balance the use of material that could imply adverse comment with the need to properly explain what happened, and why, in a fair and unbiased manner. The ATSB does not investigate for the purpose of taking administrative, regulatory or criminal action. Terminology An explanation of terminology used in ATSB investigation reports is available here. This includes terms such as occurrence, contributing factor, other factor that increased risk, and safety issue. Publishing information  Released in accordance with section 25 of the Transport Safety Investigation Act 2003 Published by: Australian Transport Safety Bureau © Commonwealth of Australia 2015 Ownership of intellectual property rights in this publication Unless otherwise noted, copyright (and any other intellectual property rights, if any) in this report publication is owned by the Commonwealth of Australia. Creative Commons licence With the exception of the Coat of Arms, ATSB logo, and photos and graphics in which a third party holds copyright, this publication is licensed under a Creative Commons Attribution 3.0 Australia licence. Creative Commons Attribution 3.0 Australia Licence is a standard form licence agreement that allows you to copy, distribute, transmit and adapt this publication provided that you attribute the work. The ATSB’s preference is that you attribute this publication (and any material sourced from it) using the following wording: Source: Australian Transport Safety Bureau Copyright in material obtained from other agencies, private individuals or organisations, belongs to those agencies, individuals or organisations. Where you wish to use their material, you will need to contact them directly.

[1] For placing under very heavy loads, the steel load spreaders measured 4.0 x 1.2 x 0.3 m and weighed 2.4 t.

On 28 December 2014, an Indonesia Air Asia Airbus A320 aircraft, registered PK-AXC, was conducting a regular public transport flight from Juanda International Airport, Surabaya to Changi International Airport, Singapore. Radar contact with the aircraft was lost while it was en route over the Karimata Strait, Indonesia. The aircraft impacted into water shortly after radar contact was lost. During the search and rescue operation, items of wreckage were found floating about 30 NM southeast of the aircraft’s last known radar position.

The National Transportation Safety Commission of Indonesia (NTSC) is the agency responsible for investigating this occurrence. The NTSC requested the ATSB to provide a flight recorder specialist to assist in the download and analysis of the flight recorders from the aircraft. The ATSB provided this assistance as part of the Australian Government Indonesia Transport Safety Assistance Package (ITSAP).

In accordance with clause 5.23 of Annex 13 to the Convention on International Civil Aviation (ICAO Annex 13), the ATSB appointed an Accredited Representative to assist the NTSC and initiated an investigation under the Australian Transport Safety Investigation Act 2003.

On 1 December 2015, the NTSC published the final report into this investigation.

The National Transport Safety Committee of Indonesia is responsible for releasing the investigation report.

National Transportation Safety Committee
Ministry Of Transportation Republic Of Indonesia
Transportation Building 3^rd Floor
Jalan Medan Merdeka Timur No. 5
Jakarta Pusat 10110
Indonesia

Phone  :  +62 21 384 7601
Email    :  knkt@dephub.go.id

Website: http://knkt.dephub.go.id/knkt/ntsc_home/ntsc.htm

Safety summary

What happened

On 16 February 2015, track maintenance was being conducted at Montgomery between Sale and Bairnsdale. To allow a passenger train to pass through the section, the maintenance gang cleared the track and track protection was lifted. However, when the gang vacated the line, an item of equipment was left on the track.

As the V/Line passenger train approached the work site, the driver saw the obstruction. In response, the driver made an emergency brake application but was unable to bring the train to a stand before it collided with the equipment. The train remained on the track and there were no injuries to passengers or rail personnel.

What the ATSB found

• The ATSB found that an item of maintenance equipment had been moved along the track away from the immediate area of works. Subsequently, this equipment was overlooked when the track was cleared of workers and other tools.
• The task of ensuring that the line was clear had cascaded to a third party within the maintenance gang. There was no formal system in place to manage this process of informal delegation.

What's been done as a result

As a result of this and other related Safeworking incidents V/Line has advised that a Safety Review of infrastructure rules, procedures and training has been implemented.

Safety message

Track maintenance personnel should be particularly vigilant to ensure that no obstruction remains on the line when authorising the passage of a train through a work site.

Network managers should ensure that systems and processes minimise the potential for maintenance equipment to be left on track.

The occurrence

On 16 February 2015, track works were planned for Montgomery, about 7 km from Sale. The works involved sleeper renewal and excavation works at a level crossing. To facilitate the site excavation works, a Hi-Rail excavator was required, and this was to travel on-track from Sale.

To protect the movement of this Hi-Rail excavator from Sale to the work site, and the on-track activities of the maintenance gang, the Signaller at Sale issued a Track Warrant at 0705. The Track Warrant meant that trains were prevented from entering this section of track.

Those involved in the works included:

• the Protection Officer, a contractor responsible for safe working and the holder of the Track Warrant
• a Ganger-In-Charge, a V/Line employee responsible for the site works
• a V/Line Ganger performing the role of a Leading Hand
• eight additional maintenance workers, mostly contracted staff.

As part of the maintenance works, hand and light mechanical tools were being used. One of the tools was a dog spike puller (Figure 1). At some point, when it was not being used, the dog spike puller was moved along the track, about 20 m away from the work area, and left unattended.

Figure 1: A dog spike puller similar to that used at Montgomery

Source: Melvelle Equipment Corp

The Melbourne to Bairnsdale passenger service 8403 was scheduled to depart Sale at 1020. To allow this train through the work site, the Track Warrant would need to be returned at least 20 minutes prior to its departure. To facilitate the return of the Warrant, the Protection Officer drove the 15 minutes back to the signal office at Sale Railway Station to arrive before 1000.

At the worksite, the gang stopped work at about 0955 and off-tracked in preparation for the returning of the Warrant and the passage of the train. In clearing the track, the gang overlooked the dog spike puller.

When the gang had off-tracked, the ‘Leading Hand’ of the gang gave the Ganger-In-Charge an ‘okay’ hand signal. This signal indicated that he believed the track was clear and that the Protection Officer could return the Track Warrant to the Signaller at Sale.

Using his mobile phone, the Ganger-in-Charge communicated the clear track to the Protection Officer, who was with the Signaller in Sale. As a result, the Track Warrant was returned at 1000. This released the section for train operations.

At the work site, the gang had commenced morning break and remained off track.

Train 8403 departed Sale at 1027. It consisted of an N class locomotive and five cars, with 120 passengers on board.

The train was approaching the work site at about 98 km/h. The locomotive driver noticed an obstruction on the line and in response made a full service brake application, followed shortly after by an Emergency application. He also sounded the air horn to attract the attention of the gang, although they thought it was sounded as a greeting.

The train could not stop in time to prevent the collision with the dog spike puller that had been left on the track. The equipment was pushed along the track by the locomotive and train 8403 came to a stop at 1033, about 380 m past the crossing.

Figure 2: Destroyed dog spike puller under the cow catcher of locomotive N452

Source: V/Line Pty Ltd

The rail service was terminated and passengers completed their journeys by bus replacement.

Context

Safeworking system

The line between Sale and Bairnsdale was a single line. It was one of a small number of lines in Victoria that still operated under the Train Staff and Ticket Safeworking system. This system uses a Staff as the authority for trains to travel or occupy the line between two locations. There is only one Staff for each section and it is physically transferred to the crew of the train that is authorised to use the section. The purpose of this Safeworking system is to prevent more than one train occupying a single line section between locations at any one time.

Track Force Protection

Requirements

Protection for workers must be provided before work can be carried out on a rail line. This protection can take several forms dependent on the activity. Protection can range from local hand signalling protection to manage through traffic, to a Track Warrant where trains are excluded from the section.

In this instance, a Track Warrant was the form of protection used. This protection covered both the movement of the Hi-Rail excavator from Sale and the maintenance activities of the gang at the work site.

A Track Warrant is the authority for infrastructure work activities. It can be issued by either a Train Controller or a Signaller to the Supervisor or Person in Charge of the work site as authority to foul the line. Track Warrants may be issued for infrastructure maintenance activities where one or more track vehicles or track machines are to be used.

On the Train Staff and Ticket Safeworking system, whenever a Track Warrant is issued for track force protection, the Staff for the section must also be handed to the holder of the Track Warrant. In this instance, the Protection Officer for the maintenance works held the Staff while the Warrant was in place.

Before returning a Track Warrant, the rules required the person holding the Track Warrant to ensure that the line was clear. Another requirement was that a Track Warrant was to be returned at least 20 minutes prior to the scheduled arrival time of a train that was to enter the affected section of line.

Personnel qualifications

The Protection Officer, Ganger-In-Charge and the Ganger acting as ‘Leading Hand’ all held Train Track Co-ordinator Level 3 competencies. These qualifications allowed them to institute and manage Track Force Protection arrangements and they were all aware of the requirement to ensure the track was clear before returning a Track Warrant.

Local contracting arrangements

Contractors are often used to perform the Safeworking role of providing track force protection. This was the situation in this instance, with the Protection Officer a contractor. This freed up the V/Line Ganger-In-Charge to supervise work site activities.

Safety analysis

Clearing track of equipment

It was known by all key parties that the track was to be cleared prior to the return of the Track Warrant at 1000 for the passage of a train. However, when clearing the site in preparation for the return of the Warrant, the dog spike puller was overlooked.

The dog spike puller had been moved away from the direct area of works and left unattended. It is probable that this led to the lead Ganger overlooking that it was still on the track.

The Ganger-In-Charge was occupied with other tasks and his view of the track was restricted by trucks parked near the track. He accepted the ‘okay’ signal from the Ganger acting as ‘Leading Hand’ that the track was clear. This confirmation of a clear track was then relayed by mobile phone to the Protection Officer who was offsite.

Safeworking systems and processes

Use of Track Warrant protection

The decision to protect both the Hi-Rail on-track movement to the worksite and the gang’s worksite activities with a Track Warrant was consistent with the provisions of the Book of Rules and Operating Procedures 1994. However, because of the Train Staff and Ticket Safeworking system on this line, use of a Track Warrant introduced the requirement for the Protection Officer to travel to Sale.

Return of Track Warrant and Staff

The requirements of the Train Staff and Ticket Safeworking system meant that the Protection Officer needed to return the Warrant and Staff to Sale before 1000 to allow the passage of the next train. To meet his obligations to ensure the line was clear, the Protection Officer adopted an informal process to inform himself of the condition of the line prior to returning the Track Warrant and Staff.

This practice of Protection Officers leaving a worksite on the Train Staff and Ticket system, prior to the line being clear, was an accepted practice. This practice allowed works to continue while the Protection Officer was in transit to the Signaller, and so enhanced productivity.

Informal delegation of task of ensuring track was clear

The Protection Officer was now relying on receiving a message that the line was clear from the Ganger-In-Charge. Instead of ascertaining the condition of the line themselves, the Protection Officer had informally delegated this task.

At the worksite, there was a further cascading of this task. To focus on other duties, the Ganger-In-Charge had passed immediate work-site supervision to another employee, the Ganger acting as ‘Leading Hand’. This included the responsibility of overseeing the off-tracking of the gang and their equipment prior to the passage of train 8403.

The cascading of this task was informal and there was no defined delegation of roles. All those involved were capable of undertaking the task. However, the task focus may have been diminished through the delegation. In any case, the cascading of this role probably heightened the potential for individual error.

Findings

From the evidence available, the following findings are made with respect to the collision between Passenger train 8403 and maintenance equipment at Montgomery, Victoria on 16 February 2015. These findings should not be read as apportioning blame or liability to any particular organisation or individual.

Safety issues, or system problems, are highlighted in bold to emphasise their importance. A safety issue is an event or condition that increases safety risk and (a) can reasonably be regarded as having the potential to adversely affect the safety of future operations, and (b) is a characteristic of an organisation or a system, rather than a characteristic of a specific individual, or characteristic of an operating environment at a specific point in time.

Contributing factor

• The dog spike puller was overlooked when the maintenance gang cleared the track to permit the passage of train 8403. As a result, the line was obstructed and the dog spike puller was impacted by the train.

Other factor that increased risk

• There were no formal systems in place to manage the accepted practice of Protection Officers leaving a work site to return a Track Warrant and Train Staff, prior to ceasing works, off-tracking and ensuring the line was clear. This practice led to the informal delegation of responsibility for ensuring the track was clear to others at the work site. [Safety issue]

Safety issues and actions

The safety issues identified during this investigation are listed in the Findings and Safety issues and actions sections of this report. The Australian Transport Safety Bureau (ATSB) expects that all safety issues identified by the investigation should be addressed by the relevant organisation(s). In addressing those issues, the ATSB prefers to encourage relevant organisation(s) to proactively initiate safety action, rather than to issue formal safety recommendations or safety advisory notices.

Depending on the level of risk of the safety issue, the extent of corrective action taken by the relevant organisation, or the desirability of directing a broad safety message to the rail industry, the ATSB may issue safety recommendations or safety advisory notices as part of the final report.

Safeworking systems and processes

There were no formal systems in place to manage the accepted practice of Protection Officers leaving a work site to return a Track Warrant and Train Staff, prior to ceasing works, off-tracking and ensuring the line was clear. This practice led to the informal delegation of responsibility for ensuring the track was clear to others at the work site.

Safety Issue No: RO-2015-003-SI-01

Sources and submissions

Sources of information

The sources of information during the investigation included:

• The Protection Officer
• The Ganger-In-Charge
• The Ganger (Leading Hand)
• V/Line Pty Ltd
• Skilled Rail Services.

Reference

• Book of Rules and Operating Procedures 1994.

Submissions

Under Part 4, Division 2 (Investigation Reports), Section 26 of the Transport Safety Investigation Act 2003 (the Act), the Australian Transport Safety Bureau (ATSB) may provide a draft report, on a confidential basis, to any person whom the ATSB considers appropriate. Section 26 (1) (a) of the Act allows a person receiving a draft report to make submissions to the ATSB about the draft report.

A draft of this report was provided to the Protection Officer, Ganger-In-Charge, Ganger, V/Line Pty Ltd, and Skilled Rail Services. Submissions were reviewed and where considered appropriate, the text of the draft report was amended accordingly.

Purpose of safety investigations The objective of a safety investigation is to enhance transport safety. This is done through: identifying safety issues and facilitating safety action to address those issues providing information about occurrences and their associated safety factors to facilitate learning within the transport industry. It is not a function of the ATSB to apportion blame or provide a means for determining liability. At the same time, an investigation report must include factual material of sufficient weight to support the analysis and findings. At all times the ATSB endeavours to balance the use of material that could imply adverse comment with the need to properly explain what happened, and why, in a fair and unbiased manner. The ATSB does not investigate for the purpose of taking administrative, regulatory or criminal action. Terminology An explanation of terminology used in ATSB investigation reports is available here. This includes terms such as occurrence, contributing factor, other factor that increased risk, and safety issue. Publishing information  Released in accordance with section 25 of the Transport Safety Investigation Act 2003 Published by: Australian Transport Safety Bureau © Commonwealth of Australia 2015 Ownership of intellectual property rights in this publication Unless otherwise noted, copyright (and any other intellectual property rights, if any) in this report publication is owned by the Commonwealth of Australia. Creative Commons licence With the exception of the Coat of Arms, ATSB logo, and photos and graphics in which a third party holds copyright, this publication is licensed under a Creative Commons Attribution 3.0 Australia licence. Creative Commons Attribution 3.0 Australia Licence is a standard form licence agreement that allows you to copy, distribute, transmit and adapt this publication provided that you attribute the work. The ATSB’s preference is that you attribute this publication (and any material sourced from it) using the following wording: Source: Australian Transport Safety Bureau Copyright in material obtained from other agencies, private individuals or organisations, belongs to those agencies, individuals or organisations. Where you wish to use their material, you will need to contact them directly.

Safety summary

What happened

On 11 February 2015, an Airbus A320 aircraft, registered VH‑VND and operated by Tiger Airways, was conducting a scheduled passenger service from Hobart Airport, Tasmania to Melbourne Airport, Victoria.

At about 1750 Eastern Daylight-saving Time, about 9 NM (17 km) north of Melbourne Airport, and after the flight crew had been cleared by air traffic control to conduct a visual approach, the aircraft descended below the minimum safe altitude, though the aircraft remained in controlled airspace.

During the descent, both flight crew became pre-occupied with other tasks inside the flight deck, which had the effect of increasing their workload and distracting them from monitoring the aircraft’s flight path and altitude. About two minutes after commencing descent on the visual approach, the flight crew levelled the aircraft after realising that it appeared to be low on profile. A safety alert issued by air traffic control soon followed, where in response, the aircraft was climbed to intercept the recommended visual approach descent profile. The remainder of the flight was uneventful and the aircraft landed on runway 16 at Melbourne Airport.

What the ATSB found

The ATSB found that after being vectored off the expected pre-planned shortened arrival route, and then cleared for a visual approach, a combination of increased workload and distraction diverted the flight crew’s attention from monitoring the aircraft’s descent. During the descent, the captain elected to intercept the final approach course by entering a radial intercept waypoint into the aircraft’s auto-flight system, which differed from the first officer’s more familiar plan to conduct a localiser intercept. This had the effect of diverting both crew members’ attention to inside the flight deck, as they discussed and demonstrated the intercept and resulting flight mode reversions. The aircraft continued to descend below the normal approach profile and entered the 500 ft vertical buffer at the base of the control area step. This reduced separation with terrain and any aircraft operating outside controlled airspace.

The flight crew’s mental model of the approach was not consistent with the actual flight path of the aircraft. This affected their ability to fly a normal descent profile and remain within the required control area step.

The flight crew miscalculated and did not adequately communicate the aircraft’s descent from 3,000 ft during the conduct of a visual approach. This limited their awareness of the descent rate and the below-profile altitude of the aircraft during a critical phase of flight.

Safety message

Flight crew should be mindful that during higher workload phases of flight, such as during approach and landing, introducing tasks that divert both flight crew members’ attention from monitoring the aircraft’s flight profile and altitude should be minimised. Further, if tasks that bring attention into the flight deck are required to be completed during a visual approach, pilots must ensure that at least one pilot monitors the aircraft’s flight path profile and energy state. Setting an appropriate lower altitude limit may be an effective risk control to alert flight crew and/or prevent the aircraft’s descent below a desired altitude. Communication and confirmation of any changes to the aircraft’s flight modes are also important during this period.

The occurrence

On 11 February 2015, an Airbus A320 aircraft, registered VH-VND (VND) and operated by Tiger Airways Australia Pty. Ltd. (Tigerair), was operating a scheduled passenger service from Hobart Airport, Tasmania to Melbourne Airport, Victoria. The flight, including departure and climb from Hobart, was uneventful until shortly prior to the commencement of descent into Melbourne.

Descent preparation and descent from flight level 360

At about 1710 Eastern Daylight-saving Time,^[1] and prior to commencing the descent, the flight crew were issued a clearance from air traffic control (ATC) to conduct a WAREN EIGHT ALPHA standard arrival route (STAR) for runway 16^[2](see Appendix A). In preparation for the arrival, the first officer, who was pilot flying (PF),^[3] entered the STAR into the aircraft’s auto-flight system.

After completing other normal pre‑descent flight deck preparation, the flight crew recalled that they conducted an approach and landing briefing, which included a review of the prescribed arrival route and any potential flight restrictions including controlled airspace limits and terrain along the intended flight path. Following the briefing, ATC advised the flight crew that they could expect track shortening. The PF sought clarification about the expected track shortening, to which ATC responded to expect ‘SANDR direct ROCKDALE approximately’ (see Appendix A). ATC also advised the arrival was to be flown at maximum speed and that any STAR speed restrictions were cancelled.

The PF reported amending the previously entered active flight plan to reflect the expected vectoring. The PF stated that a more accurate computed descent profile would be displayed by having the expected track shortening and the associated instrument approach as the active flight plan. At that time, they also selected the complete STAR as the secondary flight plan.

The PF recalled reviewing the controlled airspace limitations along the modified flight planned route. This revealed that the programmed altitude limitations along the flight path would maintain the aircraft within the limits of the controlled airspace steps. The captain reported that as the area along the expected track-shortened route was familiar, the control steps for the arrival along that route were not fully briefed.

After the descent preparations were complete, the PF commanded the aircraft’s auto-flight system to descend the aircraft from flight level (FL) 360^[4] to FL 250 using a managed descent flight mode. In this mode, the aircraft followed a pre-computed profile that allowed for aircraft deceleration and airspace restrictions along the active flight planned route.

Upon commencing the descent, ATC re-cleared the flight crew to descend and maintain 9,000 ft and to expect a left circuit for runway 16. To assist the flight crew with descent planning, ATC advised the flight crew that they had 46 track miles remaining to the runway.

After being cleared by ATC for further descent, the flight crew descended to 5,000 ft and advised that they were maintaining that altitude and reported visual.^[5] ATC followed with a stepped descent clearance to 3,000 ft and provided information to the flight crew that they would be vectored for a turn onto base leg of the circuit in 3 NM (6 km).

At 1748, which coincided with the aircraft being located at about the SANDR waypoint, ATC advised the flight crew that the STAR was cancelled and to turn left onto a heading of 290°M (Figure 1). To comply with ATC heading and descent instructions, the flight crew used the aircraft’s heading and vertical speed/flight path angle (V/S FPA) mode selectors, both located on the aircraft’s Flight Control Unit (FCU). The use of the aircraft’s selected modes allowed the flight crew to vary the aircraft’s lateral track and descent manually.

Figure 1: Recorded flight route of VH-VND showing approximate period where flight plan modification and mode changes were conducted

Source: Google Earth modified by ATSB

About one minute after turning onto a heading of 290°, ATC instructed the flight crew to turn further left onto a heading of 260°. While on that heading, ATC advised the flight crew that they were 6 NM (11 km) to the left of the runway 16 centreline and requested the flight crew to report the runway in sight.

At about 1749, the flight crew reported being visual with the runway. At 1749:46 ATC instructed them to turn further left onto a heading of 240° to intercept final, and cleared them to conduct a visual approach^[6] for runway 16. The flight crew were also instructed to change radio frequency and to contact Melbourne tower when established on final approach.

The captain reported inputting a radial intercept waypoint into the aircraft’s auto-flight system after being cleared to conduct the visual approach. This method of intercepting the localiser using a radial intercept waypoint was preferred as it provided more accurate tracking guidance. The method of inserting a radial intercept waypoint was a valid procedure highlighted in the operator’s flight crew operating manual (FCOM) for conducting an intercept (see the section titled Approach procedures) during a visual approach.

Visual approach from 3,000 ft

At 1749:51, the PF armed the approach mode (APPR)^[7] and engaged both autopilots, as the intent was to conduct a visual approach with an intercept of the instrument landing system (ILS) from the current heading. The PF also set a missed approach altitude of 4,000 ft in the altitude select window on the FCU, which was in accordance with the altitude setting requirements for the conduct of a visual approach in the Tigerair operations manual part B.

At 1749:54, when the aircraft was about 5 NM (9 km) from the extended runway centreline or about 10 NM (18 km) in a direct line from the runway, the PF commanded the aircraft to descend from 3,000 ft by using the vertical speed (V/S) mode. The captain reported being aware that the V/S mode was being used, but unaware of the descent rate that followed.

The flight crew made a number of selections on the FCU and aircraft configuration changes in the period from 1749:55 to 1751:15. At 1750:17, the ILS approach mode (ILS) disarmed and the lateral flight mode changed to the navigation (NAV) mode. This was likely the result of the flight crew selecting a direct to command in the auto-flight system after the ILS had already been armed. This initiated a series of automatic flight mode reversions.

The period of time for which both flight crew reportedly went ‘heads in’ (that is, both pilots had their attention inside the cockpit) the flight deck to program and discuss the radial intercept, flight mode reversions and auto-flight system changes was uncertain, but likely occurred from 1750:25 to 1751:14.

After the incident, the captain reported that by conducting the radial waypoint inbound intercept, the PF needed to go ‘heads in’ to confirm the new flight plan entry, which was probably not ideal for that phase of flight. The PF perceived this period was already one of high workload. Table 1 highlights the key events during that time.

Table 1: Relevant recorded flight data from VH-VND after the flight crew commenced the visual approach from 3,000 ft

Source: Operator quick access recorder (QAR) data modified by the ATSB

At 1751:15, about 1 minute and 21 seconds after commencing descent from 3,000 ft, the captain reportedly pressed the V/S push-button on the FCU to level the aircraft. The captain reported observing the aircraft’s flight profile at that time being ‘a bit low’, and was more consistent with the expected runway intercept altitude from a track-shortened SANDR direct to ROCKDALE (ROC) route. At 1751:24, the flight crew established communication with Melbourne tower after changing to the tower frequency.

At 1751:42, the Melbourne approach controller became aware of the aircraft’s altitude of 1,600 ft and attempted to contact the flight crew. This attempt to alert the flight crew to the low altitude failed as the flight crew had changed to the tower frequency earlier than instructed. The approach controller then instructed the tower controller to alert the crew of the aircraft’s low altitude, which the tower controller did.

At 1751:59, the tower controller issued a safety alert^[8] to the flight crew to check their altitude, as the aircraft was below profile altitude and had entered the 500 ft vertical buffer at the base of the control area (CTA) step, reducing separation with terrain and any aircraft operating outside controlled airspace.

At 17:52:23, the flight crew initiated a climb to 2,000 ft where the auto-flight system subsequently intercepted the ILS glide slope and made an uneventful landing.

After landing, the captain contacted Melbourne Air Traffic Control to discuss the altitude safety alert. They advised that the aircraft had descended to 1,600 ft, which was below the required altitude for flight in that section of CTA. The minimum altitude in that section of CTA for aircraft that were not visual was 2,700 ft, and for aircraft that were visual, it was 2,000 ft as they were required to maintain a 500 ft buffer above the 1,500 ft CTA lower limit (Figure 2).

Figure 2: Flight path and altitude of VH-VND (black and orange dots) relative to the controlled area steps and terrain during the visual approach to Melbourne Airport

Source: Airservices Australia modified by ATSB

_________

1. Eastern Daylight-saving Time: Coordinated Universal Time (UTC) + 11 hours.
2. Runway number: the number represents the magnetic heading of the runway.
3. Pilot Flying (PF) and Pilot Monitoring (PM): procedurally assigned roles with specifically assigned duties at specific stages of a flight. The PF does most of the flying, except in defined circumstances; such as planning for descent, approach and landing. The PM carries out support duties and monitors the PF’s actions and aircraft flight path.
4. Flight level: at altitudes above 10,000 ft in Australia, an aircraft’s height above mean sea level is referred to as a flight level (FL). FL 360 equates to 36,000 ft.
5. Visual (pilot usage): used by a pilot to indicate acceptance of responsibility to see and avoid obstacles while operating below the minimum vectoring altitude or the minimum safe altitude / lowest safe altitude.
6. AIP ENR 1.1 Subsection 12.8.6 stated that a pilot of an IFR flight conducting a visual approach by day must descend as necessary to remain not less than 500 ft above the lower limit of the controlled airspace.
7. Approach mode (APPR): the approach push-button parameter indicates a pressing of the APPR push-button on the FCU. This has the effect of arming the autopilot and/or flight director to capture approach guidance.
8. Safety alert: the provision of advice to an aircraft when an air traffic service officer becomes aware that an aircraft is in a position which is considered to place it in unsafe proximity to terrain, obstructions or another aircraft.

Context

Flight crew information

Captain

The captain held an Air Transport Pilot (Aeroplane) Licence and was appropriately qualified to conduct the flight. The captain:

• had about 14,380 hours of aeronautical experience, of which approximately 4,280 hours were on the A320/A321
• had about 7,800 hours total time in command
• held a current Class 1 Aviation Medical Certificate
• reported no recent or ongoing medical or personal issues.

Training

In 2013, the captain was certified as a ground instructor. This role was additional and separate to the captain’s usual activities as a line pilot. In fulfilling this role, the captain was responsible for the provision of specialist operational training as required by the head of checking and training. The ground instructor role differed from that of a line-training captain in that the duties did not pertain to in-flight training. Additional, role-specific training and competencies were required for a pilot to be endorsed as a line-training captain.

The captain had conducted a number of approaches for runway 16 at Melbourne Airport during initial and recurrent training assessments. The approaches were mainly instrument landing system (ILS) approaches, however a visual approach for runway 16 from a track-shortened route was conducted in 2014.

First officer

The first officer held an Air Transport Pilot (Aeroplane) Licence and was appropriately qualified to conduct the flight. The first officer:

• had about 2,690 hours of aeronautical experience, of which approximately 250 hours were on the A320/A321
• held a current Class 1 Aviation Medical Certificate
• reported no recent or ongoing medical or personal issues.

Training

The first officer conducted line training on the A320 in 2014, which included conducting a number of ILS approaches for runway 16 at Melbourne Airport. Although there was no recorded evidence that a visual approach was made for runway 16 at Melbourne Airport from a track-shortened route or radar vectors, the pilot had conducted visual approaches into other major airports. The pilot completed the required training program and was approved to commence line-flying operations in January 2015.

Aircraft information

Auto-flight descent modes

Managed descent mode

Flight crew normally control the descent of an Airbus A320 using the aircraft’s auto-flight system in either managed descent mode, or a selected descent mode. With the managed descent mode engaged, the aircraft follows a descent profile computed by the Flight Management Guidance System (FMGS), based upon the flight plan and descent conditions entered by the crew. This mode is only available if the aircraft follows a programmed lateral navigation track (NAV). During descent in managed descent mode, the FMGS optimises the descent profile and ensures compliance with all programmed altitude constraints without crew intervention.

Selected (basic) descent modes

In a selected descent mode, the flight crew controls the aircraft descent by making appropriate selections on the Flight Control Unit (FCU). Selected descent modes include vertical speed (V/S), flight path angle (FPA), and open descent.

In V/S mode, the auto-flight system adjusts the aircraft pitch attitude to maintain a set vertical speed as selected by the crew on the FCU. The aircraft descends to the altitude selected by the crew on the FCU, disregarding any intervening FMGS-programmed altitude constraints. If an FCU altitude limit lower than the aircraft’s current altitude is not set, the auto-flight system will continue descending the aircraft at the commanded vertical speed until the flight crew intervene, or there is an Enhanced Ground Proximity Warning System (EGPWS) (see the section titled Enhanced Ground Proximity Warning System) terrain or aircraft configuration warning alerting the flight crew to intervene. Flight crews sometimes prefer to use V/S mode to initiate further descent from a captured altitude, or incrementally change the descent rate to regain a preferred flight path profile.

Approach mode

Pushing the approach mode (APPR) push-button illuminates the APPR switch light and arms the FMGS for localiser and glideslope capture and tracking. One of the aircraft’s very high frequency navigation (VHF NAV) receivers must be tuned to an ILS frequency before APPR mode can be engaged. Once armed, localiser (LOC) is displayed in blue in the roll mode column of the Flight Mode Annunciator (FMA) and, in the pitch mode column, glide slope (G/S) is displayed in blue to indicate that the APPR mode has been armed.

The localiser capture point is variable and depends on intercept angle and closure rate with the centreline. The glideslope capture point is also variable and depends on the closure rate. The APPR light remains illuminated after localiser and glideslope capture and LOC and G/S are displayed in green on the FMA as the active engaged modes. ILS identifier, approach track and ILS/distance measuring equipment (DME) distance are displayed on the Primary Flight Display (PFD) on the lower left, when both LOC and G/S are the active modes. The localiser and glideslope deviation scales are displayed when the localiser frequency is tuned and the ILS (labelled LS on VH-VND) is selected on the Electronic Flight and Information System (EFIS) control panel.

Descent indications

The aircraft’s descent progress relative to the FMGS-computed descent profile is displayed by a symbol adjacent to the altitude scale representing the aircraft’s vertical deviation. The symbol moves above the central position as the aircraft descends beneath the FMGS-computed profile, and below the central position as the aircraft deviates above the FMGS-computed descent profile. The vertical deviation symbol remains displayed on the PFD despite the ILS push-button being selected on the EFIS panel, until the aircraft captures the LOC and GS. Deviation from the FMGS computed descent profile is also presented as a digital value on the progress page of the multipurpose control and display unit (MCDU).

In this case, the vertical deviation indicator would have been inaccurate as the indications related to the current aircraft altitude relative to the pre-programmed track-shortened route. The track‑shortened route was the active flight plan during most of the visual approach. Other available sources of information to assist the flight crew with determining a more appropriate descent profile were limited, but included:

• cross-referencing other data sources such as the DME displayed on the PFD
• monitoring the glideslope indication when within range
• referencing the runway offset distance displayed on the navigation display (ND)
• using the range ring on the ND to determine the approximate runway intercept distance
• requesting Air Traffic Control (ATC) for the derived radar distance to the runway from the current intercept heading
• using the runway PAPI or T-VASIS guidance when within range.

Figure 3: Example of a navigation display (ND) with distance information and cross track error displayed

Source: Tigerair FCTM, modified by ATSB

Enhanced Ground Proximity Warning System

The aircraft was fitted with an EGPWS. The EGPWS considers a range of data and in-flight parameters, and provides a distinctive warning to flight crew if the aircraft enters a potentially hazardous position in relation to the earth’s surface. No EGPWS warnings were triggered during this occurrence.

Meteorological information

The flight crew reported that the weather conditions at the time of the incident did not adversely affect the ability to conduct the flight. The flight crew confirmed that visual conditions existed prior to leaving 3,000 ft for the visual approach.

Operational information

Navigation

Operation of the Instrument Landing System (ILS)

The ILS provides lateral and vertical position data necessary to align the aircraft with the runway for approach and landing. The system uses angular deviation signals from the glideslope antennas (located approximately 1,000 ft from the touchdown point on the runway) and the localiser antennas (located past the far end of the runway). The glideslope signals provide the angular deviation from the nominal glide path (usually 3°) and the aircraft’s auto-flight system generates fly-up or fly-down commands to enable the flight crew to track the glide path down to the touchdown point on the runway. Glideslope deviation is displayed on the PFD in units of dots, where one dot equates to 0.4° deviation from the glide path.

The localiser signals provide the angular deviation from the runway centreline and the autopilot generates fly-left or fly-right commands to track the centreline until the landing roll is completed. Localiser deviation is displayed on the PFD in units of dots where one dot equates to 0.8° deviation from the localiser.

Figure 4: Primary flight display showing an example of instrument landing system (ILS) scale indications

Source: Tigerair FCOM

Approach procedures

Standard Arrival Route (STAR) information

To program a STAR such as the WAREN EIGHT ALPHA arrival into the FMGS, the flight crew were required to select the appropriate arrival from the MCDU flight plan arrivals page. After selecting the arrival, a relevant approach such as the runway 16 ILS approach is selected and appended to the STAR. This inserted additional en route waypoints such as the final approach fix (FAF) to the flight plan and would later provide localiser and glideslope guidance when within the capture area. Manual selection of the ROCKDALE (ROC) Non-directional beacon (NDB) waypoint was required to amend the flight plan route to comply with ATC clearances and other enroute constraints, before briefing the arrival and the approach.

Selection of the full WAREN EIGHT ALPHA STAR included tracking to intercept the final approach path for runway 16 at a greater distance than the track-shortened route from SANDR to ROCKDALE. To intercept a 3° profile from the track-shortened route, the aircraft would intercept final approach at about 4 NM (7 km) from the runway at an altitude of about 1,680 ft.

In contrast to the track-shortened route, the calculated runway intercept distance if the flight crew maintained a heading of 240° after flying radar vectors to the north, was about 9 NM (17 km). The profile altitude required for a recommended 3° profile at that distance was about 3,500 ft.

Visual approach

The operator’s flight crew operating manual (FCOM) outlined the standard operating procedure for the conduct of a visual approach. It included general information about conducting the approach on a nominal 3° glideslope, using visual references. The method for conducting a visual approach included:

• the autopilot is off
• both flight directors are off
• the use of flight path vector (‘the BIRD’)^[9] is recommended
• autothrust use is recommended with managed speed.

The flight crew training manual (FCTM) highlighted that, although the approach should be flown visually, having the cross-track error distance displayed on the ND provided the pilot with a visual cue as to the lateral position of the aircraft to the runway centreline. The cross-track error could be obtained by performing a direct to (DIR TO) radial inbound intercept on the last available waypoint (such as ROCKDALE or the ILS FAF) positioned on the extended runway centreline.

Although the FCTM recommended that the visual approach be conducted by amending the FMGS flight plan to include a radial inbound intercept, the company’s operations manual Part A stated that any flight path changes to the FMGS below 10,000 ft should be avoided where possible. Instead, it highlighted that using basic flight modes (selected modes) in the terminal area was preferred where a visual procedure could not be planned. This ensured that flight crew’s primary focus of attention was monitoring the aircraft’s flight path, the surrounding terrain, and potential aircraft conflicts, by having ‘two heads up’ at all times. Further, it was highlighted that flight crew must not rely solely on the FMGS, but should reference all applicable navigation aids to ensure safe navigation in the terminal area.

The missed approach altitude setting requirements for a visual approach was highlighted in the Tigerair operations manual Part B. It stated that for visual approaches, the missed approach altitude for the instrument missed approach procedure for the landing runway must be set.

Final approach course intercept

The FCTM highlighted to flight crew that, to ensure a smooth interception of the final approach course, the aircraft’s ground speed should be appropriate for the runway intercept angle and the distance remaining to the runway (Appendix B). In an attempt to ensure a smooth interception of the extended runway centreline, the Captain elected to input a radial inbound waypoint. The FCTM stated that where ATC provided radar vectors, the flight crew would use the direct to radial inbound (DIR TO RADIAL IN-BND) function. This would:

• ensure proper flight plan sequencing
• provide a comprehensive ND display
• assist lateral interception
• allow for the vertical deviation to be computed on reasonable distance assumptions.

When intercepting the final approach course using this method, the flight crew should correctly sequence the flight plan before pressing the approach push-button. If the localiser was armed or engaged before a DIR TO was actioned, the armed flight mode would revert to NAV, meaning that the localiser would have to be rearmed, which could increase workload. In this occurrence, the PF had pressed the approach push-button before the DIR TO was actioned.

Descent monitoring procedures

The operator’s procedures provided guidance regarding descent monitoring. This included the reference to appropriate pages on the MCDU and the use of vertical profile information on the PFD, where applicable. The descent procedure called for careful monitoring, including guidance that during non-precision approaches, appropriate distance/altitude checks will be called. This was of particular importance where an altitude/height versus range/fix was required.

FMA monitoring

The operator’s procedures required the pilot flying to ‘announce’ the FMA following the initiation of the descent, and for the pilot monitoring to confirm that annunciation. This required the pilot flying to state the auto-flight mode change annunciated on the FMA associated with the commencement of descent, and the pilot monitoring to check the annunciation and respond. Importantly, the effect of those changes on the flight path must be monitored on basic flight instruments associated with heading, speed, altitude, V/S and the like.

Other occurrences

The ATSB is aware of a number of occurrences on scheduled passenger transport flights where a flight crew have descended either below their normal flight path profile during a visual approach or below a minimum descent altitude while conducting an instrument approach. These involved different operators and different aircraft types to the occurrence involving VH-VND, but the fundamental nature of these occurrences is similar (see www.atsb.gov.au).

ATSB investigation AO-2011-086

At 2019 Eastern Standard Time on 24 July 2011, a Thai Airways Boeing 777-3D7 aircraft, registered HS-TKD, was conducting a runway 34 VOR approach to Melbourne Airport, Victoria. During the approach, the tower controller observed that the aircraft was lower than required and asked the flight crew to check their altitude. The tower controller subsequently instructed the crew to conduct a go-around. However, while the crew did arrest the aircraft’s descent, there was a delay of about 50 seconds before they initiated the go-around and commenced a climb to the required altitude.

The ATSB established that the captain may not have fully understood some aspects of the aircraft’s automated flight control systems and probably experienced ‘automation surprise’ when the aircraft pitched up to capture the VOR^[10] approach path. As a result, the remainder of the approach was conducted using the autopilot’s flight level change mode. In that mode, the aircraft’s rate of descent is unrestricted and therefore may be significantly higher than that required for an instrument approach. In addition, the flight crew inadvertently selected a lower than stipulated descent altitude, resulting in descent below the specified segment minimum safe altitude for that stage of the approach and the approach not being managed in accordance with the prescribed procedure.

ATSB investigation AO-2012-103

On 16 July 2012 at about 0830 New Zealand Standard Time^[11], an Airbus A320-232 aircraft, registered VH-VQA and operated by Jetstar Airways (Jetstar), was conducting an Area Navigation (Required Navigation Performance) approach to runway 05 at Queenstown, New Zealand. During the approach the aircraft descended below two segment minimum safe altitudes. Upon recognising the descent profile error, the crew climbed the aircraft to intercept the correct profile and continued the approach to land.

The ATSB found that, contrary to their intentions, the flight crew continued descent with the auto-flight system in open descent mode, which did not provide protection against infringing the instrument approach procedure’s segment minimum safe altitudes. The ATSB also found that the flight crew was not strictly adhering to Jetstar’s sterile flight deck procedures, which probably allowed them to become distracted.

The ATSB found that the Jetstar procedures did not specifically draw the flight crew’s attention to unchanged auto-flight system modes during descent or prompt crew reconsideration of the most suitable descent mode at any point during descent. Additionally, the Jetstar’s procedures allowed the crew to select the altitude to which they were cleared by air traffic control on the flight control unit altitude selector, irrespective of intervening altitude constraints. This combination of procedures provided limited protection against descent through segment minimum safe altitudes.

ATSB investigation AO-2013-047

On 8 March 2013, the flight crew of a Qantas Airways Limited (Qantas) A330 aircraft, registered VH-EBV, was conducting a visual approach to Melbourne Airport, Victoria. The captain was the pilot flying with autopilot engaged.

Soon after being cleared for the approach, on descent through 3,000 ft, the captain set an altitude target of 1,000 ft in the auto-flight system and selected the landing gear down, the first stage of wing flap and 180 kt as the target speed. The descent was continued in auto-flight open descent mode and reached a maximum descent rate of 2,200 ft/min. As the aircraft was descending through about 1,800 ft, the first officer advised the captain that they were low. The captain reduced the rate of descent by selecting auto-flight vertical speed mode but a short time later the enhanced ground proximity warning system (EGPWS) provided ‘TERRAIN’ alerts followed by ‘PULL UP’ warnings. The crew carried out an EGPWS recovery manoeuvre and subsequently landed via an instrument approach.

At the time of the EGPWS alert, the aircraft had descended to 1,400 ft, which in that area was 600 ft above ground level, with 9 NM (17 km) to run to touchdown. This was 100 ft below the control area lower limit and 1,900 ft below a normal 3° descent profile.

ATSB investigation AO-2014-003

While on approach to Melbourne, Victoria, a Jetstar Airways (Jetstar) Airbus A320 aircraft left 3,000 ft on descent, entering the 500 ft buffer above the lower limit of controlled airspace. When the aircraft passed 2,500 ft, the aircraft left controlled airspace. The aircraft again re-entered controlled airspace as it reached the airspace with a lower limit 1,500 ft, 11 NM south of Melbourne. The elapsed time from the point the aircraft left 3,000 ft to the point it re-entered controlled airspace was about 1 minute and 15 seconds. The aircraft was outside controlled airspace for about 45 seconds. There was no conflict with other known air traffic and the approach continued normally from 2,100 ft following intercept of the intended descent profile.

This incident highlighted the need for clear procedural guidance and careful auto-flight system management under conditions where the transition from a STAR to an instrument approach procedure is interrupted. Furthermore, under these conditions, awareness of the position of the aircraft relative to the intended vertical profile, relevant controlled airspace boundaries and lowest safe altitudes assumes elevated significance. The incident also highlights the importance of seeking clarification if an ATC instruction or clearance appears incomplete.

__________

9. Flight path vector on the Primary Flight Display is used to monitor the descent profile (often referred to as the BIRD).
10. A ground-based navigation aid that emits a signal that can be received by appropriately-equipped aircraft and represented as the aircraft’s bearing (called a 'radial') to or from that aid.
11. New Zealand Standard Time (NZST) was Coordinated Universal Time (UTC) + 12 hours.

Safety analysis

Introduction

While conducting a visual approach into Melbourne Airport, Victoria on 11 February 2015, the flight crew of an Airbus A320 aircraft descended below the nominal 3° descent profile and entered the 500 ft vertical buffer at the base of the control area (CTA) step. This reduced separation with nearby terrain and with any aircraft operating outside controlled airspace. The flight crew were alerted to the aircraft’s low altitude by air traffic control after levelling the aircraft close to the lower limit of controlled airspace. The flight crew subsequently climbed the aircraft to regain a normal approach profile by capturing the Instrument Landing System (ILS) glide slope. The approach continued and an uneventful landing was made onto runway 16. This analysis will examine the factors that contributed to the abnormal descent, and review the risk controls as they apply to visual approaches.

Approach and descent management

Prior to descent, the flight crew were cleared by Air Traffic Control (ATC) to commence an arrival to Melbourne by flying the WAREN EIGHT ALPHA standard arrival route (STAR). The flight crew reported fully briefing the arrival and associated ILS approach, which included reviewing the arrival route, CTA steps and surrounding terrain. Soon after, however, the STAR clearance was amended to a track-shortened route which brought the final runway intercept closer, to about 4 NM (7 km) from the threshold, with an on-profile intercept altitude of about 1,680 ft. Although the flight crew did not fully re-brief the entire track-shortened route at that time, they reported reviewing any applicable restrictions/limitations and calculated the aircraft’s descent based on those requirements.

As the aircraft approached the SANDR waypoint, however, the flight crew were radar vectored to the north of the expected track-shortened route, which positioned the aircraft in an area that was unplanned by the flight crew. It also meant that there was little time to re-brief and review the surrounding terrain and to identify the higher CTA step that the aircraft was being vectored into. During the radar vectoring, ATC was responsible for terrain clearance and flight within controlled airspace; on completion of the vectors, however, ATC had instructed the flight crew to resume own navigation and that they were cleared for the visual approach. This transferred the responsibility for terrain clearance and flight within the CTA steps back to the crew. Despite ATC giving the flight crew a position fix from the extended runway centreline, it was likely that this information alone was insufficient, and the flight crew had lost awareness of the altitude that would be required to remain above the vertical buffer for the CTA steps.

To comply with the vectors north of the SANDR waypoint, the flight crew needed to revert to basic flight modes (selected modes) and forego the programmed altitude constraints that were active to limit the aircraft had it remained on the STAR. This meant that more onus was placed on the flight crew to adjust the aircraft’s vertical profile manually to ensure the aircraft remained on the desired 3° visual approach flight path profile. Further, the altitude protections set on the aircraft’s flight control unit (FCU) were removed when the missed approach altitude for the landing runway’s instrument approach was set as per procedures. As this was higher than the aircraft’s altitude, more attention was required from the flight crew to ensure the aircraft did not descend below the desired altitude.

Reverting to basic flight modes (selected modes) and deviating off the pre-programmed route also removed some of the information available to assist the flight crew in managing the aircraft’s vertical profile. The vertical deviation indicator, which was available if the ILS push-button on the EFIS was not selected, was only accurate if the aircraft was on, or close to, the programmed route.

The captain reported that an alternative method of using the ILS glideslope as a reference was useful, and recalled mentioning this to the first officer, who was the pilot flying (PF) during the approach. The captain observed that during the approach, the ILS glide slope indications were active and indicated that the aircraft was initially above the glide slope, which was consistent with their expectation. As a result, the flight crew probably considered the glide slope indications to be valid and useful for flight path guidance. It was therefore likely that when not focussed on the activities associated with reprogramming the Flight Management and Guidance System (FMGS) and demonstrating the radial intercept, which resulted in flight mode reversions, the PF used the glide slope for profile guidance. This probably resulted in the early descent from 3,000 ft during the visual approach, and the multiple V/S changes made by the PF in an attempt to capture the on-slope indications. Following a glide slope indication with more track miles to fly than the straight-line distance to the glideslope antenna would, with reference to the track miles to run, result in a shallower flight path than the nominal 3° profile. This required a reduced descent rate from what would normally be expected. In addition, the PF also became more distracted as the aircraft approached the on-slope indications and the aircraft continued to descend below profile. Reference by the crew to the other available cues before commencing descent for the visual approach would have increased the likelihood of the aircraft remaining at an appropriate profile altitude until the runway intercept.

Approach path profile

The method used by the flight crew to calculate the required descent point to achieve an optimal 3° profile was not effective, as the descent had commenced early given the aircraft’s actual versus intended position. Further, the PF was likely uncertain about the aircraft’s profile and relative position as it descended, as indicated by the number of VS changes made, the appropriateness of those changes, and that the aircraft continued to descend well below the nominal 3° visual approach profile. The time available and the PF’s capacity to calculate and re-assess the aircrafts position and descent was likely impeded by having:

• to re-assess the relative position of the aircraft and its appropriate configuration
• to perform the tasks associated with the role of PF
• to divert their attention to observe the reprogramming of the FMGS
• to observe the demonstration of the flight mode reversions
• to complete the aircraft configuration and checklists in preparation for landing
• a resulting increased workload.

The operations manual Part A recommendation was not to use FMGS or make flight plan changes in the terminal area but instead to use basic flight modes (selected modes) to conduct an intercept. This would have provided the flight crew the opportunity to prioritise their attention on more critical tasks associated with the final stages of the approach and ensured both crew were not focussed inside the flight deck at the same time. This was important as the conduct of a visual approach using selected flight modes put more onus on the flight crew for flight path management.

In addition to the increased attention required to maintain the flight path profile, the altitude setting procedure to set the missed approach altitude for the runway instrument approach removed the only altitude constraint that would automatically level the aircraft when using selected descent modes. Other alerts, however, such as the aircraft’s enhanced ground proximity warning system (EGPWS), would still have been available to alert the crew to a low altitude/terrain proximity as a final defence, in the event that the aircraft’s descent continued undetected.

Data entry and crew co-ordination

During the visual approach, it was likely that there was a degree of demonstration or instruction between the flight crew in relation to data inputting in relation to the radial intercept and mode awareness. Although it is possible the intention of the captain was to increase the first officer’s understanding of the auto-flight system and approach management, the captain was not approved as a line-training captain. It is also likely that the captain was required to explain the radial intercept to the first officer so the first officer could check the inputs made by the captain. It is therefore likely that, independent of an instructional focus, there was probably little consideration of the effect the instruction or explanation would have on the relatively inexperienced first officer’s workload or the ability of the flight crew to manage the aircraft’s descent during that phase of flight.

Distraction

Researchers (United Kingdom Civil Aviation Authority, 2013) have found that distraction has been a major factor affecting flight crew allocation of attention when monitoring breaks down. Humans are capable of attending to more than one task through the use of selective attention techniques, however they have limited total cognitive capacity. If one of the tasks consumes all the attentional capacity of a crew member, then task shedding will occur. Distraction has been found to have been instrumental in the breakdown of monitoring in major accident investigations^[12]. In these instances, flight crew became distracted during an important phase of flight. This distraction resulted in a breakdown in monitoring and combined with the flight crew inappropriately managing their workload, this led to the loss of the pilot flying’s understanding of the state or position of the aircraft.

In the occurrence event, the captain’s decision to explain the process for programming the radial waypoint intercept to the PF meant that both flight crew diverted their attention away from other flight deck tasks. While this was necessary to ensure the data entered was accurate, this distraction resulted in the flight crew’s reduced performance in effectively monitoring the aircraft’s descent profile shortly after commencing descent on the visual approach.

Workload

Workload has been defined as ‘reflecting the interaction between a specific individual and the demands imposed by a particular task. Workload represents the cost incurred by the human operator in achieving a particular level of performance’ (Orlady & Orlady, 1999, p.203). A discussion of the effect of workload on the completion of a task requires an understanding of an individual’s strategies for managing tasks.

An individual has a finite set of mental resources they can assign to a set of tasks (for example, performing a take-off). These resources can change given the individual’s experience, training, and the level of stress and fatigue being experienced at the time. An individual will seek to perform at an optimum level of workload by balancing the demands of their tasks. When workload is low, the individual will seek to take on tasks. When workload becomes excessive the individual must, as a result of their finite mental resources, shed tasks.

An individual can shed tasks in an efficient manner by eliminating performance on low-priority tasks. Alternately, they can shed tasks in an inefficient fashion by abandoning tasks that should be performed. Tasks make demands on an individual’s resources through the mental and physical requirements of the task, temporal demands and the wish to achieve performance goals (Hart & Staveland, 1988; Lee & Liu, 2003).

The flight crew reported that they felt their workload increased once they received vectors to the north following the SANDR waypoint. They stated that the approach requirement changed from the modified route, SANDR direct ROCKDALE (ROC) approach that they had programmed into the FMGS, and this increased their workload.

The PF reported that the vector changes put them in a position where they had not been able to review the control area steps. The PF felt that by the time they had been vectored onto the final intercept heading and cleared for a visual approach, he was behind the aircraft. The increase in workload, combined with the distraction caused by both crew being involved in the re‑programming of the FMGS, decreased the flight crew’s ability to monitor and correctly assess the aircraft’s descent profile.

Flight path monitoring

Monitoring has been very broadly defined by the Flight Safety Foundation (2014) as: ‘adequately watching, observing, keeping track of, or cross-checking.’ (p.3) It has been more fully defined by the United Kingdom Civil Aviation Authority (UK CAA) (2013) as:

The observation and interpretation of the flight path data, configuration status, automation modes and on-board systems appropriate to the phase of flight. It involves a cognitive comparison against the expected values, modes and procedures. It also includes observation of the other crew member and timely intervention in the event of deviation. (p.9)

Monitoring is an extensive set of behavioural skills that all flight crew members are expected to have. This skill set is outlined in the aircraft operator’s standard operating procedures. It involves the primary roles of monitoring the aircraft’s flight path, communications, and the activities of the pilot flying. The difficulties that flight crew can have with maintaining effective monitoring is due to the difficulties in sustaining vigilance. Vigilance decreases as interaction with a system decreases. Therefore, as the pilot monitoring is not directly controlling the system being monitored, it can be harder for them to stay alert to changes. Flight crew members rarely receive direct feedback on the effectiveness or consistency of their monitoring, unlike the feedback they get by flying an aircraft manually.

In a study conducted to identify issues with flight crew checklist use and monitoring behaviour, researchers found that monitoring deviations were grouped into three clusters: late or omitted callouts, omitted verification and not monitoring aircraft state or position (Dismukes & Berman, 2010). In failing to monitor the aircraft state or position, the researchers found that most instances resulted from competing concurrent task demands on the crew’s attention. This leaves an individual vulnerable to losing track of the status of one task while being engaged in another. Crew are taught workload management in crew resource management training but this tends to focus on priorities and distributing the workload amongst crew members and not on how to manage attention when juggling concurrent task demands.

The captain stated that once the flight crew had gone ‘heads in’ to re-program the FMGC, the aircraft went below the descent profile. After programming the new intercept waypoint into the FMGC, the captain looked outside and saw the aircraft was lower than expected. The captain reported that their mental model of where the aircraft should be at the time was not consistent with where the aircraft was. The captain stated that had they been more aware of the aircraft’s position, they might have recognised they were below the profile. It took a little longer for the flight crew to determine they were low on profile, and it was not until the ATC altitude alert was issued that they commenced the climb.

Due to both flight crew’s attention being diverted from the monitoring task after commencing descent for the visual approach, their ability to detect the aircraft’s descent below profile became adversely affected. This resulted in the aircraft descending below profile altitude and entering the 500 ft vertical buffer at the base of the CTA step, reducing separation with terrain and any aircraft operating outside controlled airspace.

Mental models and perception

Researchers have stated that an individual’s mental models are representations of the world based on the individual’s knowledge and built on sensation and perception (Johnson-Laird, 2010; Wickens & Flach, 1988). Sensation is the human sensory system’s ability to detect or determine changes to the sensory inputs picked up by our sensory channels (visual, auditory, haptic, etc.). Perception occurs by assigning meaning to these sensory inputs, and the transfer of this information from perception to working memory/reasoning faculties is controlled by our attentional processes. Therefore, sensation is a passive process and perception is active in that an individual will select, organise and interpret data from the senses. Mental models are shaped by perceptions and vice versa – the process is dynamic and cyclic.

An individual’s ability to gather information relating to their current environment and task is critically influenced by the state of the individual’s knowledge or the mental model constructed (Wickens & Flach, 1988). Individuals will use mental models to reason and infer what will happen next in their world and what actions they need to take in order to get an optimal outcome based on their understanding of how the world works within the current context (Johnson-Laird, 2010; Staggers & Norcio, 1993).

Individuals can make incorrect inferences if there is a mismatch between what is sensed and the meaning given to it through the perceptual processes. In selection or placement of perceptual attention, experienced operators generally rely on a strategy of efficient sampling of reliable information sources (Wickens & Flach, 1988).

Following the vectoring to the north, the flight crew’s workload increased and they became distracted by both going ‘heads in’ to re-programme the FMGS. At that time, they did not get the opportunity to re-set their mental model of how the approach should proceed and they prepared the aircraft for landing too early. This is evidenced by the crew commencing descent early, selecting multiple vertical speeds which were not appropriate for their position, and configuring the aircraft for final approach.

Further evidence of the crew having an incorrect mental model of the aircraft’s position is in the captain’s statement that, on looking up from the FMGS re-programming task, it took 30 seconds for them to realise that the aircraft was not where they had expected it to be. This realisation was coincident with the ATC safety alert to check their altitude.

Summary

The aircraft continued descent below the recommended visual approach profile and entered the 500 ft vertical buffer at the base of the CTA step, reducing separation with terrain and any aircraft operating outside controlled airspace. The ability of the flight crew to assess the aircraft’s relative position accurately and manage the flight path profile was reduced after being vectored off the pre-programmed shortened route. The subsequent involvement of both flight crew to reprogram the FMGS and the conduct of various flight deck activities created a distraction. This increased workload, and distracted the crew from the primary task of monitoring, assessing, and managing the aircraft’s approach path.

Flight crew are reminded that descending near the terminal area during a visual approach using selected flight modes requires vigilance in flight path monitoring. This is especially the case when lower altitude constraints/limits are no longer available with a selected flight mode, and the aircraft is navigated in an unplanned area, off a pre-determined route.

Findings

From the evidence available, the following findings are made with respect to the flight path management and descent toward the lower limit of controlled airspace involving Airbus A320, registered VH‑VND and operated by Tiger Airways Australia Pty Limited (Tigerair). The incident occurred about 9 NM (17 km) north of Melbourne Airport, Victoria on 11 February 2015. These findings should not be read as apportioning blame or liability to any particular organisation or individual.

Contributing factors

• The flight crew's mental model of the aircraft's position relative to the control area steps and terrain during the vectors north of the pre-briefed track-shortened arrival route was not consistent with what was flown. This affected the flight crew's ability to recognise they were below the required altitude.
• The flight crew’s attention was diverted from the required task of flight path monitoring due to the re-programming of the Flight Management and Guidance System (FMGS) during a visual approach. This increased their workload and reduced their ability to detect that the aircraft had descended toward the lower limit of controlled airspace.
• Demonstration and discussion of flight mode reversions that occurred after re-programming the Flight Management Guidance System (FMGS) reduced the flight crew's ability to calculate and manage the aircraft's descent.

Other findings

• The Melbourne tower controller issued a safety alert to the flight crew after they had already levelled the aircraft, prompting them to climb the aircraft back to profile altitude, which re‑established terrain and traffic separation assurance.

__________

12. National Transportation Safety Board (2010). Loss of control on approach, Colgan Air, Inc., operating as Continental Connection Flight 3407, Bombardier DHC-8-400, N200WQ, Clarence Center, New York, February 12, 2009. NTSB/AAR-10/01. Washington, DC. National Transportation Safety Board (2007). Attempted takeoff from wrong runway, Comair Flight 5191, Bombardier CL-600-2B19, N431CA, Lexington, Kentucky, August 27, 2006. NTSB/AAR-07/05. Washington, DC.

About this report Decisions regarding whether to conduct an investigation, and the scope of an investigation, are based on many factors, including the level of safety benefit likely to be obtained from an investigation. For this occurrence, a limited-scope, fact-gathering investigation was conducted in order to produce a short summary report, and allow for greater industry awareness of potential safety issues and possible safety actions. Publishing information Released in accordance with section 25 of the Transport Safety Investigation Act 2003 Published by: Australian Transport Safety Bureau © Commonwealth of Australia 2018 Ownership of intellectual property rights in this publication Unless otherwise noted, copyright (and any other intellectual property rights, if any) in this report publication is owned by the Commonwealth of Australia. Creative Commons licence With the exception of the Coat of Arms, ATSB logo, and photos and graphics in which a third party holds copyright, this publication is licensed under a Creative Commons Attribution 3.0 Australia licence. Creative Commons Attribution 3.0 Australia Licence is a standard form licence agreement that allows you to copy, distribute, transmit and adapt this publication provided that you attribute the work. The ATSB’s preference is that you attribute this publication (and any material sourced from it) using the following wording: Source: Australian Transport Safety Bureau Copyright in material obtained from other agencies, private individuals or organisations, belongs to those agencies, individuals or organisations. Where you wish to use their material, you will need to contact them directly.

On 2 January 2015 a Schempp-Hirth Nimbus 2 glider, registered VH-GOV, collided with terrain near Benalla, Victoria. The pilot, the sole occupant, was fatally injured.

The Gliding Federation of Australia (GFA) requested the technical assistance from the ATSB in the recovery of flight data from a damaged data-logging device to assist with the investigation into the occurrence. To facilitate this assistance and protect the information received from the GFA, an external investigation was initiated under the provisions of the Australian Transport Safety Investigation Act 2003.

The ATSB successfully downloaded the damaged data-logging device on 23 January 2015 and provided the GFA with a copy of the data.

Further information may be obtained by contacting the Gliding Federation of Australia.