At midnight[1] on 10 December 2017, the multi-purpose, general cargo ship BBC Xingang (Figure 1) berthed at Mayfield number four berth in Newcastle, New South Wales. The cargo to be discharged comprised containers and ‘project cargo’, which consisted of large structures of various shapes and sizes. Some of the cargo had gel-coated surfaces that could be easily damaged, and these components were covered with cloth for protection during transport.
Figure 1: BBC Xingang
Source: Tropic Marine Images, Shipspotting.com
Sea fastenings (or stoppers) were used to secure the project cargo during sea passage. The stoppers were metal brackets welded to the ship’s deck, and they needed to be removed before the cargo could be discharged. To remove the stoppers, oxy-acetylene cutting and gouging techniques were required.
In preparation for removing the stoppers, the chief mate and deck crew had reviewed the ship’s risk assessment for hot work and completed the ship’s hot work permit. The master had notified Newcastle port of the intended work and a hot work audit report had been completed by a port representative.
At about 0600 on 11 December, a work team, including a site supervisor and a boilermaker, from a local engineering firm (Varley) boarded the ship. This team had a current Varley Job Safety and Environmental Analysis (JSEA) for hot work to remove stoppers on board ship and had completed a Varley hot work permit.
At 0630, the chief mate convened a safety meeting involving the Varley boilermaker and relevant crewmembers. The process of removing the stoppers was discussed along with the safety measures and procedures to be followed. Hot work requirements, including the hot work permits and risk assessments, were reviewed. Individual duties, including that of the fire watch, were explained and assigned. The ship’s crew were responsible for providing firefighting equipment (such as extinguishers and hoses) and controls such as the fire watch.
The work commenced in the number two cargo hold tween deck.[2] In preparation for the hot work, gaps between the tween deck pontoons were filled and covered with fire blankets (made from woven fibre and leather) to stop sparks from falling onto the cloth-covered cargo in the lower hold (Figure 2).
Figure 2: Worksite showing stoppers, after the incident
Source: Briese Schiffahrts annotations by ATSB
The fire watch consisted of two crewmembers in contact via VHF radio, one on the tween deck and the other in the lower hold. A small diameter fresh water hose was laid out and a makeshift water spray extinguisher readied for immediate use in the lower hold. Cargo in the lower hold was covered with the transport cloth but was not covered with fire blankets.
At 1015, the boilermaker began removing the stoppers. The work site was overseen by ship’s crewmembers, the port captain[3] and the Varley site supervisor. The port captain asked that sparks be directed away from the cargo to protect the component surfaces. This request, in some cases, resulted in the sparks being directed toward gaps between the tween deck pontoons.
The work continued as expected and, at 1100, the boilermaker stopped to relocate to the next stoppers on the tween deck. As part of checking the new work area, he lifted a fire blanket and could see small flames and smoke in the lower hold through the gap in the tween deck pontoons. He immediately raised the alarm. The lower hold fire watch was notified on the radio. At the time he was notified he was not near the area directly under where the work was being conducted. After moving to the relevant area, he quickly extinguished the fire using the water hose and water spray. Figure 3 shows the location of the work site and the fire in the ship’s hold (elevation view and plan view).
Figure 3: Number 2 cargo hold, elevation and plan views of worksite
Source: Briese Schiffahrts with annotations by ATSB
An inspection of the work site following the fire identified that molten metal and other hot material produced by the hot work had burned through the fire blankets. This hot material fell onto the material covering the cargo in the lower hold, resulting in the fire. As a result of the fire, the material covering the cargo was damaged and some surface blemishes were apparent on the cargo itself (Figure 4). No other damage was reported. Subsequent inspection of the cargo covering material found it to be 100 per cent polyester transport cloth with a maximum rated temperature of 200 °C.
Following the fire, all work to remove the stoppers was suspended and the incident was reviewed. Procedures and safety measures were reassessed during a meeting of ship’s crew and the shore contract team. The cutting procedure was amended to direct the sparks away from deck openings and containment of sparks around the work area was improved. Further, additional filling material was added to holes in the tween decks used for removable hand rails, and the lower hold fire watch was directed to monitor the area directly below the work rather than the general vicinity. Work resumed at 1315 and continued to completion without further incident.
Figure 4: Cargo covering material and cargo surface damage
Source: Briese Schiffahrts
BBC Xingang
BBC Xingang is a multi-purpose, heavy lift, general cargo ship registered in Antigua and Barbuda. The ship was built in 2013 by Tianjin Xingang Shipbuilding Heavy Industry, China and classed with DNV GL.
The ship has an overall length of 125.8 m, a moulded breadth of 22.0 m and a deadweight of 8,970 t at its summer draught of 7.60 m. The multi-purpose ship can carry containers and/or general cargo. It has two cargo holds with tween decks. Two 350 t capacity cranes are mounted on the port side and can be used in combination to lift loads up to 700 t.
At the time of the incident, BBC Xingang was owned by Briese Schiffahrts ‘Hatshausen’, operated by BBC Chartering & Logistics and managed by Briese Schiffahrts, all of Germany.
Previous incidents
The ATSB has, on two previous occasions, investigated fires on board Briese Schiffahrts managed vessels in Australian ports.[4] On both occasions the fire was started during hot work to remove sea fastenings. In the BBC Baltic fire, flammable cargo coverings were also found to have ignited. The damage caused was considerably more than that sustained on BBC Xingang.
As a consequence of the previous incidents, Briese Schiffahrts had completed significant safety actions, which were in place at the time of the current fire. These included improved shipboard procedures, risk assessments, permits to work, safety meetings and provision of equipment such as fire blankets and mats.
Safety analysis
The task of removing sea fastenings, originally welded in place to prevent movement of the cargo during the sea voyage, involved hot work, cutting and gouging techniques. This hot work produced sparks, globules of molten steel of various sizes and other hot material. To protect the sensitive surfaces of the cargo, the debris produced by the hot work was directed away from the cargo and more toward gaps/holes in tween deck pontoons.
The work area and the gaps/holes in the pontoons were protected by fire blankets. However, the molten metal produced was of sufficient size and intensity to burn through the protective fire blankets and fall onto the cargo stowed in the deck below. This cargo was covered with transport cloth to protect it during transport but was not protected from the hot debris by covering with fire blankets. Transport cloth is made of 100 per cent polyester and will ignite if directly exposed to fire. As a consequence, when the hot material fell onto the unprotected transport cloth it resulted in the fire.
A fire watch was present in the lower hold but had not been directed to closely monitor immediately below the work site. Therefore, the fire watch was not in position to immediately react when the molten metal fell from above. It was only when the flames and smoke were observed by the boilermaker that the response was initiated and the fire extinguished.
Despite the safety actions taken after previous fires, flammable cargo coverings were again ignited during sea fastening removal on this occasion. Careful consideration must be given to the use of such materials. Prior to hot work, cargo coverings should be carefully assessed and adequate protection against damage or fire due to hot material should be provided.
Findings
These findings should not be read as apportioning blame or liability to any particular organisation or individual.
The fire on BBC Xingang started during the oxy-acetylene cutting and gouging removal of sea fastenings. Molten metal and other hot material produced by this work burned through protective fire blankets in place around the site, and fell onto unprotected cargo below.
The cargo stowed below the work site was covered with flammable polyester material which had not been adequately identified and protected prior to the work commencing.
A fire watch was present in the lower hold but had not been directed to closely monitor immediately below the work site so was not in position to quickly react when the molten metal fell from above.
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 actions in response to this occurrence.
Briese Schiffahrts
On board BBC Xingang, replacement fire blankets were obtained and a safety meeting convened to discuss the incident. This meeting, attended by all ship’s crewmembers, discussed the circumstances of the incident, reviewed relevant existing fleet circulars, reviewed the risk assessment and went through procedures and processes relating to this work and the role of the fire watch.
Varley
Varley advised the ATSB that it had:
reviewed and updated the Job Safety and Environmental Analysis (JSEA) to implement added control measures for hot work
amended work procedures to:
improve coordination and communication between vessel management and third party contractors regarding expectations of fire watchers and firefighting equipment prior to starting work
ensure inspections are conducted on all cargo and materials in lower holds, prior to starting any hot work in the tween decks
ensure that if suspect items are found in the lower hold that communications with work teams and site management include focus on safety expectations and hot work controls
include using heat proof tape to cover gaps in the tween decks with fire blankets added on top
use fire blankets to protect items in the lower hold.
BBC Chartering
BBC Chartering advised the ATSB of the following safety action:
gaps/holes in the tween decks will be covered by fire resistant tape in addition to other measures such as using fire blankets to protect areas below hot work.
Vestas
Vestas advised the ATSB that as part of cargo handling guidance, storyboards explaining safety measures to be taken while welding, cutting and grinding are shared with vessel owners, carriers, and freight forwarders contracted for the carriage of Vestas cargo.
Safety message
Ship fires due to hot work to remove sea fastenings are a constant danger. The continuing incidence of hot work related fires during the removal of sea fastenings highlights the importance of maintaining vigilance throughout the entire process. This is especially important if this is a regular task and is at risk of becoming routine. Procedures and practices along with the equipment available for completion of the task need to be reviewed and assessed and used as appropriate.
Hot work requires the implementation of comprehensive risk controls and procedures. These should include, but not be limited to, detailed, task-specific appraisals, risk and hazard assessments, work permits and pre-work toolbox meetings. Ultimately, the responsibility for the implementation of these controls rests with the ship, in consultation with third parties if involved. This is especially important when shore labour is employed to complete the work and multiple organisations’ work requirements and procedures are involved.
Publishing information
Released in accordance with section 25 of the Transport Safety Investigation Act 2003
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 12 December 2017, a fractured bogie was identified on Pacific National train TM78A during a roll-by inspection at the Kooragang Coal Terminal, New South Wales. The fracture was on bogie NDCA 1199 of wagon NHBH 42954J. This wagon operated loaded between Tahmoor Colliery and Kooragang prior to identifying the fractured bogie.
What the ATSB found
The ATSB found that a fatigue crack went undetected during preventative maintenance activities prior to the structural failure of the bogie frame. It is probable that the fracture was visible at the time of unloading at Newcastle Coal Infrastructure Group on 11 December, which went undetected. The fractured bogie was identified during a roll-by inspection on 12 December, likely preventing a derailment.
What's been done as a result
Pacific National made changes to the maintenance standards used during scheduled maintenance, to increase the area of the bogie frame subjected to non-destructive testing. These changes are aimed at identifying and addressing fatigue cracking prior to the escalation of the defect.
Safety message
Asset managers should ensure that inspection techniques effectively monitor and report on asset condition. Risk controls should also be continuously assessed to control risk to an acceptable level through the life cycle of the asset.
The occurrence
What happened
On the morning of 12 December 2017, train TM78A operated by Pacific National (PN) was unloading coal at the Kooragang Coal Terminal (KCT).[1] At the time, a maintenance worker was conducting a roll-by inspection[2] after the train moved through the unloading bay. The worker visually identified a crack on the lead bogie of wagon NHBH 42954J, this was the 35th wagon in the direction of travel. The train was stopped and the wagon was removed to allow for inspection.
The train operated between KCT to Clarence on the 10 December, before returning to Newcastle Infrastructure Coal Group (NCIG). On 11 December the train operated empty from NCIG to Tahmoor before returning loaded to KCT (Figure 1). During these journeys, the trackside condition based monitoring equipment recorded elevated bearing temperatures on axle 4 of wagon NHBH 42954J, as well as an increased angle of attack (AoA).[3] These recordings did not generate an alert at the time of passing the various detection sites and the train continued to operate until the fractured bogie was visually identified by the maintenance worker.
Figure 1: Incident location
Source: Geoscience Australia, annotated by OTSI
Safety analysis
A PN maintenance worker was located inside the unloading bay at NCIG on 11 December, this worker’s primary function was to observe the wagon doors close. It was not clear if this worker was deemed to additionally be conducting a roll-by inspection. The worker was positioned on the opposite side (to the bogie fracture) of wagon NHBH 42954J and did not report any defects. The bogie was probably visibly fractured at the time based on the increased AoA.
The roll-by inspection completed on 12 December was completed by a single worker, with the worker positioned on the same side as the fractured bogie and the defect was identified. The axle showed signs of rotating forward and up relative to the frame as the fracture opened under the weight of the wagon (Figure 2). As the fracture opened the effective axle spacing changed, lengthening the wheelbase on the side of the cracked frame. PN maintenance standards permit roll-by inspections being completed by a single worker. The defect was detected during a roll-by inspection, however could have been missed if the worker was positioned on the opposite side.
Figure 2: Fractured bogie frame NDCA 1199
Source: Pacific National, annotated by OTSI
The train underwent unit train maintenance (UTM) and a full train examination (FX) on 8 December 2017 at the PN Lithgow maintenance depot. During this inspection, there were no defects reported in relation to bogie NDCA 1199. Prior to identifying the fractured bogie, the train was compliant with the PN technical maintenance plan.
Bogie NDCA 1199 underwent schedule maintenance in July 2015. The bogie frame was subjected to non-destructive testing (NDT) consisting of magnetic particle inspection (MPI) and tested in accordance with PN procedure WMM 11-08_06 One Piece Bogie Inspection. There were no defects detected at the time of testing and the location of the fracture was at the extremity of the area subjected to NDT. Evidence of remnant MPI marker paint was present on the outboard plate, but there was no apparent marker paint on the inboard plate at the region of the fracture.
Fatigue cracking has historically been identified in the NDCA bogies at the welded junction joining the vertical plate and bottom plate section due to the design and manufacturing process used to build this class of bogie at the time. The NDCA bogie is comprised of welded steel plates to form a single piece box section bogie frame.
Prior to this occurrence, a derailment occurred in 2011 at Leigh Creek, South Australia. This derailment was reportedly the result of a fractured NDCA bogie. The bogie fracture increased the wheelbase on the side of the fractures, likely placing additional loading on the axle bearing as the axle tracked abnormally, this led to the bearing overheating and failing.
Post-occurrence inspection of the bogie NDCA1199 showed a fatigue crack most likely initiated at the toe of the weld on the inboard corner of the bogie side frame before progressing across the lower plate and inboard vertical plate (Figure 3). The inboard fracture face was smooth and displayed beach marks (slow progression), while the outboard face was jagged in appearance indicative of rapid progression.
Figure 3: NDCA 1199 fracture analysis
The image shows the bogie fracture from the underside of the bogie to show the three fracture faces. Source: OTSI
The trackside condition based monitoring equipment recorded increased bearing temperatures from 10 December after departing Clarence, as well as an increased AoA on 11 December. Analysis of this recorded data following the event, indicates that the change was likely the result of the bogie frame fracture progressing. As the fracture progressed, the wheelbase (spacing between the axles) increased between axle 3 and 4, affecting the AoA and placing additional loading on the bearing.
At the time of passing through the various trackside condition based monitoring sites, the recorded bearing temperatures did not trigger an automatic alarm to notify the rail infrastructure manager (RIM) or PN. The track side detection equipment utilises algorithms to determine critical levels[4] that trigger automatic alarms to advise the RIM or operator. In relation to bearing temperature, only warm or hot bearing alarms will trigger an automatic alarm that requires immediate attention. Low bearing temperature alerts and AoA are recorded by the system but do not generate automatic alarms.
The wagon loading for the previous journeys was reviewed and indicated, that between 10 December and 12 December, the wagon operated within the allowable 100 t gross limit. The operation of the train and wagon loading does not appear to have contributed to this occurrence.
Findings
These findings should not be read as apportioning blame or liability to any particular organisation or individual.
A fatigue crack likely originated at the toe of a weld on the inboard lower plate of the bogie frame that was not identified during the preventative maintenance activities, prior to the structural failure of the frame. The fatigue cracking most likely progressed across the full face of the lower plate and inboard side plate before rapidly progressing across the outboard side plate.
The NDCA bogie design has a history of fatigue cracking which increases the risk of derailment if the cracking is not identified. Pacific National had processes in places to identify fatigue cracking but this crack was likely not detected due to the location of the defect.
Trackside condition based monitoring equipment recorded elevated bearing temperature and increased angle of attack in the days prior to the fractured bogie being identified. This was likely a result of the bogie fracture progressing. The recorded temperature did not trigger an automatic alert to advise the rail infrastructure manager or operator.
Safety action
The ATSB has been advised of the following safety action in response to this occurrence.
Pacific National
As a result of this occurrence, Pacific National has advised the ATSB that they have taken the following safety actions:
Bogie Survey
Pacific National undertook a visual inspection of all NDCA bogie frames for evidence of cracking as directed by Office of the National Rail Safety Regulator (ONRSR).
Revised Maintenance Standard
Pacific National revised maintenance standard WMM 11-08_06 One Piece Bogie Inspection to increase the area subjected to NDT testing during schedule overhauls. The entire side frame is now tested on the inboard and outboard plates as well as the underside of the bogie frame through the horn cheek area.
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
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 6 December 2017, at about 0800 Central Standard Time,[1] a Cessna U206G (C206) aircraft, registered VH-HPA (HPA) and operated by Hardy Aviation, taxied at Darwin Airport, Northern Territory (NT). A Cessna 210L (C210) aircraft, registered VH-SYT (SYT) and operated by Chartair, taxied not far behind HPA.
The pilots of the aircraft were both operating charter flights to Port Keats, NT, under the visual flight rules (VFR),[2] and had planned to track at 8,500 ft. Their initial clearance, in accordance with the traffic management plan for Darwin, was to track via VFR route 5, which tracked south for 5 NM then south-west (Figure 1). Both aircraft were taxiing to depart from runway 18. Prior to take-off, the pilot of each aircraft was advised by air traffic control[3] that the other aircraft would also be tracking to Port Keats at 8,500 ft.
Figure 1: Extract of Darwin Visual Terminal Chart showing VFR Route 5 and direct track to Port Keats
Source: Airservices – annotated by ATSB
The pilot of HPA, with two passengers on board, observed SYT taxiing and saw only the pilot on board. He assumed, therefore, that SYT would probably be light and expected it would quickly overtake and out-climb HPA.
Shortly after take-off, the pilot of HPA contacted the Darwin approach controller and reported passing 800 ft on climb to 2,000 ft. The Australian Defence Air Traffic System (ADATS) controller’s situation display indicated the aircraft’s altitude as 700 ft at that time, based on the transmission from the aircraft’s mode C transponder (see the section titled Radar altitude). The controller then cleared HPA to track direct to Port Keats at 8,500 ft.
The pilot of SYT was cleared for take-off about 1 minute after HPA. Once airborne, he immediately looked for, and sighted HPA. Just before 0808, the pilot of SYT contacted the approach controller and reported passing 600 ft on climb to 2,000 ft, with HPA in sight (Figure 2). By the end of that transmission, the controller’s situation display radar altitude for SYT indicated 100 ft. The displayed level information differed from the pilot’s reported altitude by more than the permitted tolerance of 200 ft (see the section titled Radar altitude) but there was no indication that the controller identified that discrepancy.
Figure 2: Google Earth image overlaid with aircraft tracks
Source: Google Earth and radar data – annotated by ATSB
In response to the transmission by the pilot of SYT, the controller advised that the preceding traffic (HPA) was now tracking direct to Port Keats at 8,500 ft, and then asked whether direct tracking was also being sought for SYT. The pilot of SYT responded ‘affirm, when available’. The controller advised him to expect that clearance in 1 minute.
The approach controller reported that SYT was left on VFR route 5 for about 1 minute to segregate it from HPA, before being re-cleared to track direct to Port Keats at 8,500 ft. The pilot of SYT reported that the direct track to Port Keats put SYT on a similar track to HPA. Therefore, he kept HPA in sight, and manoeuvred to the right of its track, expecting to overtake it.
The pilot of HPA expected that SYT would be either just to the left (based on the VFR route’s initial southerly track compared to the direct track to the south-west) or directly behind HPA. At 0812, the pilot of HPA requested (and received) a clearance to stop climb at 6,500 ft ‘due to faster following traffic’. This transmission was intended to make the pilot of SYT aware that he knew SYT would be close behind, and for vertical separation between the two aircraft.
Three minutes later, the approach controller asked the pilot of SYT if he still had the preceding traffic in sight, and the pilot responded ‘affirm’. The controller then advised him that HPA was now on climb to the amended level of 6,500 ft.
As SYT approached 5,500 ft, the pilot of SYT could see his aircraft was gaining on HPA, and assessed it would soon overtake HPA. Shortly after, HPA moved through the 10 o’clock position[4] of the pilot of SYT who then lost sight of it behind SYT’s left wing.
At 0816, when about 15 NM south-west of Darwin Airport, the pilot of SYT reported to the approach controller that he was ‘just coming up on HPA’s 3 o’clock position and lost sight’ of that aircraft. The controller advised that HPA was ‘climbing through 6,200 [ft] at the moment, probably half a [nautical] mile to your left’. The pilot of SYT responded ‘Sierra Yankee Tango’, and did not sight HPA.
The controller then gave HPA traffic information: ‘SYT is on your right 3 o’clock, half a [nautical] mile, climbing through 5,500 [ft]’. The pilot of HPA responded that he was ‘looking’. The pilot of HPA looked to his right, but did not see SYT.
The radar data at that time indicated lateral separation between the aircraft was 0.27 NM (500 m), with SYT behind and slightly right of HPA and 800 ft below. About 30 seconds later, an ADATS conflict alert (CA)[5] activated on the controller’s situation display (SYT indicated 5,600 ft and HPA 6,400 ft). When HPA reached 6,500 ft, the pilot levelled the aircraft off and completed the top of climb checks.
At 0817, the approach controller asked the pilot of SYT ‘Do you have that traffic in sight or would you prefer a different level or tracking?’ The pilot of SYT responded ‘Negative, traffic not in sight… happy to maintain this track if HPA has us in sight’. The controller then asked the pilot of HPA if they had SYT in sight. The pilot of HPA responded ‘Negative’ and advised that HPA was now maintaining 6,500 ft.
The pilot of SYT reported that, at that time, he still thought the aircraft were half a nautical mile (900 m) apart and was expecting HPA to be out to his left. However, the radar data indicated that HPA was 800 ft vertically above, and 0.18 NM (333 m) laterally away from SYT.
At 0818 the CA activated again while neither pilot had the other aircraft in sight. The radar data at the time indicated that SYT was at 5,900 ft, HPA at 6,500 ft with a lateral separation of 0.13 NM (241 m).
The approach supervisor reported that, at that stage, there was still just over 500 ft vertically between the two aircraft and that he instructed the approach controller to maintain 500 ft separation between the aircraft.
In response to the direction from the supervisor, the controller asked the pilot of SYT if he would ‘like to stop climb 6,000 [ft]?’, to which he responded ‘Negative, I’ll continue on to 8,500 [ft], happy to take 3 [nautical] miles right of track if that puts us out of conflict’. The pilot of SYT reported that he was puzzled by the question because at that stage his altimeter was indicating an altitude of 6,300-6,400 ft. The radar data at the time showed SYT at 6,000 ft, HPA at 6,500 ft and 0.086 NM (160 m) lateral separation between the aircraft, which were then about 20 NM south-west of Darwin Airport.
The pilot of SYT commented that as soon as he made that request to deviate three nautical miles right of track, HPA came from the left top corner of his windshield across the nose to the bottom right in front of him and filled the windscreen. He estimated the two aircraft passed 3‑4 m apart.
The pilot of HPA was looking out to his 3 o’clock position when the pilot of SYT was in the process of requesting the deviation right of track. SYT appeared in the pilot of HPA’s 5 o’clock position and he reported being surprised by its close proximity. He and the passengers in HPA estimated the aircraft came within 5 m of each other.
The controller then cleared SYT to deviate up to three nautical miles right of track and advised that ‘HPA by radar is on top of you 6,500 [ft]’.
Both pilots reported that by the time the approach controller cleared SYT to deviate right of track, the two aircraft had already passed each other and the separation between them was increasing. SYT was now to the left of HPA and the pilot of SYT had HPA in sight. The closest proximity according to the radar data was 100 ft vertically and 0.02 NM (37 m) laterally.
The approach controller reported that the CA activated again and SYT did not seem to be deviating to the right, just continuing to climb. The controller observed SYT passing 6,200 ft on the situation display and, as it was now less than 500 ft below HPA, they issued a safety alert. The approach supervisor reported also directing the controller to issue a safety alert at that time.
After the aircraft had already passed each other in close proximity, the pilot of SYT responded ‘Say again’. The approach controller then responded ‘Safety alert: HPA by radar is on top of you 6,500 [ft]. Deviate up to 3 [nautical] miles right of track and maintain 6,000 [ft] until clear’.[6]
The approach controller recalled assessing that SYT might be required to descend to the cleared altitude of 6,000 ft at the time the instruction was issued. In response to that transmission, the pilot of SYT advised that HPA was now to the right of their track, so SYT would stay to the left.
The two aircraft continued to Port Keats without further incident.
VFR aircraft in Class C airspace
In Class C airspace, VFR aircraft are provided with traffic information on other VFR aircraft and are not separated by air traffic control (ATC). The approach supervisor commented that there was no responsibility for ATC to provide separation between those two aircraft. However, they were entitled to a traffic service and a safety alert if the controllers assessed there was a problem.
Civil Aviation Regulation (CAR) 163 stated that ‘The pilot in command of an aircraft must not fly the aircraft so close to another aircraft as to create a collision hazard.’ In addition, CAR 163A states that ‘the flight crew of an aircraft must… maintain vigilance so as to see, and avoid, other aircraft.’
The approach supervisor commented that they had an expectation that the pilots of the involved aircraft would have had an understanding of the relative performance of each other’s aircraft and that they would maintain visual contact and be able to keep clear of each other.
While pilots of VFR aircraft are responsible for avoiding other VFR aircraft, according to MATS 2.2.1.1, the objectives of Air Traffic Services include to ‘prevent collisions between aircraft.’ The approach supervisor commented that when pilots of VFR aircraft lose sight of each other, the requirements of the safety alert come into place.
Darwin VFR aircraft operations
A large number of VFR aircraft operate charter flights to and from Darwin Airport. The controllers reported that VFR aircraft often tracked in close proximity to each other, although when there were more than two aircraft departing on the same track, ATC usually provided instructions to segregate them. In addition, ATC often separated VFR aircraft from instrument flight rules aircraft by issuing the pilots of the VFR aircraft with tracking instructions.
The controller commented that when multiple aircraft depart together bound for the same location, they are usually from the same operator, but in this incident they were not. He also commented that this might have contributed to the slower aircraft departing ahead of the faster one. If the two aircraft were from the same operator, the pilots may have sequenced themselves so the faster aircraft would depart first. The pilot of SYT said in future, taxiing out, he would contact the ground controller and ask if they could depart ahead of the slower aircraft. In addition, in a similar situation, as soon as possible, he would put his aircraft on a track 5 NM right of the other aircraft.
The approach controller commented that the C206 and C210 were very similarly performing aircraft, and their speeds can vary depending on how many people and how much cargo and fuel is on board, so the type of aircraft is not always a good indication of relative performance.
In accordance with Darwin’s traffic management plan, all VFR traffic are cleared to depart via a published VFR route, except those going to the nearby islands. After departure most aircraft are cleared for direct tracking as soon as practicable. The VFR route 5 tracks 5 NM south of Darwin then south-west, and would have been within a couple of miles left of the direct track. The pilot of SYT had planned to track via VFR route 5 in accordance with their standard operating procedures.
The pilot of HPA was using an electronic flight bag[7] and tracked direct to Port Keats once cleared. The pilot of SYT reported maintaining a constant heading direct to Port Keats after receiving clearance to track direct, and that he ‘was checking the heading continuously’ and the two aircraft were ‘paralleling until he lost sight of HPA’.
Conflict alert
In Class C airspace in Darwin, when aircraft operate within 3 NM horizontally and 1,000 ft vertically, a CA activates. In response, the controller usually announces to the other controllers and supervisor what separation standard is in place – whether it is ‘traffic’, or another standard such as ‘500 ft’ vertical separation. The approach supervisor commented that it was normal for Darwin for the conflict alert to activate between VFR aircraft.
ATSB investigation AO-2011-011 identified a safety issue at Williamtown (Newcastle Airport), New South Wales associated with conflict alerting. The conflict alerting function had been disabled following a risk analysis and advice from safety specialists, due to numerous, unavoidable spurious alarms at Williamtown. During a trial period in which the alerts were activated, a Department of Defence investigation found that the alert function did not assist controllers in the identification or resolution of traffic conflicts and that false alerts may have resulted in controller desensitisation.
Limitations of see‑and‑avoid
The limitations of see-and-avoid practices are well known and documented. Civil Aviation Advisory Publication (CAAP) 166-2(1) Pilots’ responsibility for collision avoidance in the vicinity of non‑controlled aerodromes using ‘see-and-avoid’[8] discusses see-and-avoid in non-controlled airspace, but much of the content is relevant to pilots of all VFR aircraft including in controlled airspace.
Alerted see-and-avoid, where the pilot is directed where to look to sight another aircraft, is much more effective than un-alerted. However, there are still a number of factors that affect a pilot’s ability to sight another aircraft. In this incident, the pilot of SYT lost sight of HPA when it was above and diagonally to his left, as it was then obscured by SYT’s left wing. Additionally, when the pilot advised ATC that he had lost sight of HPA, the controller responded that HPA was half a mile to his left, and then advised the pilot of HPA that SYT was half a nautical mile to his right in his 3 o’clock position. A review of the radar data identified that the relative positions of the aircraft differed from that advised by the controller.
Radar altitude
An aircraft’s transponder operating in mode C transmits a signal that permits a secondary surveillance radar ground station to determine the distance and bearing to the aircraft as well as its altitude.
Aircraft altimeters and ATC situation display show barometric altitude below the transition altitude (10,000 ft in Australia). Aircraft mode C transponders sense and transmit static air pressure to ground‑based surveillance equipment. That equipment converts the value into the corresponding pressure altitude and, if below the transition altitude, applies an input from a ground‑based QNH to obtain the equivalent barometric altitude. The calculated altitude is displayed to controllers with a resolution of 100 ft.
MATS section 9.7.5.5 Display tolerance stated that ‘when the displayed pressure altitude-derived level information differs from the pilot-reported or known altitude by more than 200 ft: a) advise pilot; b) request check of pressure setting; and c) confirm current level.’
When the pilot of HPA reported passing 800 ft, the displayed altitude was 700 ft and therefore within tolerance. However, when the pilot of SYT first contacted the approach controller and reported passing 600 ft, the aircraft’s label on the situation display indicated 100 ft by the end of the pilot’s transmission, which was outside the 200 ft tolerance. The 500 ft discrepancy between SYT’s displayed radar altitude and actual aircraft altitude correlated with the recorded radar data from Airservices Australia while the aircraft was in cruise. There was no indication that the controller detected the discrepancy.
Because of this discrepancy, SYT was about 500 ft higher and therefore about 500 ft closer vertically to HPA than displayed. Allowing for this, the closest proximity between the aircraft around the time of the occurrence was within 100 ft, and this proximity occurred before the controller issued the safety alert. Given the radar is only displayed to the nearest 100 ft, this was consistent with the pilots of both aircraft reporting that the aircraft came within 3–5 m of each other at the same altitude and that they had passed before the safety alert was issued.
Safety analysis
Development of the occurrence
The pilots of the two aircraft were required to see and avoid each other. SYT was trailing, but faster than, HPA. That relative positioning reduced the opportunity for the pilot of HPA to identify the developing proximity event, as SYT was below HPA as the two aircraft converged. As SYT closed on HPA, the pilot of SYT lost visual contact as the wing structure visually obscured HPA. The pilot of SYT identified that his aircraft was about to overtake HPA, but did not manoeuvre to keep HPA in sight or request an alternative clearance. The pilot of SYT lost sight of HPA while the aircraft were about 500 m and 800 ft apart.
Although the pilot of SYT did not manoeuvre to maintain sight of HPA, he did advise ATC that he had lost visual contact with that aircraft. When requested, ATC will provide VFR flights in Class C airspace with a suggested course of action to avoid other VFR flights, but the pilot is still required to see and avoid other aircraft.
As neither pilot requested ATC to provide avoiding action, the controller did not issue alternative segregation instructions at that time and the two aircraft continued to converge. The controller did, however, provide traffic information to both pilots and offered alternative tracking.
The controller issued a safety alert and instructions when the vertical separation depicted on the situation display between the aircraft reduced to about 500 ft. However, due to a combination of radar accuracy/resolution and the altitude of SYT on the situation display not being within the required 200 ft tolerance, the aircraft were much closer than the indicated 500 ft. Consequently, the safety alert was issued after the near collision and so neither pilot had an appreciation of just how close the aircraft were until they re‑sighted each other as they passed in close proximity.
Had the controller verified the initial altitude of SYT on first contact, they would have identified the discrepancy between the reported and displayed altitudes. The controller would then have had an accurate indication of the vertical distance between the two aircraft.
See-and-avoid limitations
Obstruction by the aircraft’s wing is one of many factors identified by ATSB research that affect a pilot’s ability to sight another aircraft. This occurrence therefore highlights the difficulties of the see-and-avoid principle, even when the pilot is given information about (or alerted to) the other aircraft’s position. Airborne collision avoidance systems (ACAS) provide valuable information to alert pilots of other aircraft in their proximity and can direct the pilot to take avoiding action, thereby reducing the risk of collision.
Findings
These findings should not be read as apportioning blame or liability to any particular organisation or individual.
The approach controller did not verify the initial altitude of VH-SYT, which was outside the allowable 200 ft tolerance. That resulted in the two aircraft being significantly vertically closer than displayed and, in turn, the controller issuing a safety alert after the near collision had occurred.
After the pilot of VH-SYT lost sight of VH-HPA, he advised air traffic control and took no further action to ensure segregation between the aircraft.
Safety message
Pilots and air traffic controllers have a joint responsibility to avoid collisions between aircraft. In controlled airspace, air traffic controllers are not required to provide pilots of aircraft operating under the VFR with separation from other VFR aircraft. While air traffic controllers can provide traffic information to pilots of VFR flights, see-and-avoid is the primary means of preventing collisions between VFR aircraft.
The limitations of see-and-avoid techniques are well known and are detailed in the ATSB publication Limitations of the See-and-Avoid Principle. When pilots are alerted to the location of another aircraft, this significantly improves their ability to sight the other aircraft. However, as detailed in the publication, many factors can affect a pilot’s ability to sight another aircraft. The publication was written in 1991 and states that ‘Because of its many limitations, the see-and-avoid concept should not be expected to fulfil a significant role in future air traffic systems.’
Both VH-SYT and VH-HPA were equipped with mode C transponders, but neither aircraft was equipped with any airborne collision avoidance system (ACAS) technology, nor were they required to be by regulation. Such a system provides information to increase a pilot’s awareness of nearby aircraft and therefore reduces the risk of a mid-air collision.
Recent advancement of ACAS technologies has made them viable for general aviation aircraft and they should be considered. The ATSB report (AO-2016-015) into a near collision involving a Saab 340 aircraft and a glider in 2016 outlined a proposal by the industry body, Australian Strategic Air Traffic Management Group, to the Civil Aviation Safety Authority (CASA). It recommended the adoption of standards for ADS‑B technology to be fitted in general aviation aircraft to enable awareness of other aircraft traffic and thereby reduce the risk of mid-air collisions.
Following industry consultation, CASA is proposing to relax the equipment and installation standards for ADS-B fitment in VFR aircraft. The aim is to make it cheaper and easier for aircraft operating under VFR to purchase and use the technology.
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
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 the morning of 1 December 2017, an M-18A Dromader aircraft (Pratt & Whitney PT6A engine), registered VH-WHR (WHR) prepared for an aerial agricultural spraying operation at a private airstrip, 9 km north of Emerald Airport, Queensland. The aircraft was operated by Central Highlands Aerial Services and was conducting a spray operation on an 81-hectare paddock (Figure 1, shown in green), about 6 km west‑south‑west of the airstrip.
Figure 1: Application area (green), accident site (blue) and track flown (yellow)
Source: Google earth, annotated by the ATSB
The pilot conducted a pre-flight inspection of WHR and found everything to be serviceable. He was also the last person to fly the aircraft, three days prior to the accident, and had not identified any problems.
The wind was reported to be 4‑6 kt from the northeast. The spraying operation was standard with no major hazards in or around the paddock ‑ the pilot had sprayed the paddock several times in the past conducting the same spray pattern flying east and west.
At about 0604 Eastern Standard Time,[1] the aircraft arrived at the paddock and the pilot conducted several short spray runs in the south-western corner to ensure there was no overspray onto an adjoining paddock. He then conducted several back-to-back spray runs in the same corner (Figure 1) and at about 0614, he commenced a racetrack pattern over the remainder of the paddock.
At about 0620, a witness located about 1 km from the paddock observed the aircraft complete the ninth racetrack pattern run and commence a turn to the right to line up for the next run. The witness estimated that about three quarters of the way through the turn, while lining up for the next run, the aircraft rapidly pitched down and the right wing collided with the ground. The aircraft subsequently flipped, and came to rest inverted and facing in the opposite direction to the flight path about 20‑30 m from the initial contact point (Figure 2).
Figure 2: Accident site
Source: Police
At about 0622, the witness called emergency services and proceeded to the accident site. When the witness arrived at the wreckage, he found the pilot had exited the aircraft through the broken cockpit side window.
Ambulance officers treated the pilot before transporting him to hospital. He was admitted to treat his injuries, which included a fractured left leg, three fractured left ribs, bruises, cuts to his left side, and concussion. The pilot was wearing a helmet at the time of the accident and it was damaged from impact with the aircraft structure (Figure 3). The pilot had no recollection of the accident and no mechanical issue was identified that may have contributed to the accident.
Figure 3: Damage to the left side of the pilot’s flight helmet
Source: Police
While the track of the aircraft during the spray operation was recorded, other parameters such as airspeed, time, altitude, and aircraft attitude were not. Figure 4 shows the final four racetrack pattern turns back towards the west. The last inbound turn (shown in blue) was conducted at a smaller turn radius than the previous three turns in that direction. The track data finished about 460 m from the accident site.
Figure 4: Final four right turns from an easterly track (last shown in blue)
Source: Google earth, annotated by the ATSB
The pilot joined the operator in January 2015, gaining about 800-flight hours in WHR (with the installed PT6A engine) and flew the previous flight in WHR three days prior to the accident. He did not report any concerns with the aircraft.
The maintenance release for WHR was issued about 40 flight hours prior to the accident and no outstanding maintenance or defects were recorded.
Safety analysis
About three quarters of the way through a turn, as the aircraft was lined up for the next racetrack pattern spray run, the aircraft was observed to rapidly pitch down and collide with the ground. Analysis of the limited available recorded data showed that the final turn was flown at a tighter radius than the previous racetrack pattern turns. In combination, this could indicate that the accident was the result of an aerodynamic stall. However, there was insufficient information to determine if that occurred.
The pilot was unable to remember the final turn and could not provide a reason for the track variation or why the aircraft pitched down. No mechanical defects were noted with the aircraft, on the maintenance release, during the previous flight, or up to the section of the accident flight the pilot could remember. A post‑accident inspection of the aircraft by the operator did not identify any defects. From the limited available information, it was not possible to determine the reason for the accident.
The pilot was wearing his own personal flight helmet at the time of the accident. During the accident, the left side of the helmet struck the internals of the cockpit. Based on the degree of damage to the helmet, it probably prevented the pilot receiving more serious head injuries.
Findings
These findings should not be read as apportioning blame or liability to any particular organisation or individual.
During a turn, and for reasons that could not be determined, VH-WHR pitched down and collided with the ground.
The helmet worn by the pilot probably prevented more serious head injury.
Safety message
The International Civil Aviation Organization circular 85-AN/71 Safety in aerial work Part 1. Agricultural Operations discusses the importance of reducing serious head injuries by wearing a correctly fitting flight helmet. Pilots operating aircraft in agricultural operations are particularly vulnerable to accidents involving major or fatal head injury. The circular also discusses the need to select a helmet which provides effective protection and that is part of the pilot’s personal flying equipment as was the case in this accident.
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
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 24 November 2017, an Airbus A320 aircraft, registered PK-AZE and operated by PT Indonesia AirAsia, departed Perth, Western Australia on a scheduled passenger flight to Denpasar, Indonesia. Shortly after take-off from runway 21, the aircraft turned left, contrary to the cleared standard instrument departure, and at a height of about 223 ft above ground level, which was below the minimum height for turns specified by the operator. Air traffic control (ATC) assigned a series of headings to the flight crew. The aircraft was also turned through one of the headings assigned by ATC. An additional heading was subsequently issued by ATC to return the aircraft back to its planned track. The flight continued to Denpasar without further incident.
What the ATSB found
The ATSB found that during pre-flight preparations, the first officer (FO) as the pilot flying programmed the flight management guidance system (FMGS) using runway 03 based on the assumption that their departure runway would be the same as the runway on which they had previously landed. That programming occurred prior to the FO obtaining the automatic terminal information service (ATIS) information, which stated that runway 21 was in‑use. In addition, the captain did not separately listen to the ATIS as required by the operator’s procedures and likely relied on the briefing conducted by the FO.
Despite the various cues available to the flight crew, including several ATC instructions for using runway 21 and airport signage, and the flight crew reporting feelings of unease about the flight preparations, the incorrect programming of the FMGS was not detected. Shortly after take-off, the aircraft’s flight director indicated a left turn was required. The FO commenced a left turn and engaged the autopilot, which continued the left turn to navigate towards the first programmed waypoint (in-line with the opposite runway direction and behind the aircraft). However, ATC quickly noticed the diversion from their cleared track and corrected the crew’s heading.
Once the crew had detected the error, the captain reprogrammed the correct flight plan in the FMGS, rather than selecting the heading assigned by ATC. This resulted in the aircraft turning through the assigned heading, requiring further instructions by ATC.
It was also established that ATC were unaware of the nature of the problem and were not utilised fully as an additional resource to assist the flight crew.
What's been done as a result
Following the incident, Indonesia AirAsia have included a similar ‘change of departure runway’ scenario in their line operations flight training. Further, they have plans to launch a cross-departmental initiative to increase the awareness and skill sets of pilots, especially in the area of threat and error management.
Safety message
This incident demonstrates that deviating from standard procedures, even slightly, can render them ineffective and result in errors. Data input errors continue to be one of ATSB’s SafetyWatch priorities.
It further highlights the significance of stopping and re-evaluating the situation while on the ground when there is a feeling of uncertainty about the flight, even if it results in undesirable delays. This provides an opportunity to detect errors before they affect operations. Once airborne, workload and time limitations become even more critical due to the rapidly changing situation.
The occurrence
Pre-flight preparation
On 24 November 2017, at about 1145 Western Standard Time,[1] an Airbus A320 aircraft, registered PK-AZE and operated by PT Indonesia AirAsia, was being prepared for a flight from Perth Airport, Western Australia to Denpasar (Bali) Airport, Indonesia. The captain was the pilot monitoring (PM) and the first officer (FO) was the pilot flying (PF).[2] The captain conducted the exterior inspection of the aircraft while the FO entered the flight plan into the flight management guidance system (FMGS) as per the required cockpit preparation procedure.
The crew had operated the aircraft into Perth on the previous sector, and the aircraft had arrived 46 minutes after the scheduled arrival time. The previous flight had landed on runway 03. Believing that they would be using the same runway for take-off, as he had done on previous occasions, the FO entered runway 03 into the FMGS.
The FO then listened to the automatic terminal information service (ATIS),[3] which stated the runway‑in‑use for departure was runway 21, and wrote ‘runway 21’ on the paper flight plan. He then calculated and entered the aircraft’s take-off speeds based on runway 03 into the FMGS. The FO reported he would normally listen to the ATIS first, then input data into the FMGS, but to make it easier on this occasion he programmed the FMGS first and then listened to the ATIS.
When the captain returned to the flight deck after conducting the exterior inspection, he asked the FO which runway would be used for take-off. The FO stated that it was runway 03. The captain reported that he did not listen to the ATIS. However, he cross-checked the information entered in the FMGS, which included comparing the flight plan with the anticipated standard instrument departure (SID)[4] chart.
The FO then conducted the take-off briefing. He later recalled stating that the departure would be from runway 03. The captain also recalled that the FO’s take-off briefing stated that they would be departing from runway 03.
At 1201, the flight crew received their airways clearance from air traffic control (ATC) to depart for Denpasar using the AVNEX TWO SID and to climb to 5,000 ft as per the SID. The AVNEX TWO SID was available for both runway 03 and runway 21, with different initial waypoints before converging (refer to section titled Perth departure procedures).
The FO reported he ‘felt that there might be something wrong’ with the FMGS programming and advised the captain, but they both continued with the preparation.
Taxi to runway 21
At 1213, the flight crew received a taxi clearance from ATC and commenced taxiing (Figure 1). The clearance included a series of instructions to taxi to the holding point for runway 21 on taxiway ‘Whiskey’. The flight crew read back this clearance correctly.
Figure 1: Taxi and take-off path (in red) of PK-AZE
Source: Airservices Australia (modified by the ATSB)
At 1218, the Perth tower controller provided instructions to hold short of runway 21. Shortly after, he cleared the flight crew to line up on runway 21. The flight crew read back both instructions correctly. Prior to reaching the holding point for runway 21, the flight crew received an electronic centralised aircraft monitor (ECAM)[5] message BRAKE SYS 1 FAULT.[6] The captain reported that he actioned the message using the appropriate checklist. After actioning the message, the captain reported feeling there was something wrong, but was unsure of the issue.
At 1220, the controller cleared the aircraft for take-off on runway 21, which the flight crew read back correctly. At that time, the flight crew noticed the runway mode[7] had not activated on the flight mode annunciator.[8] They did not think it was unusual as sometimes the mode did not activate, particularly if the instrument landing system[9] was unavailable.
Take-off from runway 21
After becoming airborne during take-off, both pilots’ navigation displays showed the next waypoint[10] was MIDLA (Figure 2). Both flight crew noticed the flight director[11] on their primary flight display was commanding a left turn. Shortly after, following the flight director, the FO turned the aircraft left and then engaged the autopilot. Recorded flight data showed the aircraft was manually turned at 223 ft above ground level (AGL) (refer to section titled Minimum height of turns).
Figure 2: Flight path of PK-AZE from runway 21
Source: Google Earth, modified by the ATSB
About 20 seconds later, the tower controller instructed the flight crew to turn right onto a heading of 200°. The captain read back the instruction correctly, selected the assigned heading on the flight control unit and engaged heading mode. The autopilot began to bank the aircraft to the right. As this happened, the controller asked the flight crew to confirm the instruction of heading 200°, which the flight crew read back incorrectly as 300°. The tower controller consulted another controller about the aircraft’s track, the issued instruction, and whether to cancel the aircraft’s assigned SID.
About this time, the captain took over PF duties from the FO. The flight crew were instructed to maintain the assigned heading of 200° and contact the departures controller. After the flight crew contacted the departures controller, they were cleared to climb to flight level (FL)[12] 180 and instructed to turn right onto a heading of 270°, which the flight crew acknowledged and selected. Shortly after, the captain gave the PF duties back to the FO and soon began reprogramming waypoints in the FMGS. The FO recalled being unaware the captain was reprograming the FMGS.
At 1225, the controller instructed the flight crew to maintain heading 290°. At this time, the captain selected waypoint SWANN in the FMGS. Flight data showed that the autopilot lateral mode changed from heading mode to navigation mode at that time, which resulted in the aircraft diverging from the instructed heading. The controller repeated the instruction to turn to heading 290°. The flight crew selected 290° on the flight control unit and re-selected heading mode, and the autopilot turned the aircraft to that heading (Figure 2).
The departures controller asked the flight crew whether they were issued ‘the AVNEX SID’, which the flight crew confirmed. The controller then asked whether operations were normal, which the flight crew also confirmed. The flight continued to Denpasar without further incident.
The captain held an Air Transport Pilot (Aeroplane) Licence, multi-engine command instrument rating, and a valid Class 1 Aviation Medical Certificate. He had a total of 13,487 hours of aeronautical experience, of which 3,974 hours were on the Airbus A320. The captain had been assessed as meeting the requirement of the International Civil Aviation Organization English proficiency at a level 5 standard.[13]
The captain’s last instrument rating line check was on 18 November 2017 and the last line check was completed on 29 August 2017. Although the line check was passed with an overall marginal score of two out of five, the results indicated mostly satisfactory (required standard) scores of three. Deficiencies noted in the check were in knowledge and application of standard operating procedures and operations manual, and threat and error management as both PF and PM. Previous simulator training and check sessions identified similar issues.
First officer
The first officer (FO) held a Commercial Pilot (Aeroplane) Licence, a multi-engine instrument rating, and a valid Class 1 Aviation Medical Certificate. He had a total of 955 hours of aeronautical experience, of which 766 hours were on the Airbus A320. The FO had been assessed as meeting the requirement of the International Civil Aviation Organization English proficiency at a level 4 standard.
The FO’s last instrument rating line check was on 15 August 2017 and the last line check was completed on 30 June 2017. The line check was passed with all items scored as three out of five, which indicated a satisfactory result and the required standard was met. Previous simulator training and check sessions included comments about reading and understanding paper checklist procedures.
Fatigue considerations
A review of the flight crew’s rosters and sleep obtained found there was a low likelihood they were experiencing a level of fatigue known to have a demonstrated effect on performance.
Meteorological information
The Perth automatic terminal information service (ATIS) details relevant at the time of the aircraft’s departure indicated that the wind direction and strength was variable,[14] with a maximum tailwind of 3 kt on runway 21. The temperature was 32 °C, with CAVOK[15] conditions.
Perth Airport information
Perth Airport, with an elevation of 67 ft, is located about 10 km to the east of the city. A further 10 km to the east is the ‘Darling Scarp’ (Darling Range), an escarpment extending 345 km north-south from about Bindoon to Pemberton.
Perth Airport has two runways, 03/21 and 06/24. The movement areas, including the taxiways and runways, were marked with guidance signs, designed to assist pilots and other users with navigating the airport. These included signs that identified the holding point position with the runway designation.
Perth departure procedures
A standard instrument departure (SID) is a designated instrument flight rules departure route linking either an aerodrome or a specified runway with a specified point, normally on a designated air traffic services route, at which the en route phase of a flight commences.
At Perth, the ‘AVNEX TWO’ SID was used for both runways and in all directions (Figure 3). When departing from runway 03 using this departure, after take-off the crew were to maintain a heading of 016° (magnetic) for 4 NM (about 7 km) and conduct a left turn at waypoint MIDLA. For runway 21, the crew were to maintain a heading of 196° for 4 NM (about 7 km) and turn right at waypoint NAVEY at or above 2,500 ft above mean sea level.
Figure 3: AVNEX TWO AirAsia SID chart
Source: Aerostratus, modified by the ATSB
Operational information
Procedures and checklists
Cockpit checklists are an essential tool for overcoming limitations of pilot memory, and ensuring that action items are completed in sequence and without omission. According to Degani & Wiener (1993):
The major function of the flight deck checklist is to ensure that the crew will properly configure the airplane for any given segment of flight. It forms the basis of procedural standardization in the cockpit.
From the pre-flight preparations to the take-off, the crew completed a number of procedures and checklists where there were multiple opportunities to detect a data entry error. The relevant extracts from these checklists and procedures from the Indonesia AirAsia Flight Crew Operating Manual are below. The captain and FO were assigned specific tasks applicable to all flights.
Preliminary cockpit preparation procedure
The Preliminary Cockpit Preparation procedure specified that both flight crew were to separately obtain and calculate the aircraft’s take-off performance data. In this case, the pilot monitoring (PM) was required to cross‑check the pilot flying’s (PF’s) calculations. Specifically, the procedure stated that:
Each flight crewmember independently computes the preliminary performance data in accordance with the technical condition of the aircraft/or other criteria that may impact the aircraft performance (e.g. NOTAM [notice to airmen], runway condition, aircraft configuration).
Obtain data needed for initializing the system, preparing the cockpit and for preliminary take-off performance computation. The airfield data should include: RUNWAY IN USE, ALTIMETER SETTING, and WEATHER DATA.
PRELIMINARY TAKEOFF PERFORMANCE……………COMPUTE AND CROSSCHECK
Cockpit preparation procedure
As detailed below in the Cockpit Preparation procedure, the PM was required to check the flight plan (‘F-PLN’) entered by the PF into the flight management guidance system (FMGS) with the relevant navigation chart and ‘paper’ flight plan. The PM was also required to check the data, including the airfield information entered into the FMGS by the PF. The take-off briefing was then conducted by the PF. During that briefing, both flight crew were required to cross-check the parameters referred to by the PF to ensure that they had been set or programmed correctly.
F-PLN A page..............................................................................COMPLETE AND CHECK
The flight crew must check, modify, or insert (as applicable) the F-PLN in the following order, according to the data given by ATIS, ATC, or MET.
It is recommended to display the F-PLN page on the PM side.
Taxi procedure
For the Taxi procedure, the PM was to obtain the taxi clearance for the assigned runway, and both flight crew confirm any changes to the take-off briefing.
The Before Take-Off procedure required the PM to obtain the take-off clearance from ATC and both flight crew confirm to the runway for departure using the available cues, such as those listed below.
TAKEOFF OR LINE UP CLEARANCE………………………………………………..OBTAIN
TAKEOFF RUNWAY…………………………………………………………….……CONFIRM
Confirm the line up is performed on the intended runway. Useful aids are:
The runway markings
The runway lights
Take-off procedure
In the Take-off procedure below, the PM was to monitor the navigation display (ND) and confirm the aircraft was lined-up on the runway centreline.
PFD/ND…………………………………………………………………….....………MONITOR
If an ILS [instrument landing system] that corresponds to the departure runway is tuned, RWY mode appears. If not, no lateral mode appears until the aircraft lifts off.
Check the FMS position on the ND (aircraft on runway centerline).
Minimum height of turns
Although flight crews were to follow the designated SID procedure after take-off, the Indonesia AirAsia Flight Crew Operating Manual specifically stated:
Procedures to be followed after take-off, during the approach and go around are as per SID (including Radar departure) and STAR. No turns shall be commenced below 400ft AGL[original emphasis].
Operational philosophy
Flight crew guidance
The Indonesia AirAsia Flight Crew Training Manual and standard operating procedures provided guidance to flight crew for both normal and abnormal situations. There was no specific procedure listed on managing data entry errors detected in-flight, but general Airbus principles were included in the manuals.
Golden rules for pilots
The Flight Crew Training Manual states the pilots’ responsibility is to ‘fly, navigate, communicate’ in that order, along with additional considerations. The flight crew must perform these three actions in sequence and must use appropriate task sharing in normal and abnormal operations, in manual flight or in flight with the autopilot engaged.
‘Fly’ indicates that the PF must concentrate on ‘flying the aircraft’ in order to achieve and maintain flight parameters such as pitch attitude, bank angle, airspeed, and heading. The PM must assist the PF and must actively monitor flight parameters, and call out any excessive deviation using standard phraseology. The PM's role of ‘actively monitoring’ is very important. Therefore, both flight crew must focus and concentrate on their tasks to ensure appropriate task sharing and maintain awareness of the situation and immediately resolve any uncertainty as a crew.
‘Navigate’ refers to and includes the following ‘know where ...’ statements, in order to ensure awareness of the situation, including know where you are, know where you should be, and know where you should go.
‘Communicate’ involves effective and appropriate crew communication between the PF and the PM, and between flight crew and ATC. In abnormal and emergency situations, the PF must recover a steady flight path, and the flight crew must identify the flight’s status. The PF must then inform ATC and the cabin crew of the flight’s status and the flight crew’s intentions.
Communication also applies to communicating adjustments or changes to the information and/or equipment on the flight deck, such as FMGS alterations or flight path modifications. The other flight crew must be informed and an acknowledgement obtained.
The Flight Crew Training Manual also stated to take action if things do not go as expected. If the aircraft does not follow the desired vertical or lateral flight path, or the selected targets, and if the flight crew does not have sufficient time to analyse and solve the situation, the flight crew must immediately take appropriate or required actions, as follows:
the PF should change the level of automation from managed guidance[16] to selected guidance,[17] or from selected guidance to manual flying.
the PM should perform the following actions in sequence: communicate with the PF; challenge the actions of the PF, when necessary; and take over, when necessary.
Similar occurrences
A search of the ATSB’s database found similar occurrences involving incorrect data entry into the aircraft’s systems and navigation errors:
On 10 March 2015, an Airbus A330 aircraft operated by AirAsia X was conducting a regular passenger service from Sydney, New South Wales. On departure from runway 16R (right) the aircraft was observed by ATC to enter the departure flight path of the parallel runway, 16L (left). Following advice from ATC, the flight crew identified a problem with the on-board navigation systems. Attempts to troubleshoot and rectify the problem resulted in further degradation of the navigation system, as well as to the aircraft’s flight guidance and flight control systems. The crew elected to discontinue the flight but were unable to return to Sydney as the weather had deteriorated in the area and the available systems limited the flight to approaches in visual conditions. The aircraft was instead radar vectored to Melbourne, Victoria and the flight completed in visual conditions.
The ATSB found that when setting up the aircraft’s FMGS, the captain inadvertently entered the wrong longitudinal position of the aircraft. This adversely affected the onboard navigation systems. However, despite a number of opportunities to identify and correct the error, it was not noticed until after the aircraft became airborne and started tracking in the wrong direction. The flight crew attempted to troubleshoot and rectify the situation while under a heavy workload. Combined with limited guidance from the available checklists, this resulted in further errors by the flight crew in the diagnosis and actioning of flight deck switches. Finally, the ATSB identified that effective monitoring and assistance by ATC reduced the risk to the occurrence aircraft and other aircraft in the area.
ATSB occurrence 201806598
On 26 September 2018, a foreign-operated Boeing 737-800 was conducting a regular passenger service from Melbourne, Victoria. During initial climb, the aircraft did not adhere to the assigned runway 34 KEPPA 1 SID and conducted an early left turn. ATC queried the crew about which waypoint they were tracking to. The crew responded they were tracking to ATNOL, which was the first waypoint on the KEPPA SID 1 from runway 27. It was subsequently determined the crew had entered the incorrect departure runway of runway 27 into the flight management computer.
Very shortly after take-off from runway 21 at Perth, Western Australia, the aircraft was turned at a low height in the opposite direction to their clearance and towards an escarpment. Air traffic control (ATC) assigned a series of headings to the flight crew to correct their path. During that time, the aircraft also turned through one of the assigned headings. An additional heading was issued by ATC to return the aircraft back to its planned track. The flight continued without further incident.
Although there were time delays for the flight crew’s sectors that day, there was insufficient evidence to indicate that this contributed to the development of the incident.
This analysis will examine why the incorrect runway was programmed into the flight management guidance system (FMGS) and why this error was not detected by the flight crew. It will also discuss the flight crew’s pre-flight preparations with regard to obtaining the automatic terminal information service (ATIS) and explore their actions in response to the early turn.
Incorrect runway-in-use
Based on their recent landing on runway 03, the first officer (FO), who was the pilot flying had the expectation that the take-off runway would remain the same, as experienced on previous occasions. Consequently, he entered runway 03 into the FMGS in preparation for the departure. This was done prior to listening to the ATIS and obtaining the airfield data, which indicated that runway 21 was in-use. This was performed out-of-sequence, where the operator’s procedure required the airfield data to be obtained before programming the FMGS. While the FO commented in his interview that he would normally follow the procedure, on this occasion, he elected to program the FMGS first to make it ‘easier’.
Although the FO correctly noted the ATIS details for runway 21 on his paper flight plan, he did not detect that it was different to what he had programmed. In this case, the FO may not have detected the runway he used for programming was different due to not having an expectation it would have been incorrect. Expectations can influence perception, as people often hear what they expect to hear and see what they expect to see (Hawkins 1987). Research has also found that the pilots checking their own work are less likely to detect their own error than cross-checking by other crew members (Thomas, Petrilli, and Dawson 2004).
Independently cross-checking the runway-in-use
The captain did not separately obtain the ATIS as required by the operator’s procedures and likely relied upon the FO’s briefing that runway 03 was the runway-in-use. Further, he subsequently either did not check or detect that runway 21 was written on the flight planning documents nor that it was inconsistent with the information programmed into the FMGS.
Cross-checking is a fundamental element in all multi-crew operations and is a vital mechanism for detecting errors. An exploratory study of error detection processes during normal line operations conducted by Thomas and others (2004) identified that cross-checking and monitoring of other crew actions was the most frequently observed error detection process. By not obtaining the ATIS independently and cross-checking the runway-in-use, this removed an opportunity for the captain to identify the programming error prior to take-off.
Non-detection of error
Research conducted during normal line operations has found over half of errors made by flight crew remain undetected (Thomas, Petrilli, and Dawson 2004). In this case, the error in the pre‑flight programming remained undetected until after take-off. However, aside from cross‑checking procedures, there were multiple opportunities available to the flight crew that would have provided them with the opportunity to identify the incorrectly entered runway. These were:
Prior to, and during taxiing, the flight crew received taxi instructions and several clearances from ATC referencing runway 21, which were all read back correctly.
The flight crew reported that, during take-off, they both noticed that the runway mode had not activated. However, they explained that the mode might also not activate if the instrument landing system was not available, rather than considering this as a programming error.
While taxiing to the runway, there were guidance signs identifying the runway designation.
Of note, the flight crew received an electronic centralised aircraft monitor (ECAM) message around the same time ATC issued a hold‑short instruction for runway 21. This message directed the flight crew’s attention to responding to the ECAM message at a time potential cues to the error were available, such as the instruction and airport signage. Further, although the flight crew reported at two separate times, they felt there was something wrong with the preparation, they decided to continue with the flight. Even when cues to an error are provided there is a likelihood they can be dismissed or explained away if the information is not consistent with a person’s expectations (Hawkins 1987).
Turn contrary to clearance and published procedure
The FMGS was programmed for the runway 03 AVNEX TWO departure, which required a left turn after take-off at waypoint MIDLA. Therefore, after take-off from runway 21, the flight director indicated a left turn was required to capture the flight path to waypoint MIDLA to the north. The recorded flight data showed that the aircraft was turned left when at 223 ft above ground level. This was contrary to the runway 21 departure, which required a right turn at waypoint NAVEY to the south not below 2,500 ft. This was also below the operator’s stipulated minimum turn height of 400 ft.
The left turn continued momentarily after the autopilot was engaged but was quickly corrected when ATC assigned a heading instruction for the flight crew to regain the cleared flight path. During the left turn, the flight crew had not realised the autopilot was flying a different path to what was actually cleared. A left turn was consistent with the crew’s expectations based on their pre-flight briefing for runway 03, although the low height of the directed turn should have been unusual and not consistent with the briefing. Following the flight director at this stage suggests the crew were relying on automation at the expense of their own monitoring, possibly due to an over-reliance on automation (Parasuraman and Riley 1997).
Managing an unexpected situation after take-off
The flight crew were in an abnormal situation when they identified the programming error in the FMGS. Air traffic control provided the flight crew with a number of heading instructions to regain the cleared flight path and also asked if operations were normal, which they confirmed was the case. As highlighted by Tullo (2010), ATC can potentially assist with problem-resolution in abnormal situations. In this case, ATC were unaware of the nature of the problem and the flight crew did not utilise them fully as an additional resource. By making ATC aware of the abnormal situation, it could have potentially reduced the flight crew’s workload in deciding subsequent action to evaluate and solve the problem. This may have reduced the need for the flight crew to reprogram the FMGS at a time when their workload would have already been high, as they were required to complete other checklists.
Although the flight crew correctly identified the problem, the captain’s chosen solution of reprogramming the FMGS increased workload during an already high workload situation. Such circumstances reduces the ability to monitor the flight path (Dismukes and others 1998).
With reference to the operator’s ‘fly’ then ‘navigate’ golden rule for pilots, the role of the captain as pilot monitoring was to focus on the flight parameters including heading and communicate any deviations to the pilot flying. In this case, while ATC had provided the necessary navigation information (heading), the captain attempted to achieve this objective by manipulating the aircraft’s FMGS rather than selecting or flying to the given heading, at the expense of monitoring the flight parameters. During that time, the aircraft also turned through one of the assigned headings and an additional heading instruction was needed to be issued by ATC to return the aircraft back to its planned track.
Findings
From the evidence available, the following findings are made with respect to the data entry error related operational non-compliance of an Airbus A320, registered PK-AZE that occurred at Perth Airport, Western Australia, on 24 November 2017. These findings should not be read as apportioning blame or liability to any particular organisation or individual.
Contributing factors
Based on a recent landing, the first officer assumed runway 03 would be in use and programmed this runway for take-off into the flight management guidance system (FMGS) before listening to the automatic terminal information service (ATIS). Although the first officer copied runway 21 from the data recorded onto the flight plan, he did not notice this differed from what he had programmed into the FMGS and briefed the captain for a runway 03 take-off.
The captain did not obtain any independent information about the runway‑in‑use for pre‑departure checks, including listening to the ATIS and reviewing data recorded on the flight plan, and likely relied on verbal information from the first officer.
The incorrect programming of the FMGS was not detected before take-off despite numerous cues that the departure runway and flight path was different to what was briefed. Although the flight crew sensed there was something amiss with their pre-flight preparation, they continued without further checking.
Shortly after take-off from runway 21, the aircraft was turned left at 223 ft above ground level. This was below the minimum allowable height of 400 ft stipulated by the operator, and well before and in the opposite direction to the cleared standard instrument departure.
Other factors that increased risk
The flight crew did not communicate the nature of the problem to air traffic control and so did not effectively utilise air traffic control as an available resource. This resulted in the captain unnecessarily reprogramming the FMGS at a time when workload was already high.
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 taken by Indonesia AirAsia in response to this occurrence.
Line operations flight training, and threat and error management
Following the incident, Indonesia AirAsia have included a similar change of departure runway scenario in their line operations flight training. Further, they have plans to launch a cross-departmental initiative to increase the awareness and skill sets of pilots, especially in the area of threat and error management.
General details
Captain
Licence details:
Air Transport Pilot’s Licence (Aeroplane), issued July 2008
Endorsements:
A320 systems
Ratings:
Class ratings: single-engine land, multi-engine land
Class ratings: single-engine land, multi-engine land
Type ratings: A320
Instrument rating: multi-engine land
Medical certificate:
Class 1, valid to June 2019
Aeronautical experience:
955 hours
Last flight review:
June 2017
Sources and submissions
Sources of information
The sources of information during the investigation included:
flight crew
Indonesia AirAsia
Airservices Australia.
References
Degani, A and Wiener EL 1993, Cockpit checklists: Concepts, design, and use, Human Factors, 35(2), 28-43.
Dismukes RK, Young GE, Sumwalt III RL, and Null CH 1998, Cockpit interruptions and distractions: Effective management requires a careful balancing act. ASRS Directline, issue 10, pp. 4-9.
Hawkins, FH 1987, Human factors in flight. Aldershot: Ashgate.
Orasanu, J 2010, Ch. 5. Flight crew decision-making. In Kani, B, Helmreich, R, Anca, J. Crew resource management (2nd ed). San Diego: Academic Press.
Parasuraman, R and Riley, V 1997, Humans and automation: Use, misuse, disuse, abuse. Human factors, 39(2), 230-253.Thomas, MJW, Petrilli, RM, and Dawson, D 2004, An exploratory study of error detection processes during normal line operations. Proceedings of the 26th Conference of the European Association for Aviation Psychology.
Thomas MJ, Petrilli RM, and Dawson D 2004, An exploratory study of error detection processes during normal line operations (Doctoral dissertation, European Association for aviation psychology).
Tullo FJ 2010, Teamwork and Organizational Factors. In Kani, B, Helmreich, R, Anca, J. Crew resource management (2nd ed). San Diego: Academic Press.
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 flight crew, Indonesia AirAsia, Airservices Australia, and the Civil Aviation Safety Authority.
Submissions were received from Indonesia AirAsia, Airservices Australia, and the Civil Aviation Safety Authority. The submissions were reviewed and where considered appropriate, the text of the 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
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.
Preliminary report
Report release date: 29/06/2018
This preliminary report details factual information established in the investigation’s early evidence collection phase and has been prepared to provide timely information to the industry and public. Preliminary reports contain no analysis or findings, which will be detailed in the investigation’s final report. The information contained in this preliminary report is released in accordance with section 25 of the Transport Safety Investigation Act 2003.
Sequence of events
On 24 November 2017, at about 1200 Western Standard Time,[1] the crew of an Airbus A320 aircraft, registered PK-AZE and operated by AirAsia Indonesia, was being prepared to depart on a scheduled passenger service from Perth Airport, Western Australia for Denpasar (Bali) Airport, Indonesia. The captain was designated as the pilot monitoring and the first officer (FO) was designated as the pilot flying.[2]
While the captain was conducting the pre-flight walk around, the FO entered the flight plan into the flight management guidance computer (FMGC). Believing that they would be using runway 03 for take-off, as they had recently landed on this runway, he entered this into the FMGC. He then listened to the automatic terminal information service,[3] which indicated the runway-in-use was runway 21. When the captain returned to the fight deck, the FO completed the pre-flight and departures briefing using runway 03. At 1201, the crew received their clearance from air traffic control (ATC) to depart for Denpasar using the AVNEX TWO standard instrument departure (SID)[4] and to climb to 5,000 ft using the SID (Figure 1). At 1213, the crew commenced taxiing. The crew also received ATC clearances to taxi to, and line-up on runway 21, which was read back correctly by the crew.
Figure 1: AVNEX TWO standard instrument departure
Source: Naviga, modified by the ATSB
At 1220, the aircraft took off from runway 21. Shortly after take-off, the aircraft was turned left at 260 ft above mean sea level (AMSL) (Figure 2), which was contrary to the SID procedure and below the minimum safe altitude stipulated by the operator. The runway 21 SID required a right turn at or above 2,500 ft at waypoint[5] NAVEY (Figure 1) and the operator stipulated that turns should not be commenced below 400 ft above ground level.
After observing the aircraft turning left on radar, ATC re-cleared the crew onto an assigned radar heading. ATC later confirmed with the crew they were issued with the AVNEX TWO SID and asked if operations were normal. The crew reported operations normal and the aircraft was turned to intercept the flight planned route and continued to Denpasar without further incident.
Figure 2: Flight path of PK-AZE (in white) showing the left turn (waypoints highlighted in blue)
Source: Google earth and Air Asia Indonesia, modified by the ATSB
Recorded data
The aircraft’s flight data recorder was downloaded, and a copy was provided to the ATSB. A review of that recording found:
the waypoint MIDLA was the first selected waypoint in the flight management guidance computer, which was the first waypoint on the AVNEX TWO SID for runway 03 (Figure 1)
when manually flown, the aircraft was turned left at 260 ft AMSL
after multiple heading changes were made by the crew, waypoint SWANN was selected at 8,104 ft, which was the second waypoint on the runway 21 AVNEX TWO SID.
Ongoing investigation
The investigation is continuing and will consider the following:
operator pre-flight procedures and checklists
crew training and qualifications
aircraft systems.
______________ The information contained in this preliminary report is released in accordance with section 25 of the Transport Safety Investigation Act 2003 and is derived from the initial investigation of the occurrence. Readers are cautioned that new evidence will become available as the investigation progresses that will enhance the ATSB's understanding of the accident as outlined in this preliminary report. As such, no analysis or findings are included in this update.
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
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 6 November 2017, the fully laden, mini cape-size dry bulk carrier Orient Centaur was transiting the South Channel, Weipa, Queensland, outbound under the conduct of harbour pilots as part of a trial introduction of this size of ship to the port. While in the South Channel, the ship’s main engine shut down due to a loss of water from a cracked engine cooling component, and propulsion was lost. Shortly after, the ship grounded on the northern batter of the channel. The stern then slowly swung across the channel and grounded on the southern batter.
Under the guidance of the harbour pilots, three tugs were used to successfully refloat the ship. The ship was subsequently towed out of the channel to an anchorage. Surveys conducted over the following days identified that the ship did not sustain any damage.
What the ATSB found
The ATSB found that the approval process for this size of ship had only considered the risks associated with a main engine failure before the departing ship’s entry into the narrow South Channel. The port risk assessment did not require the use of an escort tug for any ships transiting the South Channel during outbound voyages, and the tug masters had not been trained in the specifics of escort towage nor in emergency response.
What's been done as a result
The ship’s managers advised there is now weekly on board testing of cooling water and every 6 months at a shore laboratory. Also, only manufacturers’ original spares are to be used during maintenance.
In addition to the trial conditions, an escort tug is now used for all departing bulk carriers. All departing ships over 200 m now have an escort tug made fast, from the wharf to the South Channel exit. Further, two continuously manned, 85 tonne bollard pull azimuth stern drive tugs are now based at the Port of Amrun, about 60 minutes steaming time from the Port of Weipa.
All pilots and tug masters have completed emergency response and escort towage training at the Smartship simulator. The training included emergency towage scenarios, positioning and use of tugs in an emergency, competency in standard escort tug manoeuvres and indirect towage.
Arrival and departure briefs developed for the Port of Weipa have been updated with safety settings for the electronic chart display and information system, tidal information, weather conditions, wind limits and tug positions/line lengths. All masters and ship bridge teams are briefed before arrival and departure.
Safety message
The introduction of this size of ship to the Port of Weipa was preceded by identification, and risk assessment, of a number of hazards. However, aspects of tug escort and the potential for malfunction of the heavily‑laden ships within the confined channel were not considered. This occurrence highlights the importance of considering potential hazards from end to end in order to provide the best opportunity to manage safety risk.
Video animation
The occurrence
At 1120 Eastern Standard Time[1] on 6 November 2017, the 255 m dry bulk carrier Orient Centaur (cover) completed loading a cargo of bauxite in Weipa, Queensland. The draught survey was completed shortly thereafter, with departure draughts of 12.97 m fore and aft.
At about 1415, the Port of Weipa’s dynamic under keel clearance (UKC) system became inoperative. As a result, the Weipa Vessel Traffic Services (VTS) duty officer was unable to determine the ship’s UKC. Two harbour pilots allocated to Orient Centaur for its departure then calculated that, based upon the ship’s draughts, tide and channel information, the static UKC would be 1.31 m.
At 1515, the ship’s crewmembers commenced pre-departure checks for a 1615 departure from Lorim Point Wharf (Figure 1), to sail on the evening high tide[2] to China. Shortly after, and as required for this size of ship (mini cape-size)[3], the two harbour pilots boarded the ship. The pilot taking the conduct[4] (conning pilot) had worked in Queensland ports for 9 years and was a check pilot for the Port of Weipa. The second pilot (comms pilot) had worked in Queensland ports for 10 years and was also a check pilot for the Port of Weipa. Both pilots had also recently piloted sister ships to Orient Centaur departing Weipa.
Figure 1: Port of Weipa
Source: Australian Hydrographic Service annotated by the ATSB
The pilots were escorted to the navigation bridge and met by the master. By 1537, the ship’s crewmembers had completed the pre-departure checks. At 1540, the master and pilots commenced the master-pilot information exchange (MPX) and discussed the pilotage. During the exchange, they discussed the Port of Weipa pilotage plan and the Full Mission Bridge Simulation (FMBS) departure report. [5] The ship’s route, speeds, and UKC in the channel were detailed and the tug requirements, position and repositioning were explained. Two tugs were to be in attendance, Peter Crooke (48 t bollard pull)[6][7] on the centre lead forward, and SL King (69 t bollard pull)[8] on the centre lead aft. The master was informed Peter Crooke would let go when the ship was off the berth and in the centre of the channel. SL King would let go at Gonburg Point.
The MPX form also noted that a flood tide setting from the north-west was expected at Urquhart Point, the entrance to the South Channel. The master and pilots completed the MPX at 1602, and the tugs were made fast shortly after. At 1610, the ship’s last mooring line was retrieved on board, and shortly after, the tugs started pulling the ship off the berth. By 1615, the ship was in the centre of the channel off Lorim Point, and the conning pilot ordered slow ahead. At 1618, the conning pilot ordered half ahead, and Peter Crooke was let go and stood by off the starboard quarter as the ship proceeded north of Cora Bank, to Gonburg Point.
At 1643, as the ship passed Gonburg Point, SL King was let go and stood by on the quarter, and the conning pilot ordered full ahead. Shortly after, as the ship passed beacon SC24, the conning pilot ordered port helm and started altering course into the South Channel. At about 1654, when the ship was steady in the South Channel, the conning pilot released the tugs as they passed SC22, and they headed back to their berth (Figure 2). The ship then proceeded outbound in the South Channel as planned, at a speed of about 8 knots.
Figure 2: Orient Centaur passing Urquhart Point and entering the South Channel
Source: Orient Centaur VDR, annotated by ATSB
At 1716, as the ship passed SC13 and SC14, the ship’s main engine jacket water header tank went into low‑level alarm, followed by low‑pressure and high‑temperature alarms. The chief engineer phoned the bridge and asked the master to stop the main engine for repairs, as there was a cooling water issue. The conning pilot advised the master that half ahead was acceptable and the master moved the telegraph to half ahead. He then advised the chief engineer, the main engine could not be stopped as the ship was in a narrow channel.
At 1717, the conning pilot asked the master to contact the chief engineer and confirm if half ahead was available for about an hour. Shortly after, the main engine’s automatic slowdown was activated. The master tried to contact the chief engineer regarding the main engine status. The comms pilot reported the ship was experiencing engine problems to Weipa VTS, and then issued a general call to all tugs for assistance.
At 1718, the main engine shut down, with the ship’s speed at 8.6 knots (Figure 3). Peter Crooke was the first tug to respond to the call for assistance, and the comms pilot asked the tug master to attend the ship as soon as possible. The comms pilot then made several further radio calls to Weipa VTS, stating ‘the ship had lost the main engine, taking a sheer, and was about to ground in the vicinity of SC12’.
Figure 3: Orient Centaur’s main engine shutdown to the grounding at 1 minute intervals
Source: Orient Centaur VDR, annotated by ATSB
The ship’s speed reduced to 8.3 knots as it passed beacons SC11/SC12, and it started to turn to starboard with an increasing rate of turn. The conning pilot ordered hard to port, but this had little effect. The ship continued to turn to starboard and at 1723 (Figure 4), the ship’s starboard shoulder grounded on the northern batter (bank of the channel) of the South Channel, at a speed of 6.7 knots, on a heading of 270°, and the rudder positioned amidships.
Figure 4: Orient Centaur aground in the South Channel at 1723
Source: Orient Centaur VDR, annotated by ATSB
Shortly after, SL King’s tug master responded to the comms pilot’s radio call and was asked to keep him advised of their position, as the ship’s stern started to swing across the channel. The tug master advised that SL King had just passed the wharf and would be at the ship as soon as possible. Peter Crooke’s master also advised that they were about 3 minutes behind SL King. The comms pilot then made a radio call to Weipa VTS and requested a third tug, due to the need to control the situation before the tide dropped (high water was at 1737).
At 1728, Weipa VTS advised the comms pilot that the third tug, Harry Evans (43 t bollard pull)[9] had been raised and would head out to the ship as soon as possible. At 1730, the comms pilot phoned the Regional Harbour Master (RHM) to inform him of the grounding. Unable to reach him, he called the Assistant Regional Harbour Master (ARHM). Subsequently, the ARHM gave the pilots authorisation to act in whatever manner was required to get the ship off the bank and prevent it from being aground forward and aft, across the channel, with a falling tide.
Several minutes later, the comms pilot requested the pilot boat, Windah, to read the draughts on both sides of the ship and take depth soundings around the ship’s stern. The comms pilot then advised SL King’s master to have lines ready to do heavy lifting upon their arrival.
By 1739, the ship had come to rest with the starboard shoulder aground on the northern batter, on a heading of 279°, south-west of SC 12 with the port quarter across the channel and drifting towards the southern batter.
At 1745, the comms pilot managed to contact the RHM and provided a situation update. At the same time, the ship’s crew activated their emergency procedures for a grounding. The ship’s stern continued moving slowly across the channel, before coming to rest on the southern batter at 1759, on a heading of 289°.
At 1800, the ship’s engineers identified that the cooling jacket of the cylinder head of number one main engine unit cylinder had cracked, which led to the loss of the jacket water, and the subsequent shutdown of the engine. They then started a cylinder liner jacket overhaul for the failed unit.
Shortly after, SL King arrived off the ship’s starboard quarter and put a line up through the starboard side panama lead. At 1805, the tug moved abeam and started holding the stern. At about the same time, the third tug, Harry Evans departed its berth and proceeded to Orient Centaur.
At 1808, the starboard anchor was lowered to one shackle[10] in the water to control the bow on refloating of the ship.
By 1810, Peter Crooke had arrived off the port quarter and put a line up through the port side panama lead. After the tug’s line was fast, the tug started laying back to start pulling the ship astern from the grounded position.
Orient Centaur’s stern started to slowly swing clear and away from the southern batter and, at 1820, the ship was refloated. Shortly after, SL King was relocated forward, to run a long line up the ship’s centre lead. By 1823, the line was fast and the ship’s deck crewmembers started heaving up the anchor.
By 1826, the ship’s anchor was aweigh and, at 1830, Orient Centaur was back in deep water and being manoeuvred back into the centre of the channel. The pilot’s plan from here involved SL King towing the ship, and for Peter Crooke (port quarter) and Harry Evans (starboard quarter), to steer the ship (Figure 5). The comms pilot then updated the RHM on the situation and the intended plan. The RHM advised the pilots that the ship should be towed to the emergency anchorage, that the Australian Maritime Safety Authority (AMSA) had taken charge of the incident, and that the ship was officially detained pending inspections.
Under towage, the ship passed between SC10 and SC9 at 1838, at a speed of 3.8 knots with an UKC of 3.2m. At 1843, the third tug, Harry Evans was off the starboard quarter and by 1848, had made fast a line through the starboard side panama lead.
By 1856, the ship had passed SC7 at a steady speed of 4.5 knots and shortly after 1948, passed the fairway buoy and was clear of the South Channel.
The pilots and Orient Centaur’s master checked the proposed route to the emergency anchorage, and determined that the emergency anchorage position would not provide a satisfactory UKC for the ship in the loaded condition. At 2010, the comms pilot informed RHM that the emergency anchorage area did not have sufficient water for the ship’s draught and an alternative position was agreed.
Figure 5: Tugs Harry Evans arriving at Orient Centaur and Peter Crooke fast on the port quarter
Source: Ports North pilots
At 2120, the tugs started slowing the ship in preparation for letting go the starboard anchor. Shortly after, SL King was cast off and the ship’s crew prepared the anchor for letting go.
The ship’s starboard anchor was let go at 2137. By 2206, all three tugs were clear of the ship and the pilots disembarked at 2218.
Orient Centaur remained at anchor throughout 7 November 2017, during which time the ship’s engineers completed the cylinder liner jacket replacement. The main engine was tested and made ready for departure.
At 0800, 8 November 2017, marine surveyors from AMSA and the ship’s classification society attended the ship with a dive team. The dive report noted some loss of paint on the starboard shoulder and no damage or paint loss on the port quarter.
The AMSA surveyor revoked the detention order at 1345, and at 1518, the ship weighed anchor and proceeded to China.
Orient Centaur was fitted with a Hyundai MAN B&W 6S60MC-C7 slow speed engine (appendix A). The six-cylinder engine is cooled by a central cooling system.[11]
Figure 6: Jacket water cooling diagram
Source: MAN Two-stroke MC/MC-C Engines
The cylinder liner is made of alloyed cast iron and suspended in the cylinder frame with a low‑situated flange. The top of the cylinder liner is fitted with a cooling jacket. The cylinder cover, made of forged steel, is attached to the cylinder liner and has bores for cooling water.
The jacket cooling water system (Figure 6) is used for cooling the cylinder liners, cylinder covers, and exhaust valves of the main engine, and for heating the fuel oil drainpipes.
The engine manufacturer had issued a service letter (SL2014-584/LRA) in January 2014, regarding the cooling water systems associated with engines of this type. The service letter provided a number of possible actions to prevent corroded or cracked engine parts (cylinder cover, cylinder liner and cooling jackets) and cooling water leakage. It stated:
…most cases of heavy corrosion in the cooling water system or cracked cylinder cooling jackets on the cylinder cover, were related to insufficient cooling water maintenance. Heavy deposits tended to find the way into the clearances between the cylinder cover and the cooling jacket. This made heat expansion of the cylinder cover impossible without also exposing the cooling jacket to significant stress as the heat expansion is reducing the clearances. In combination with heavy corrosion of the engine components caused by insufficient cooling water treatment, the cooling jacket may consequently crack.
Consistent with the detail in the service letter, the engine jacket water needed to be carefully treated, maintained and monitored to avoid corrosion, corrosion fatigue, cavitation, and scale formation.
The crewmembers who were in the engine room at that time of the occurrence sighted water coming from a crack in the cylinder cooling jacket of the cylinder head of cylinder number 1 (Figure 7). The loss of jacket water triggered the main engine to shut down.
Figure 7: Cylinder cooling jacket failure and crack (after cleaning)
Source: Interorient accident investigation report annotated by ATSB
The engine manufacturer subsequently inspected the failed cylinder cooling jacket and identified that:
the jacket had one vertical crack that passed through a securing bolt hole
there were four bulges around the crack but no evidence of any impact marks
the cooling space of the inner cooling jacket had heavy corrosion/rust and metal peeled off.
The cooling water systems were tested for nitrites, chlorides and pH after the incident. All were within limits, however the chloride tests of the high temperature cooling water were slightly high.
Port of Weipa
Located on the north-west coast of Cape York Peninsula, the Port of Weipa’s primary export is bauxite (aluminium ore) from the Rio Tinto Alcan mine (RTA). North Queensland Bulk Ports Corporation (NQBPC) is a statutory Queensland government‑owned corporation charged with overseeing the commercial activities in the port, including the maintenance of the port infrastructure. The primary function of NQBPC, under the Transport Infrastructure Act 1994, is to establish, manage and operate effective and efficient facilities and services within the port, while maintaining appropriate levels of safety and security.
Port Procedures
Maritime Safety Queensland (MSQ), the maritime regulator, administers the State’s shipping legislation. MSQ appoint Regional Harbour Masters (RHM), who are authorised to give direction under the relevant provisions of the Transport Operations (Marine Safety) Act 1994. An RHM controls the pilotage areas within their region and has the authority to direct the master of a ship to navigate or operate a ship in a prescribed way.
Each pilotage area has a Port Procedures and Information for Shipping Manual (PPM).[12] The PPM details services, mandatory regulations and procedures to be observed. The manual also includes guidelines and information to assist masters, owners and agents of vessels arriving and departing the area.
MSQ operates a Vessel Traffic Service (VTS) for the Port of Weipa. The VTS is based at the RHM's office in Cairns, Queensland. VTS operators at Cairns have direct contact with shipping. In the event of an emergency, the VTS centre is the key notification and communications facility and activates the appropriate response agencies.
Pilotage
All ships that are 50 m or more in length, proceeding within a Queensland pilotage area, must carry a licensed marine pilot or be under the command of a master who holds a pilotage exemption certificate for the area. Far North Queensland Ports Corporation Ltd (Ports North) provide the compulsory pilotage service for the Port of Weipa, as well as for the ports of Cairns, Cape Flattery, Mourilyan, Karumba, Thursday Island (Port Kennedy) and Skardon River.
Sailing times
Ships can only depart Lorim Point on a flood tide and are scheduled so as to depart the berth 1 hour 15 minutes before the high water at Humbug Point (Figure 1). This scheduling allows ships about 30 minutes to enter and steady in the South Channel by the time they are off Urquhart Point. By that time, the tidal stream should have reduced to less than one knot.
First-strike response
The Port of Weipa had an emergency response plan (First-strike Oil Spill Response Plan), prepared by the Queensland Department of Transport and Main Roads, for dealing with pollution from ships. The plan detailed the operational arrangements between MSQ and Ports Corporation of Queensland and the response and handover arrangements for oil spills within the port. Further, the plan identified available resources and key contact information for responding to an oil pollution incident.
The plan was to be used in conjunction with the following:
the Queensland Coastal Contingency Action Plan (QCCAP)
Maritime Safety Queensland's Standard Operating Procedures for oil spill response
the Oil Pollution First-Strike Response Deed for the Port of Weipa.
MSQ is both the statutory and combat agency for ship‑sourced oil spills and the incident controller for all incidents within the scope of the plan. North Queensland Bulk Ports (NQBP) are responsible for ensuring that an adequate first‑strike oil spill response capability is maintained within the port. The incident controller was the RHM, assisted by the NQBP Port Manager. The local first-strike response team consisted of six personnel, with either the NQBP Port Supervisor or MSQ representative as local incident controller, with five personnel from Rio Tinto Alcan.
Although the grounding incident did not result in any pollution, the incident debrief attended by all directly involved parties identified an issue with the first-strike response. As NQBP and MSQ have only one officer each based in Weipa, NQBP utilises Rio Tinto personnel to meet their responsibilities. It was not clear whether the first‑strike response team could have been assembled in a timely manner with the number of local personnel as required by the QCCAP.
Mini cape-size bulk carrier
Weipa arrival and departure simulation
In 2017, Rio Tinto Marine submitted a written application to the RHM for mini cape-size ships that were 255 m in length with a 43 m beam to export bauxite from Weipa. As the ships exceeded the port’s size limits, Full Mission Bridge Simulation (FMBS) exercises were required. In May and July 2017, the RHM, pilot manager, several senior pilots and a Rio Tinto shipping specialist, attended the MSQ Smartship simulator in Brisbane, Queensland, to conduct the FMBS exercises. The exercises (Figure 8) used a model of the mini cape-size bulk carrier for arrivals in a ballast condition and departures in a fully laden condition, to a maximum draught of 12.8 m.
Figure 8: Port simulation exercise report
Red arrows show areas of bank effect, yellow arrows show flood tide, other annotations show ship’s course, past positions and tidal information.
Source: MSQ regional harbour master (Cairns)
The ship’s manoeuvrability was tested in normal[13] and extreme[14] weather conditions. The recommendations included, among others:
The vessels push the limits for the channel, but simulation indicated they can safely enter and leave the port in conditions expected at Weipa.
Before final sign off can be given the following will need to be completed:
All pilots will need to spend two days in the simulator before handling the vessels to ensure they get “the feel”. These vessels are marginal and it was shown in the initial simulation that the pilots need to get used to the lack of power and the effect on the steering capability.
It was further recommended another 2-day simulation exercise be conducted to train two more pilots to handle the vessels. On the completion of the (Full Mission Bridge Simulation) exercise and risk assessment, a vessel of this class is chartered to Weipa for a couple of voyages to “prove” the feasibility. Provided this is successful more pilots would be trained and the number of mini cape vessels be limited to 5 vessels of the same class.
Risk assessment
The RHM’s risk treatment action plan, based on the outcomes of the FMBS exercises for mini cape-size ships entering the Port of Weipa was completed prior to any shipment trials. The plan covered four areas, general, channel entry, berthing and departure. The plan also covered the risks, current controls, interventions and identified the responsible officer in each area.
The general risk controls included the PPM requirements, simulated arrival and departure conditions, bridge resources management issues, and ship characteristics among others. The risk controls also required that two pilots, appropriately qualified and trained in the simulator, were to be on board for the ship’s transit.
The arrival channel entry risks detailed environmental effects, wind and leeway limits related to the 106 m wide South Channel widening to Cora Bank (Bell Mouth). The average draught of the mini cape-size bulk carriers in ballast was 7.5 m and the ships arrive on a rising tide, with the maintained channel depth at 11.1m. The risk assessment identified the possibility of main engine failure during arrival, and the ship’s anchors ready and tugs available as the mitigating risk controls. This was on the basis that ships in ballast have shallower draught, and therefore have less chance of grounding in the 11.1 m deep approach channel. Further, with this greater clearance under the keel (compared to fully‑laden deep draught departure channel clearance) the ships’ anchors can be used to assist manoeuvring and control.
In addition, during arrivals, a tug is required to be on stand-by at SC24 beacon, at the entry into Cora Channel. The tug is positioned adjacent to the section of approach channel with the more shallow water outside the channel and thus the greater risk to grounding should a ship leave the channel. From this position, the tug is immediately available to assist ships on approach to Urquhart Point if needed.
The berthing risks included control measures for the ship’s draught, usage of tugs, speed of approach and use of sound navigational methods. An intervention measure required that only pilots with an endorsement for mini cape-size ships would be assigned to this type of ship.
The departure risks identified the same areas of risk as for the channel entry, with the same controls and intervention measures. Engine failure was also identified as a risk, with controls being escort tugs, and that they be available.
At the conclusion of the exercises, the RHM recommended ship trials into the Port of Weipa under the following conditions:
the vessels must be one of the five sister ships
vessel to be loaded no deeper than 12.8 m and use static UKC of 10 per cent draught
entry and departure for first six visits to be in daylight only
two senior pilots who have attended the simulation to be on board
wind conditions maximum 25 to 30 knots
departure from Lorim Point to be 1 hour 15 minutes before high water at Humbug Point to minimise current at Bell Mouth
tugs to be available to escort until vessel has entered the South Channel – this may restrict other departures or arrivals
6 trial shipments to be permitted subject to the above conditions
following trial, review pilot observations
should further shipments be approved all, pilots who conduct the vessel to have completed 2 days simulator training.
Tug requirements
The PPM stated that all ships over a certain size that used the Lorim Point (Figure 1) berths were required to use a minimum of two azimuth stern drive (ASD) tugs for all movements. In addition, guidance for ships departing Lorim Point stated that the forward tug should be let go after the ship responded to helm movements, and the aft tug as required, when the ship was steady and passing Gonburg Point. Further, during the trials, tugs were to be available to escort until the ship had entered the South Channel. One of the objectives of escorting a ship is to apply high steering and braking forces via a towline to limit to impact of grounding. Escort tugs are specifically designed for this role to operate over long distances and at relatively high speeds. At the time of the occurrence, the Port of Weipa did not have any escort-rated tugs and the tug masters were not trained in the specifics of escort towage nor in emergency response.
ASD tugs can be operated in either direct or indirect towage methods (Figure 9). The method used in the Port of Weipa was direct towage, as the tugs were predominantly used for harbour work and assisting ship berthing and un-berthing manoeuvres. Direct towing is performed at low speeds from about 0‑ 5 knots. The tug uses its engines and propellers to generate the tow force (directly), which is governed by the conventional bollard pull of the tug.
Indirect towing can be performed at speeds in the range of 5‑12 knots. The tug generates a tow force (indirectly) by putting tension on the tow line using a combination of the tug’s weight, water pressure (from the tug’s skeg) and the resultant drag of the propeller wash. The tug can assist swinging and steadying by working around the radius of the towline. When working at the stern of a ship, the pull on the towline can be increased by a factor of 1.5 to 2 times the bollard pull.
The effectiveness of the towage methods is also dependent on the availability of safe water to work in. In a narrow channel, the ability to work around the radius of the towline is restricted. Therefore, with the ship at speed (about 8 knots) in the South Channel (106 m) it was prohibitive for a tug to effectively steer it using dynamic arrest mode. However, direct arrest mode remained an effective option when retarding (braking assistance) forces were required.
Figure 9: Azimuth stern drive tugs - direct and indirect towage methods
Source: Tug Use in Port, A Practical Guide, Captain Henk Hensen, annotated by ATSB
As Orient Centaur proceeded outbound in the South Channel, the main engine shut down due to low cooling water flow, and a loss of propulsion.
The loss of cooling water was the result of a crack in a main engine unit’s cylinder cooling water jacket through a securing bolt hole, an area of inherent stress concentration. Furthermore, consistent with the engine manufacturer’s service letter, it is possible that the effects of corrosion products found in the cooling water space contributed to the cracking due to increased stress in the area of the failure.
Mini cape-size bulk carriers operation
Port risk assessment
The hazards associated with engine failure (loss of propulsion) during departure from Weipa was considered during the approval risk assessment process and were to be mitigated by having tugs in attendance until the ship was steady in the South Channel. On the day of the occurrence, and after the ship had passed beacon SC 21, the tugs were stood down and instructed to return to the berth. Consequently, when the loss of propulsion occurred in the narrow South Channel, there was no effective mitigation against grounding. The time taken for tugs to reach the ship would depend on its position in the channel, as well as the proximity of the tugs.
The subsequent investigation report observations prepared by the Regional Harbour Master (RHM) stated:
Completed Risk Assessment shows having tugs standing by until vessel is steady in channel would help prevent the vessel “T boning” the shoreline in areas where the channel is wide enough for the vessel to sheer sufficiently to be 90 or almost 90 degrees to the bank when grounding could be more violent and lead to hull damage.
Slope of batter in vicinity of Beacon SC 12 was beneficial to grounding as it slowly brought the vessel to a stop.
Further to seaward increase in depth and reduction in slope of batter would have seen the vessel get further out of the channel and may have been grounded for more of the length of the vessel.
No vessel discharge facility or equipment is available in Weipa.
As the South Channel was only 106 m wide, the ship was too long to ‘T bone’ in the channel. However, after grounding on the northern batter, the stern swung on the falling tide and grounded on the southern batter, effectively blocking the channel. There was then a significant risk the ship’s structural integrity could be compromised and at that time, the pilots estimated they had about 30 minutes to refloat the ship. Fortunately, the actions taken by the tugs and the pilots on board enabled the ship to be re‑floated before the falling tide rendered the task impossible.
The subsequent grounding debrief, attended by the RHM, Maritime Safety Queensland, Far North Queensland Ports Corporation Ltd (Ports North), North Queensland Bulk Ports and Smit Lamnalco, identified this risk:
If the Orient Centaur hadn't been successfully floated the vessel may have grounded forward and aft across the channel. The steep channel batters coupled with the ebbing tide may have compromised the hull's integrity. This could have blocked the channel and essentially closed the Port of Weipa. Due to the potential complexity of the salvage and also the remote location and long transit times for suitable salvage equipment this could have led to a salvage and port closure lasting many months, which would have an effect on the Queensland economy.
Further:
If the vessel had grounded further west where the gradient of the batter was more gradual, the refloat may have been more difficult, although hull integrity may not have been such a concern.
Simulation
The pilots stated that, during departure, Orient Centaur handled in the same manner as during simulation. However, during the simulation exercises for the mini cape-size ships, emergencies such as loss of propulsion and/or steerage were not simulated in the South Channel. The pilots had previously been trained in simulated emergency scenarios and escort towage, but only for events such as loss of propulsion and engine failure in smaller vessels when entering/departing the less confined Bell Mouth.
After the incident, the ATSB attended the Smartship simulator to observe several pilots participating in a 2‑day mini cape-size ship simulation. During this time, the incident was simulated with and without tug assistance. With tug assistance (tow line centre lead aft), the ship was kept directionally stable and its speed reduced significantly, to control a grounding. Without tug assistance, the ship grounded in an uncontrolled manner.
The inclusion of a loss of propulsion simulation in a confined channel during the mini cape-size approval process would have informed the likelihood of grounding and the subsequent impact on port operations and provided the opportunity to practice the response to such an event.
Tug usage and availability in the Port of Weipa
In addition to the requirements of the Port Procedures and Information for Shipping Manual and the Ports North pilotage operations safety management system, the RHM required two tugs to escort a mini cape-size ship until it was steady in the South Channel. The tug crews then had to remain with the tugs until the ship had departed the South Channel.
The contingency for a ship grounding in the South Channel was that the tugs would be manned and ready to proceed to the ship. After Orient Centaur grounded, the tugs started to arrive at the ship after about 47 minutes.
The pilot’s orders to the tug masters during the refloating of the ship were standard and well understood, and the required tug movements were also standard practice. However, during the recovery, the operation of two tugs working in close proximity on the stern of the ship was different to that of the day-to-day operations of the tugs.
With a ship at speed, and with two tugs working at the stern, interaction forces between water flow patterns are generated by both the ship and tugs, and between the tugs themselves. When tugs are towing on a line, the situation is further complicated, as they are frequently changing position and heading, therefore operating in areas where they experience different interaction effects. The tug masters were conscious of these effects and reported spending a lot of time avoiding contact with each other during the recovery. This was particularly apparent when asked to either layback, or move out to lift the ship’s stern from port or starboard. This shows that a higher skill set than those associated with standard harbour towage practice was required during the response to the occurrence and for escort towage more generally.
However, the tug masters had not been trained in indirect towage methods, multiple tug usage or emergency scenarios. Additionally, at the time of the occurrence, the tugs were considered to be used as a means of controlling grounding and were not escort rated.
Emergency response to a pollution incident
This grounding occurrence did not produce any pollution however, in different circumstances activation of the pollution emergency response plan would been required. That response relied on trained personnel from North Queensland Bulk Ports to form the First-strike Oil Spill Response team. However, due to the limited availability of personnel at the time of the occurrence it was unlikely that that an emergency response to a pollution incident in the Port of Weipa would have been initiated in a timely manner. As detailed in the Safety issues and actions section of the report, appropriate proactive action has been taken to improve the emergency response.
Findings
From the evidence available, the following findings are made with respect to the grounding of Orient Centaur, which occurred in the South Channel, Weipa, Queensland on 6 November 2017. 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
As the ship was transiting the South Channel, its main engine shut down due to a loss of cooling water associated with failure of the number 1 cylinder cooling jacket. Following the loss of propulsion, the ship’s starboard shoulder grounded on the northern batter and its port quarter drifted across the channel to ground on the southern batter.
Other factors that increased risk
In pre-trial simulations, the risks associated with engine failure during departure were only considered up to when a ship had entered the South Channel. Consequently, the tugs were not in attendance to assist if propulsion was lost. [Safety issue]
The conditions for trial shipments required all pilots to complete extensive arrival and departure simulations, including transit through the South Channel. However, the simulations did not include contingencies, such as loss of propulsion and/or steerage that would have assisted in determining the risk of, and response to, such occurrences.
Tugs were to be available to escort the mini cape-size ships until they had entered the South Channel, where they were stood down. However, the tug masters had not been trained in the specifics of escort towage nor in emergency response. [Safety issue]
Other findings
At the time of the incident, it was unlikely that that an emergency response to a pollution incident in the Port of Weipa would have been initiated in a timely manner. Subsequently, appropriate proactive action has been taken to address the emergency response issues.
The actions of the pilots, ship's crewmembers and tug masters following their arrival at the ship were timely and significantly reduced the consequences of the occurrence.
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 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 initiate safety action proactively, 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 marine industry, the ATSB may issue safety recommendations or safety advisory notices as part of the final report.
All of the directly involved parties were provided with a draft report and invited to provide submissions. As part of that process, each organisation was asked to communicate what safety actions, if any, they had carried out or were planning to carry out in relation to each safety issue relevant to their organisation.
Descriptions of each safety issue, and any associated safety recommendations, are detailed below. Click the link to read the full safety issue description, including the issue status and any safety action/s taken. Safety issues and actions are updated on this website when safety issue owners provide further information concerning the implementation of safety action.
Safety issue description: In pre-trial simulations, the risks associated with engine failure during departure were only considered up to when a ship had entered the South Channel. Consequently, the tugs were not in attendance to assist if propulsion was lost.
Safety issue description: Tugs were to be available to escort the mini cape-size ships until they had entered the South Channel, where they were then stood down. However, the tug masters had not been trained in the specifics of escort towage nor in emergency response.
Additional safety issues
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.
Interorient, the ship’s managers advised the following proactive actions will be taken:
• water analysis on board weekly
• water analysis at shore every 6 months
• checking appropriate tightening of Cylinder Head every 8000 hrs
• proper overhaul of cooling jacket (replacement of o rings and cleaning) and use of original spares according to manufactures instructions, service letters and manual every overhaul
• checking securing bolts as per manufactures service letters every overhaul
• proper tracking of running hours monthly
• original spare cooling to be available on board at all times
• all cylinder jackets renewed within 1 year from incident
• maker guidance/recommendations and service letters are followed as per planned maintenance system schedule.
North Queensland Bulk Ports (NQBP) advised their operations have been reviewed and have taken the following proactive actions:
NQBP have prepared a tender package for the provision of Salvage and Casualty Advice for all NQBP Ports to include:
- guidelines to assist NQBP on items that may be included in vetting requirements to minimise the likelihood of an incident in the future
- the establishment of appropriate minimum insurance requirements for ships using NQBP ports
- provide response timing and plans, which are considered for reasonably foreseeable events within their Ports.
Obtained quotes for Vessel Vetting Service (Rightship) to facilitate implementation of additional vetting protocols. Implementation is planned following the review of Salvage Planning, as this work will also consider vetting parameters that can be enforced to minimise the likelihood of an incident occurring.
As part of this process, NQBP has already sought from each of our Terminal Operators detail of their vetting processes, specifically the criteria that they are vetting to.
Completed a First Strike Oil Response exercise in Weipa
NQBP was added to the Weipa VTS Incident Notifications
NQBP are currently in the process of undertaking a Risk Assessment & Management review of our Ports for marine operations. This process is being completed in conjunction with the RHM, Terminal Operators and Towage Providers for each Port.
General details
Ship details
Name:
Orient Centaur
IMO number:
9464572
Call sign:
5BYD2
Flag:
Cyprus
Classification society:
Lloyd’s Register
Ship type:
Bulk carrier
Year built:
2010
Owner(s):
Orient Centaur Shipping
Manager:
Interorient Marine Services
Gross tonnage:
63993
Deadweight (summer):
114841 t
Summer draught:
12.97
Length overall:
255.258
Moulded breadth:
43.00
Main engine(s):
Hyundai MAN B&W 6S60MC-C7
Total power:
13560 kW
Speed:
14.5
Damage:
nil
Appendices
Appendix A – Engine cross section MAN B&W 6S60MC-C7
Sources and submissions
Sources of information
The sources of information during the investigation include:
the master and involved crewmembers of Orient Centaur
Interorient
the Australian Maritime Safety Authority
the Regional Harbour Master (Cairns), Maritime Safety Queensland
the Manager Pilotage Services, Ports North
North Queensland Bulk Ports
Smit Lamnalco.
References
Rowe, Captain RW (with Russell, Captain PJD). The Ship Handler’s Guide, The Nautical Institute in conjunction with the Warsash Maritime Centre, London.
Hensen, Captain Henk. Tug Use in Port, A Practical Guide, The Nautical Institute, London.
Maritime Safety Queensland (MSQ) 2017, Port Procedure and Information for Shipping – Weipa, MSQ, Brisbane. Available at https://www.msq.qld.gov.au
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 master and directly involved crew and managers of Orient Centaur, the ship’s flag State, the pilots and manager pilot services, the Regional Harbour Master (Cairns), North Queensland Bulk Ports, Smit Lamnalco and the Australian Maritime Safety Authority and Regional Harbour Master (Cairns).
Submissions were received from the Australian Maritime Safety Authority, the pilots and manager pilot services and the Regional Harbour Master (Cairns). The submissions were reviewed and where considered appropriate, the text of the 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
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.
At 1242 Eastern Standard Time[1] on 18 November 2017, a Bell 206 helicopter, registered VH‑SDZ (SDZ), departed from Middlemount Airport, Queensland with a pilot and crew member on board to conduct power line inspections 32 km to the south.
At approximately 1255, the pilot positioned the helicopter in a hover 30 ft above the ground and 40 metres from a transmission tower so that it could be photographed by the crew member. After 3-4 minutes of hovering, the pilot heard what was described as a ‘very loud bang’ through the airframe, which was also felt through the controls. The helicopter began to shake violently and bounce vertically. The pilot also reported seeing tiny pieces of debris falling in front of the helicopter.
In response, the pilot immediately lowered the collective, intending to land in a clear area below the helicopter, but it did not respond to collective or cyclic control inputs. Instead, the helicopter began to pitch upward and drift backwards. The helicopter then yawed to the right, most likely due to contact with trees behind. The yaw could not be controlled with the tail rotor pedals so the pilot moved the throttle to the idle position. Despite attempts to level the aircraft, the left skid contacted the ground first and the helicopter rolled over.
The pilot moved the throttle and fuel shut off valve to the off position and switched off the battery before both occupants exited the helicopter. The pilot and the crew member were uninjured as a result of the occurrence.
Figure 1: VH-SDZ following collision with terrain
Source: Operator
Component examination
Examination of the helicopter following the occurrence identified that the mast, both pitch links, and the swashplate were fractured. The mast collar set was also fractured and not engaged with the mast[2]. The operator initially suspected that the mast collar set had disengaged in flight.
The mast collar secures the swashplate’s rotating ring to the mast. It is designed to drive the rotating ring at the same speed as the main rotor. If the mast collar is not engaged, the pitch links connecting the swashplate to the rotors are exposed to transverse forces for which they are not designed. This can result in the pitch links winding around the mast.
The fractured components and trunnion bearings (connecting the pitch links to the rotor blades) were sent to Bell Helicopter’s Engineering Laboratories for detailed examination (Figure 2). The initial examinations were also attended by the United States Federal Aviation Administration. Bell provided a report of their findings to the ATSB.
Figure 2: Components examined at Bell Helicopter’s Engineering Laboratories
Source: Bell Helicopter
The hypothesis of mast collar separation was considered by Bell Engineering Laboratories. It was determined that an in-flight separation would likely have resulted in rotational damage on the mast and mast collar splines, which was not observed. The damage observed in the pitch links was also not consistent with a separated mast collar.
Bell Helicopter concluded that all of the fractures, including that resulting in the mast collar separation, were a result of overstress. No pre-existing defect was found and all damage observed was secondary to the occurrence. Seized or damaged trunnion bearings were also considered for inhibited movement of the pitch links, which could result in abnormal bending loads. However, the condition of the trunnion bearings was found to be typical for bearings removed from service.
In summary, there were no findings made during the inspection at Bell Helicopter’s Engineering Laboratories that identified the probable factors contributing to this occurrence. The reason for the loss of control was therefore not determined.
Findings
While hovering, the pilot experienced a loss of cyclic and collective control that resulted in a ground collision. The reason for the loss of control was not able to be determined.
Safety message
This accident highlights how rapidly an emergency situation can develop. Recognising that the pilot had limited control authority on this occasion, regular practice and/or briefing of emergency actions will increase the likelihood of a correct response.
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
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 8 November 2017, at about 1508 Eastern Standard Time,[1] the pilot of a Robinson R44 helicopter, registered VH-WRR and operated by Whitsunday Air Services, ditched about 49 km north of Hamilton Island Airport, Queensland. In addition to the pilot, there were three passengers on board.
Prior to the flight, the passengers received a safety briefing and were instructed to wear seatbelts and life jackets. At about 1430, the helicopter departed Hamilton Island for a 1 hour scenic (charter) flight over Whitehaven Bay and the Great Barrier Reef.
When about 40 minutes into the flight, on return to Hamilton Island, the pilot heard the engine sound decrease and noted that the helicopter was unable to maintain the cruise altitude of 1,000 ft. The pilot checked the manifold pressure[2] gauge and noted it was at 22 inches Hg, however, it was set at 24.5 inches Hg when leaving the reef between 500 and 1,000 ft. In response to the reduction in power, the pilot raised the collective[3] to increase power and maintain altitude, but the manifold pressure did not change. The pilot reported that the indicated main rotor revolutions per minute (RPM) decreased and the low rotor RPM horn activated. The engine RPM indication was also oscillating throughout the range.
In response to the low RPM horn, the pilot increased throttle and again raised the collective. The rotor RPM initially spiked and then decreased with associated re-activation of the low rotor RPM horn. The engine RPM gauge continued to oscillate and the helicopter could not maintain altitude. The pilot noticed the engine noise was changing with the fluctuations in the indications. However, there were no further indications that suggested a problem with the helicopter.
Consequently, when at 700 ft, the pilot elected to conduct an autorotation[4] onto the water. The pilot activated the emergency flotation system[5] and broadcast a MAYDAY[6] call on the common traffic advisory frequency. The engine RPM gauge continued to provide erratic readings and the low engine sound continued during the landing. About 30 seconds later, the helicopter landed on the water with the emergency floats deployed (Figure 1).
The pilot contacted another company pilot who was operating in the area and informed them of the situation and their location. The pilot shut down the engine and applied the rotor brake.[7] The pilot activated the emergency locator transmitter and instructed the passengers to prepare to inflate their life jackets and undo their seatbelts in readiness to exit the helicopter.
About 10 to 15 minutes later, the company aircraft was circling overhead. The pilot and passengers remained in the helicopter until they were rescued by the crew of a local vessel about 1 hour later. All occupants were uninjured. While there appeared to be no observable damage sustained to the helicopter, it later sank and was unable to be recovered.
Figure 1: VH-WRR after ditching about 49 km north of Hamilton Island Airport, Queensland
Source: Australian Maritime Safety Authority, modified by the ATSB
Additional comments
The following additional comments were made with regard to the accident:
The pilot reported refuelling the helicopter prior to the flight and conducting a check for contaminants, none of which was found. After the helicopter had landed on the water, the fuel gauges indicated that the tanks were half full.
A review of the meteorological conditions around the time of the accident indicated they were conducive to serious carburettor icing conditions with descent power selected.
The Civil Aviation Safety Authority noted the partial power loss described in this accident may be consistent with a magneto/governor failure, which has reported to have occurred on other R44 helicopters.
A maintenance logbook entry around 2 weeks prior to the accident noted that the helicopter had intermittent tachometer/governor fluctuations. The right magneto points were found to be out-of-tolerance and adjusted. A ground run to test the adjusted magneto was completed satisfactorily. Robinson Helicopter Company had issued Service Letter 62, which stated that the throttle governor signal source was ‘the tachometer breaker contact (points) assembly located within the engine-right magneto…excessive wear causing insufficient point gap…(that) could cause an erratic tachometer indication.’
Post-accident discussions between the manufacturer, operator and maintenance provider resulted in a consensus of opinion that the power reduction was associated with either a governor control failure and or a compromised engine RPM signal from poor tachometer points.
Similar occurrences
A search of the ATSB’s database found the following occurrences involving an engine power loss followed by a ditching in helicopters:
On 3 January 2011, a Robinson R44 helicopter, departed Cairns Airport, Queensland for a 30-minute charter flight (ATSB investigation AO-2011-001). About 25 minutes into the flight, when at about 400 ft above mean sea level, the engine failed and the rotor low RPM horn sounded. The pilot broadcast a MAYDAY and entered autorotation. During the descent, he deployed the emergency floatation system, however, the right float did not fully inflate. When at about 50 ft above the sea, the helicopter entered an uncommanded 360⁰ yaw to the left. The pilot was unable to control the yaw and the helicopter impacted the water heavily and turned onto its right side. The pilot assisted the passengers to egress and inflated their life jackets. A post-occurrence engine strip and examination found no fault that would give reason for the engine to fail in flight.
On 26 January 2011, a Robinson R44 helicopter departed the Knuckle Reef helipad, Queensland, for a 20-minute charter flight (ATSB investigation AO-2011-008). On board the helicopter were the pilot and three passengers. While returning to the helipad 15 minutes later, at about 950 ft above sea level, the helicopter experienced engine problems, including a sudden loss of cylinder head temperature indication and variations in the engine manifold pressure. The helicopter was unable to maintain altitude and began to descend at 200 ft per minute. The pilot inflated the emergency floatation system, commenced an autorotation and landed on the sea. Following the safe recovery of the occupants, the helicopter was unable to be recovered.
Findings
These findings should not be read as apportioning blame or liability to any particular organisation or individual.
During the cruise, for undetermined reasons, the helicopter experienced a partial engine power loss. After unsuccessful attempts to increase power, the pilot manually deployed the emergency floatation system and performed a successful ditching.
Safety message
Power loss
Without the recovery of the helicopter, the reasons for the partial engine power loss could not be conclusively determined. However, the indications were consistent with a magneto/governor failure, which Robinson Helicopter Company have published a revised service letter Governor troubleshooting / magneto maintenance describing how the failure occurs and the compliance procedure.
The pilot commented that this was the first time he had been involved in an emergency situation and the training he received prepared him well for executing the autorotation and managing the situation after landing on the water.
The United States Federal Aviation Administration Helicopter Flying Handbook stated that, when rotor RPM begins to decrease, it is essential to recover and maintain RPM. Low rotor RPM and ensuing blade stall can result in a total loss of rotor lift, allowing the helicopter to fall to the surface and possibly resulting in blade strikes to the tail boom and other airframe damage. Low rotor RPM during an autorotation may result in a less than successful result.
Survival factors
This accident highlights the importance of being adequately prepared for an emergency situation such as a ditching. In this case, the helicopter was fitted with an emergency floatation system, the passengers were wearing life jackets, and the pilot’s post-landing actions resulted in a positive outcome where no injuries occurred. Without a floatation system the risk of the helicopter sinking with the occupants on board would be greatly increased.
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
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 18 November 2017, the flight crew of a Boeing 777-319(ER) aircraft, registered ZK-OKN and operated by Air New Zealand, was conducting a scheduled passenger service from Auckland, New Zealand, to Brisbane, Queensland. The flight crew consisted of the aircraft captain, who was the pilot flying (PF), and the first officer, who was the pilot monitoring (PM).[1]
The flight crew commenced duty at 0525 Eastern Standard Time,[2] and the aircraft departed Auckland at 0630. As the aircraft approached descent into Brisbane, the flight crew copied the expected arrival procedures and weather conditions from the ATIS.[3] The ATIS, information ‘OSCAR’ (O), stated that arriving aircraft were to expect an instrument approach to runway 01,[4] and that the wind at the threshold of runway 01 was 180° at 5 kt, with a maximum tailwind of 5 kt. The flight crew programed and briefed for an expected SAVER1P standard arrival procedure with an RNAV-P (RNP) RWY 01 approach. The PF reported that, in response to the expected tailwind on final approach and landing, the approach briefing included the importance of ensuring that the aircraft did not get high during the descent and approach.
Prior to commencing the descent, the aircraft was cleared to conduct a SAVER1A standard arrival procedure for an ILS approach to runway 01 (Figure 1).The flight crew reprogrammed and re-briefed the arrival before commencing the descent.
Figure 1: An extract of the Jeppesen SAVER1A chart
The aircraft’s quick-access recorder (QAR)[5] provided information on the aircraft’s flight parameters, autoflight system altitude targets, and autoflight system modes. During the descent, the aircraft was being controlled by the PF through the use of an autopilot, with flight profile changes being achieved through selections on the mode control panel (MCP) (see Figure 2). Under normal procedures, the flight crew were required to verify autoflight system mode changes, such as those selected on the MCP, had been activated through the required mode being displayed on the flight mode annunciator (FMA). The FMA, located just above the primary flight display (see Figure 2), displayed the active flight modes for, from left to right, the autothrottle, roll and pitch.
Figure 2: Flight deck panels identifying the mode control panel with associated controls, and the primary flight display with the flight mode annunciator expanded.
Source: Boeing, annotation by ATSB.
The aircraft commenced descent at 0856 with the autopilot selected on and the autoflight system being selected to the LNAV and VNAV modes.[6] Approaching DUNNI, at 0919 with a descent clearance limit of 5,000 ft, air traffic control (ATC) cleared the aircraft to continue the descent to 4,000 ft. In response, the PF set 4,000 on the MCP and the aircraft continued the descent while tracking towards VETIS. At about 0921:15, a change of controllers commenced at the ATC workstation. The handover to the new controller was completed at about 0922:30. During this period the aircraft was approaching VETIS, and the flight crew reported that they were keen for further descent, being mindful of the desire to ensure that the aircraft did not get high on the descent.
Shortly after passing VETIS, at 0922:45, and while maintaining 4,000 ft, ATC cleared the aircraft to descend to 3,000 ft and for the ILS runway 01. The PF selected 3,000 ft in the MCP altitude window (see Figure 3 at 0922:52) and pressed the altitude selector[7] to initiate further descent. The PF later observed that the altitude selector was probably not properly pressed. As a result, the expected flight mode change did not occur and the aircraft did not commence the descent as expected. The QAR data recorded the autoflight pitch remaining in the VNAV mode, however, the flight crew later reported that the VNAV mode changed from VNAV PTH to VNAV ALT. The PF reported that this pitch mode change was unexpected and unfamiliar. The flight crew operations manual stated the following with respect to the pitch mode entering VNAV ALT:
When a conflict occurs between the VNAV profile and the MCP altitude, the airplane levels and the pitch flight mode annunciation becomes VNAV ALT. The airplane maintains altitude. To continue the climb or descent, change the MCP altitude and push the altitude selector or change the pitch mode.
In response, and with the intent of ensuring that the aircraft did not get high on the desired descent profile, the PF selected the V/S[8] mode, and then the FLCH[9] mode, to initiate the descent. The aircraft commenced descending and at 0923:33 the autoflight system commenced reducing the rate of descent to capture the cleared altitude of 3,000 ft—indicated by the ALT[10] mode activating. Shortly after, and as the aircraft passed through LOGAN, the autoflight system transitioned back into the VNAV mode and then the ALT mode.
Figure 3: Flight data for the period 0922.45 (at about VETIS) to 0925.00 (just before GLENN).
Source: ATSB
As the aircraft was turning towards GLENN, and maintaining the cleared altitude of 3,000 ft, the PF selected 1,000 ft in the MCP and then the FLCH mode with the intent of continuing the descent further. About 5 seconds later, at 0924:03, the aircraft commenced descent from 3,000 ft. At 0924:14, as the aircraft had passed 2,850 ft, ATC instructed the aircraft to maintain best speed, and at least 180 kt until 5 NM final. The PM acknowledged the instruction.
At 0924:35 the PF raised the MCP target altitude to 2,000 ft and changed the flight mode to V/S. At 0924:50, ATC alerted the aircraft that it was cleared to 3,000 ft and that it was descending through 2,200 ft, which was acknowledged. The aircraft levelled at about 2,000 ft and maintained that altitude until it intercepted the glideslope for the ILS. The aircraft landed at 0929 without further incident.
Pilot comments
The PF later reported that, as a result of not properly pressing the altitude select push button to commence the descent from 4,000 ft, the workload experienced increased significantly. This resulted in what the PF described as a loss of situational awareness. After being alerted by ATC that the aircraft was below the cleared altitude of 3,000 ft, and at that time having the runway in sight and being able to maintain visual conditions until landing, the PF decided to maintain 2,000 ft until the aircraft re-joined the approach profile to reduce the workload.
The PM reported that the PF’s actions in descending below the cleared level were not challenged, as this was the first time that the PM had flown this approach.
Operator’s comments
The crew underwent a training session in the B777 simulator to replicate the event and to identify how the error was able to manifest. The training also represented an opportunity to reinforce the use of the various modes for the approach phase.
The RNAV approaches are regularly treated as step down approaches by controllers. The RNAV approaches were designed to reduce pilot and controller workload, however, they regularly increase the pilot workload and increase the opportunity for crew errors. It is also not uncommon to be taken off an RNAV path, given radar vectors then put back on the RNAV at some point in the RNAV path sequence which also increases pilot workload.
Related occurrences
A number of ATSB investigations have examined occurrences that included altitude deviations in flight path involving foreign crew operating within Australia. Of these, two recent examples are summarised below. Full reports are available at the ATSB website.
On the morning of 22 February 2017, a Singapore Airlines Boeing 777-212, from Singapore Changi Airport, Singapore, to Canberra Airport, Australian Capital Territory, was conducting a standard arrival into Canberra when it descended through an altitude constraint associated with the arrival procedure. The error was the result of FMC entry omissions and the flight crew not identifying the relevant altitude limitation.
During a second approach into Perth Airport, Western Australia, an Airbus A320 aircraft operated by PT Indonesia AirAsia was conducting a VOR approach to runway 06 when the flight crew descended the aircraft earlier than normal, but believed that they were on the correct flight path profile. While descending, both flight crew became concerned that they could not visually identify the runway, and focused their attention outside the aircraft. At about that time, the approach controller received a ‘below minimum safe altitude’ warning for the aircraft. The controller alerted the crew of their low altitude and instructed them to conduct a go-around.
Safety analysis
The approach and landing are phases of flight known to have high workload. That workload can increase exponentially when errors and the unexpected occur. As a result of not pressing the altitude selector properly, and then checking that the required flight mode change occurred on the FMA, the PF’s workload increased substantially. This resulted in the PF losing awareness of the aircraft’s descent profile, and in particular the aircraft’s altitude versus the distance to landing, a component of what is commonly referred to as situational awareness. This loss of descent profile awareness, combined with a preconception that the aircraft should not get high on profile during the positioning for final approach, resulted in the PF initiating a descent below the altitude limit of 3,000 ft that was required to be maintained until the aircraft was established on final approach. The PF identified the error and began levelling the aircraft just before the flight crew were notified by ATC that the aircraft had breached the descent clearance limit.
The PM role requires an awareness of the PF actions as well as an awareness of the aircraft’s flight profile relative to the clearance limit and any limitations associated with the approach procedure. While the PM had not flown this approach before, the monitoring role during the approach was not effectively executed.
Findings
These findings should not be read as apportioning blame or liability to any particular organisation or individual.
As a result of high workload, a loss of awareness of the aircraft’s descent profile, and a preconception with ensuring that the aircraft did not get high on the approach flight path, the flight crew initiated a descent below the cleared altitude of 3,000 ft as the aircraft was positioning for final approach. The flight crew corrected the error and levelled the aircraft at about 2,000 ft shortly before being alerted of the altitude breach by ATC.
Safety message
This investigation identifies how an error can increase workload, particularly during a phase of flight that has an already high workload. It also highlights the importance of confirming mode changes on the FMA. Handling of approaches to land continues to be a safety priority for the ATSB.
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
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 27 October 2017, at about 1900 Co-ordinated Universal Time,[1] a Jetstar Airways Airbus A320 aircraft, registered VH-VGY, was being prepared for a scheduled passenger service from Auckland International Airport, New Zealand to Sydney, Australia. The captain was designated as the pilot flying and the first officer was the pilot monitoring.[2]
At about 1909, the leading hand had finished loading the last container into the aircraft hold and was organising his paperwork. As it was raining, he decided to put the clipboard in the right engine (No. 2) cowling to stop his paperwork from becoming wet and blown by the wind, with the intention to retrieve it later. The leading hand went to the flight deck, gave some paperwork to the flight crew, and returned to the ground to organise the aircraft’s push back.
At about 1919, the dispatcher cleared the ground and servicing equipment from the aircraft and conducted the ‘duty of care’ walk-around. During the walk-around, she noticed the clipboard in the right engine and thought that the leading hand would return for it, so she continued with the walk-around. Soon after, the engines reportedly started normally.
At about 1925, when the aircraft was taxiing, the leading hand realised his clipboard with the paperwork was missing and thought the dispatcher had the paperwork. The leading hand asked the dispatcher about the clipboard and she mentioned she saw it in the right engine during the walk-around. The ground crew returned to where they were preparing the aircraft and noticed paper debris on the ground. The ground crew organised for their operations area to contact the flight crew.
At about 1937, the aircraft departed. Shortly after, when on climb through flight level[3] 150, the flight crew received a radio call from the Auckland Approach air traffic controller to contact the surface movement controller. The captain handed control of the aircraft to the first officer and contacted the surface movement controller who advised that the ground crew had lost their paperwork and it may have been placed on the engine. The captain requested further information about the paperwork, specifically whether the paper was on top of the engine or inside the inlet. The flight crew checked the engine instruments and there were no abnormal indications. The surface movement controller confirmed that the paperwork was placed within the inlet and paper debris was found on the tarmac (Figure 1). The captain then contacted the company engineer at the airport and asked whether it was just paperwork or a clipboard with a metal clip. The engineer advised that a piece of sheared metal had been found. The flight crew decided to return to Auckland.
After landing at about 2048, the engine was inspected by engineers and paper was found throughout the engine. They also found minor damage to an engine fan blade and attrition liner.[4]
Figure 1: Foreign object debris on the ground behind the aircraft (circled)
Source: Aerocare, modified by the ATSB
Additional comments
The leading hand stated that, due to the wind and rain, he felt the need to shelter the paperwork. Normally, staff use the pushback tractor for shelter during adverse weather and to prepare paperwork for the flight. There is a metal box on the loader to store the folder. However, as the pushback tractor was not yet present at the bay, he used the engine cowling. He recalled that he did not feel pressured to rush the departure.
The dispatcher stated that she did not view the clipboard as a foreign object as it belonged to the leading hand and had the paperwork for the flight. She assumed that the leading hand would retrieve it later, prior to engine start-up.
The captain stated that, to obtain more information about the incident, numerous calls were made to other agencies, which took considerable time. Further, due to poor communications, he was unable to contact the operator’s maintenance controller to discuss the engine’s status.
Procedures
The internal investigation into the incident by the ground handling operator, Aerocare, noted that the Jetstar Airways operational manual detailed the responsibilities of the dispatcher when conducting the ‘duty of care’ walk-around and provided a table of the steps involved for this process. While there was no specific requirement to check the engine cowlings/intakes for foreign objects, the manual stated that all staff operating near the aircraft were to be constantly observant for abnormalities and to report these to the leading hand or supervisor prior to the aircraft departing.
The investigation also noted that there was no procedure for the ground crew to establish communications with the flight crew in the event of a non-normal or emergency situation, either prior to or after the aircraft had departed. Further, there was no guidance on how paperwork was to be prepared and managed by ground crew during adverse weather conditions.
Previous occurrences
A search of the ATSB database found the following occurrence where a foreign object has been left on an aircraft:
On 3 November 2015, the pilot conducted a maintenance test flight in a Eurocopter AS365 N3 helicopter, registered VH-WPX, at Jandakot Airport, Western Australia (ATSB investigation AO-2015-127). During the post-flight inspection, following this test flight, a licenced aircraft maintenance engineer noticed two large gouges to the leading edge of one of the main rotor blades. A spanner that had been used during the track and balance related adjustments could not be located. It was later located on an adjacent taxiway about 43 m from the hangar. Due to the scuff marks and scratches found on the spanner, it was determined that it had been left in the rotor head area and was likely ejected during engine start up.
Findings
These findings should not be read as apportioning blame or liability to any particular organisation or individual.
While preparing the aircraft for departure, the leading hand placed a clipboard in the right engine, which was subsequently ingested during start-up.
During the walk-around, the dispatcher noticed the clipboard in the right engine, but believing it would be retrieved prior to the aircraft departing, the dispatcher did not notify the leading hand or supervisor of the foreign object debris as per company procedures.
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.
Safety notices
Both Jetstar Airways and Aerocare have released a notice, which outlines that foreign object debris also includes items accidentally left behind. It further states that engines are not to be touched or used for the placement of items, and emphasises the responsibilities of ground crew to manage foreign object debris by clearing them and reporting their presence to other crew.
Updated procedures
As a result of this incident, Jetstar Airways released an updated aircraft dispatch procedure, which included:
a specific warning about not placing items in the engine cowling
improved detail around checks and responsibilities
a section on emergency and non-normal procedures
detailing methods for re-establishing communications between ground crew and flight crew such as visually gaining the attention of the flight or contacting them via radio.
Safety message
The presence of foreign object debris poses a significant threat to aircraft safety. It has the potential to cause aircraft damage during critical phases of flight, costing airlines and airports millions of dollars each year. This incident demonstrates the effect foreign object debris has on aircraft operations and emphasises the importance of not placing objects in aircraft engines. It further highlights that all staff operating near aircraft are responsible for reporting any non-normal events they encounter. It should not be assumed that others will perform a task where a hazard has been identified. Assuming other people will undertake a task, such as removing a hazard increases the risk of the task not being completed.
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
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.
