On 15 April 2016, the instructor and student of a Pacific Aerospace CT/4B aircraft, registered VH-YCO (YCO), conducted a dual (military) training flight under the instrument flight rules.[1] The aircraft departed from Tamworth Airport, and tracked to Narrabri Airport, before commencing the area navigation (RNAV) runway 11 approach to Gunnedah Airport, all in New South Wales.
Prior to commencing the approach, the instructor of YCO contacted the military radar controller (see Military radar control), who advised that they were not aware of any other aircraft in the area. The instructor reported that they broadcast on the Gunnedah common traffic advisory frequency (CTAF), when 18 NM from the aerodrome, advising that they were going to conduct the RNAV runway 11 approach, and stating their estimated time at the airport (Figure 1). The crew reported hearing a voice-back response from the aerodrome frequency response unit,[2] indicating that no one else had transmitted on the CTAF in the previous 5 minutes, and no response from any other aircraft on the CTAF.
At that time, an instructor and student pilot of a Piper PA-28-181 aircraft, registered VH-WJZ (WJZ), were conducting a local dual training flight from Gunnedah Airport. The instructor reported that they were broadcasting on and monitoring the CTAF.
At about 1450 Eastern Standard Time (EST), YCO was 13 NM north-west of Gunnedah on the RNAV approach for runway 11, and at 4,500 ft above mean sea level (AMSL), when they were alerted by the aircraft’s traffic collision avoidance device[3] of another aircraft. The device indicated that the other aircraft was 200 ft below them and 2 NM away. The instructor looked south and within 20 seconds sighted WJZ at the same level. The instructor took control of the aircraft from the student, and conducted a 60° angle of bank turn to the left to avoid WJZ.
The instructor of WJZ sighted YCO when about 13 NM north-west of Gunnedah aerodrome at about 4,000 ft AMSL. YCO was then to their north in their 2 o’clock[4] position. The instructor of WJZ conducted a left turn and reported sighting YCO commence a left. The aircraft passed at the same level about 150 to 200 m horizontally apart.
The instructors of the two aircraft subsequently communicated on the CTAF. The instructor of WJZ reported that they had not heard any relevant calls on the CTAF leading up to the incident.
Figure 1: Recorded track of VH-YCO, approximate track of VH-WJZ, and approximate location of near collision
Source: Instructor of VH-YCO – annotated by ATSB
Military radar control
The ADF established and operated a radar system in the vicinity of Tamworth to provide directed traffic information (DTI) to pilots operating CT4B aircraft in the Tamworth training areas. This was an interim measure to treat the risk of separation breakdown in the Tamworth training areas while those training areas and associated procedures underwent redesign and other systems were established. The DTI was provided on a discrete frequency monitored only by ADF aircraft. Directed traffic information was an advisory only service and controllers did not provide control or direction to pilots, but gave information aimed at increasing their situational awareness.
The military controller did not report any conflicting traffic before or during the incident.
Traffic collision avoidance device
YCO was fitted with a traffic collision avoidance device (TCAD), which warns of the presence of threat aircraft if the other aircraft is fitted with a functioning transponder that is being interrogated by a radar transmitter. The TCAD detects Secondary Surveillance Radar (SSR) transponders in aircraft within a certain proximity.
The TCAD displays threats detected within a predetermined volume of airspace known as a shield. The shield setting for the TCAD during the incident flight was +/- 1,000 ft in altitude and 2 NM. When a detected aircraft enters the pre-set shield, the pilot is alerted via aural and visual indications.
The TCAD system will not detect a threat aircraft that is not equipped with a transponder, the transponder is inoperative, or the transponder is operating but not being interrogated by either an SSR or a TCAS fitted aircraft. The TCAD is designed as an aid to situational awareness and should not be relied on for traffic separation.
In this incident, the TCAD identified WJZ as a threat and alerted the pilots of YCO.
Pilot comments
Instructor of VH-WJZ
The instructor of WJZ commented that the other aircraft was conducting military training and they have a radar in the Gunnedah training area, to provide them with traffic warnings. The military radar service did not identify any conflicting traffic, however, YCO’s TCAD identified WJZ, which indicated that WJZ’s transponder was functioning.
The instructor reported that there was some distortion in the broadcasts from YCO heard after the incident.[5]
The instructor further commented that in future, they would broadcast their position in the training area every 15 minutes; even if there were no broadcasts to indicate there may be nearby aircraft.
Instructor of VH-YCO
The instructor of YCO commented that if pilots of aircraft conducting instrument approaches broadcast their aircraft’s position with reference to a compass, this may assist visual flight rules’ pilots to assess whether there could be a conflict.
Safety action
Whether or not the ATSB identifies safety issues in the course of an investigation, relevant organisations may initiate safety action in order to reduce their safety risk. The ATSB has been advised of the following safety action in response to this occurrence.
Operator of VH-WJZ
As a result of this occurrence, the operator of VH-WJZ has advised the ATSB that they are taking the following safety actions:
Airspace procedure
As the instrument approach lies within the Gunnedah training area, student pilots are required to remain below 3,000 ft when within 5° of the approach path. Additional risk assessments were conducted at the other company flying school locations and a similar hazard was found at Scone, where similar de-conflicting provisions have been made.
Safety message
The ATSB report Limitations of the See-and-Avoid Principle outlines the major factors that limit the effectiveness of un-alerted see-and-avoid. In this occurrence, un-alerted see-and-avoid did lead to the instructor of one aircraft sighting the other. However, insufficient communication between pilots operating in the same area is the most common cause of safety incidents outside controlled airspace and near non-controlled aerodromes. A broadcast that does not provide a clear understanding of the location of an aircraft, or the intentions of the pilot, is often ineffective in directing other pilots where to focus their lookout.
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 13 April 2016, an instructor and student of a Jabiru J170-D aeroplane, registered 24-7750 (7750), conducted a local training flight from Bathurst Airport, New South Wales. At about 1442 Eastern Standard Time (EST), as they were returning to Bathurst, the instructor broadcast on the Bathurst common traffic advisory frequency (CTAF) that they were inbound from the south-west, and added that they were estimating arrival in the circuit at 1446. As they subsequently arrived in the circuit, the instructor broadcast that they were joining the circuit on an early downwind for runway 17, for a full-stop landing.
The wind was from the east-south-east. Powered aircraft were operating on runway 17 and gliders (and towing aircraft) were operating on runway 08. Bathurst aerodrome elevation is 2,435 ft above mean sea level (AMSL) (Figure 1).
About a minute after broadcasting their arrival in the circuit, the pilot of 7750 asked Glider Ground[1] how many gliders were in the air. Glider Ground advised that there were ‘two gliders, NGH and NDQ, just thermalling,[2] at 4,000 ft off the threshold of runway 26.’ The pilot of 7750 confirmed sighting two gliders.
Meanwhile, a student pilot of a Glaser-Dirks DG-1000S glider, registered VH-NDQ (NDQ) was conducting a solo flight at Bathurst. The student had been briefed prior to the flight to make a downwind call, stay close to the runway in use by the gliders, and to keep a good lookout. At about 1449, about 90 seconds after the pilot of 7750 had communicated with Glider Ground regarding glider traffic in the air, the pilot of NDQ broadcast on the Bathurst CTAF that they were on left downwind for runway 08.
Immediately following the downwind call by the pilot of NDQ, the pilot of 7750 broadcast that they were on left base for runway 17, and soon after, broadcast that they were on final approach to runway 17 for a full stop landing. The pilot of NDQ reported hearing both those broadcasts, but did not make any broadcasts or directed radio calls in response.
After 7750 touched down on runway 17, about 100 m before the intersection with runway 08, the pilot sighted a glider (NDQ) on short final for runway 08, at an estimated 100 ft above ground level. The pilot assessed that they did not have sufficient time to stop before the intersection of runway 08, so applied full power to cross runway 08 as quickly as possible.
When at about 500 ft above ground level and on final approach to runway 08, the pilot of NDQ sighted 7750 their 10 o’clock[3] position at about the same altitude. As 7750 landed, the pilot of NDQ assessed that there was the potential for a collision, closed the glider’s airbrakes[4] and initiated a climb to pass over 7750. As the glider passed over 7750 near the intersection of the two runways, the pilot of NDQ heard the aircraft’s engine increase power. The glider then landed ahead on runway 08 (Figure 1).
The instructor in 7750 lost sight of NDQ as it passed overhead. As 7750 accelerated with a high power setting, the instructor elected to continue the take-off. The pilot of 7750 then conducted a circuit before landing safely.
Figure 1: Layout of Bathurst aerodrome showing indicative tracks of 7750 and NDQ
Source: Airservices Australia – annotated by ATSB
Pilot comments - Pilot of 24-7750
The pilot of 7750 commented that the circuit was very busy at the time of the incident. They were maintaining a good lookout and listening intently to the CTAF for positional information from the gliders, noting that gliders would have ‘right of way’ over powered aircraft. During final approach to runway 17, the instructor was communicating with the student in 7750 for teaching purposes.
The pilot also commented that they now discuss operational intentions with the glider operator at the commencement of each day’s operations.
Safety message
Simultaneous operations on crossing runways can be problematic, particularly where the volume of traffic is high and where the nature of the potentially conflicting operations are dissimilar (such as powered flight and gliding operations). Organisations responsible for the coordination and conduct of such activities are encouraged to carefully assess and manage the risks involved. This is particularly important when operations are likely to involve instructional flights and relatively inexperienced pilots, where workload and the potential for pilot distraction may be elevated.
This incident highlights the importance of effective communication. The primary purpose of communications on the CTAF is to ensure the maintenance of appropriate separation through mutual understanding by pilots of each other’s position and intentions. Where a pilot identifies a risk of collision, that pilot should alert others as soon as possible to allow a coordinated and effective response.
Civil Aviation Advisory Publication 166-1(3) stated that ‘whenever pilots determine that there is a potential for traffic conflict, they should make radio broadcasts as necessary to avoid the risk of a collision’.
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 6 April 2016, Pacific National (PN) grain train 5422N parted at Parkville on the Main North Line. A fractured yoke on the trailing end of the fifth wagon allowed the yoke pin to fall out and the coupler shank to disengage from the wagon. The coupler shank fell into the four foot of the track and damaged the underbelly discharge doors on 10 wagons as the train progressed over the coupler. Approximately 10 tonnes of barley was released from the wagons as a result. The trailing wheelset on the thirteenth wagon was derailed, then ran in a derailed state for approximately 45 metres causing damage to several track sleepers before the train came to a stop. Regular services were able to continue under caution past the incident site via the adjacent crossing loop. Wagon recovery and track repairs was completed without any further incident.
What the ATSB found
A draftgear component (the yoke), that was not compliant with a PN maintenance standard, was not identified during a maintenance inspection and re-entered service undetected. The yoke was an earlier design and susceptible to fatigue failure. PN had identified the issue and completed a programme to replace this yoke design across their grain wagon fleet however, the yoke on this wagon had been overlooked.
What's been done as a result
PN issued an additional Rolling Stock Notice to their maintenance teams to advise of the incident and to mandate that all yokes in the grain wagon fleet were checked to ensure no yoke of this design remained in service.
Safety message
Maintenance systems must ensure that when a non-compliant component is identified as needing replacement, all such units are located and replaced.
Figure 1: 5422N at Parkville
Source: ATSB
The occurrence
At 0230 on 6 April 2016 PN grain train 5422N parted at Parkville (approximately 322.350 km[1]) on the Main North Line, Hunter Valley, NSW. The train parted due to a fractured yoke (in the draw gear assembly) on the trailing end of the fifth wagon (NGKF 35898X). The crew of 5422N comprised of a driver and second person.
The fractured yoke allowed the yoke pin to fall out and the coupler shank to disengage from the wagon. The coupler shank fell into the four foot of the track and damaged the underbelly discharge doors on 10 wagons as the train progressed over the coupler. Approximately 10 tonne of barley was released from the wagons as a result.
The trailing wheelset on the thirteenth wagon was derailed when the fallen coupler struck the wagon’s fourth axle. The wheelset ran in a derailed state for approximately 45 metres before both parts of the train came to a stop. A number of track sleepers were damaged during the incident.
The driver notified the Network Control Officer Upper Hunter 2 (NCO) that 5422N was stopped approximately 80 m from signal 06-12M. The driver told the NCO, he had experienced a sudden loss of air from the train which led him to believe the train had parted.
The second person conducted a walking inspection of the train and radioed the driver that the train had parted behind the sixth wagon and he could not see the remainder of the train.
The driver advised the NCO that an adjacent train in the Parkville crossing loop was not to move as he could not confirm the location of his detached train. The NCO told the driver of 5422N that he would warn the other train of the incident. The NCO also advised the driver that based on his indicator board, the two portions of the train were between the two yard signals and the line was clear for the other train to come out of the crossing loop.
With that knowledge, the NCO cleared the signal for the other train to depart from Parkville crossing loop. The second person continued his inspection and found the detached portion of 5422N and the flashing tail light marker[2] on the final wagon, confirming the train was clear of the yard signals.
The driver asked the NCO for confirmation of track protection for 5422N. The NCO confirmed that protection of the train was being provided by the yard signals being blocked[3].
The second person secured the detached portion of 5422N by applying the handbrakes on all wagons. The driver secured the locomotives and the front wagons of the train.
At approximately 0400 the NCO informed the driver that a Condition Affecting the Network (CAN) had been issued.
The driver and second person were post incident drug and alcohol tested and ceased duties for the rest of their shift in accordance with PN procedures. Both crew members returned a negative result from the post incident testing.
Site clean up started at approximately 1500. PN arranged for the damaged wagons and the spilled barley to be cleared from the site. PN then shunted the two portions of 5422N and re-railed the derailed wagon.
At 0210 on 7 April 2016 Australian Rail Track Corporation (ARTC) track maintainers were able to commence emergency track repairs and were completed by 0600. In accordance with ARTC procedures, a temporary speed restriction of 40 km/h was placed on the main line through Parkville to allow operations to continue.
Parkville is located between Scone to the south and Murrurundi to the northwest (see Figure 2). Train 5422N was travelling from Werris Creek to Carrington Grain Terminal near Newcastle on the Main North Line. The incident occurred adjacent to the township of Parkville in the Upper Hunter region of NSW.
Figure 2: Incident location map
Source: Geoscience Australia, annotated by ATSB
Rail infrastructure information
There are two tracks at Parkville, namely a main line and a crossing loop. A single, bidirectional line connects Parkville and Murulla to the north. The track from Murulla to Parkville is predominantly on a falling grade or downhill (see figure 3). Murulla sits between Murrurundi and Parkville at approximately 339.4 km.
Figure 3: Curve and Gradient diagram of track between Murulla and Parkville
Source: ARTC, annotated by ATSB
Safeworking system
The safeworking system in place was Centralised Traffic Control (CTC). In this system, lineside signals and associated track circuits allow the NCO, from the control centre at Broadmeadow, see the location of any given train operating along the Main North Line.
Train information
5422N was a loaded grain train consisting of 40 wagons and hauled by three locomotives. It was approximately 636 m long with a mass of 3040 tonne excluding the locomotives. It was typical of many grain trains working throughout NSW.
The yoke and draw gear assembly
The yoke is an integral component of a wagon’s draw gear assembly. While a number of different designs are in service, they are all designed to retard the train’s longitudinal draft and buffering[4] forces. The major components of a draw gear assembly are typically: coupler shank, buffer package, yoke and yoke pin (see Figure 4).
The yoke houses the buffer and is connected to the coupler by a pin or key (in this case a pin). The yoke is a part of the draw gear and transmits draft and buffering forces between wagons. Under compression, the force is transferred between the back-strap of the yoke and the wagon. In operation, the back-strap is subjected to repetitive loads. To mitigate against this effect, later yoke designs (post-1980) have internal rounded smooth radius fillets at high stress locations which are better able to withstand the cyclic stresses that are imposed on these areas. The later yokes are manufactured from a higher tensile steel.
The yoke in this incident was a pre-1980 design made from a lower tensile steel and without rounded smooth radius fillets in the high stress locations.
PN has maintenance systems in place to manage it’s total wagon fleet of approximately 15,000[5] wagons, across several locations in Australia.
PN was aware that the early design yokes (in grain wagons manufactured before 1980) needed to be removed from their fleet and replaced with later design yokes made from higher tensile steel and with internal rounded smooth radius fillets.
As such, specific instructions in PN’s Wagon Maintenance Manual (WMM) had been developed to provide instruction to PN’s wagon maintainers to identify and change out the early-design yokes as the wagons came in for preventative maintenance.
Site observations
ATSB observed the train had parted five wagons behind the locomotive (see Figure 1). The draw gear assembly was missing from the rear of the fifth wagon, NGKF 35898X (see Figure 5). The coupler shank was found under the rear axle of the rear bogie of the sixteenth wagon, NGKF 35864R (see Figure 6).
Figure 5: Location for draw gear assembly
Source: ATSB
Figure 6: Coupler shank in four foot
Source: ATSB
The bottom discharge doors on 10 wagons had been damaged. Approximately 10 tonne of barley had spilled from the wagons (see Figure 7).
Figure 7: Wagon doors and spilled grain
Source: ATSB
The coupler shank struck the axle of the trailing wheelset on the thirteenth wagon, NGPF 36032J. This resulted in the wheelset derailing near the crossing loop (see Figure 8). The wheelset travelled in a derailed state for approximately 45 m, damaging a number of concrete sleepers, before coming to a stop. (see Figure 9).
The ATSB found that a draftgear component (an older design yoke), that was not compliant with a Pacific National (PN) maintenance standard, and not identified during a periodical wagon maintenance inspection and the yoke re-entered service undetected.
PN had a system in place to remove this early design yoke from their grain wagon fleet and replace it with a later design yoke better suited to the application. However, this yoke had been overlooked by the wagon maintainers and consequently continued in service up until its failure on 6 April 2016.
The yoke fitted on the rear of the fifth wagon, NGKF 35898X, was of an early design and was known to be susceptible to fatigue failure. The undetected older style yoke eventually failed in-service and led to the train parting and the subsequent derailment (see Figure 10).
While the train data logger information indicated that some brake applications on 5422N did not comply with PN procedures, ATSB concluded train handling did not contribute to the incident. After the yoke pin fell out of the assembly, the downhill operation most likely kept the two portions of the train together until the train negotiated a slight uphill gradient at Parkville. Here the train experienced a longitudinal tensile load that allowed the two portions of the train to separate.
The track leading to the incident site was inspected by a track specialist on behalf of PN. It was concluded there were no track defects or anomalies that could have contributed to the incident.
Figure 10: Event sequence
Source: ATSB
Yoke failure
The yoke from wagon NGKF 35898X had a manufacture date stamp ‘10/71’. NGKF wagons were manufactured in 1976. The yoke was approximately five years older than the wagon itself. It was therefore, likely that this yoke was the original yoke fitted to this wagon.
The yoke design had recently been identified by PN as having a finite life due to a propensity to fatigue failure. The tensile strength of the parent metal and the lack of internal rounded relief fillets in high stress areas of the yoke contributed to this issue.
The yoke had fractured at the rear section where the top and bottom longitudinal straps connected with the backing face (see Figure 11).
Figure 11: Fractured yoke
Source: Bureau Veritas, annotated by ATSB
PN commissioned Bureau Veritas (BV) to conduct a metallurgical examination to identify the nature of yoke’s failure mechanism. The examination identified that the yoke had failed due to fatigue cracking which had initiated at multiple points on the inner face of the top yoke strap (see Figure 12). Fatigue cracking had progressed through approximately 85 per cent of the material’s cross-section before the ductile failure of the remaining material. The bottom yoke strap subsequently failed with approximately 10 per cent of the material’s cross-section subject to fatigue cracking before the ductile fracture, failing after the top arm failed.
The BV report stated that:
‘it was suspected that the fracture may be due to a material with low toughness properties[6] being in service for a long time, which was exposed to cyclic stress.’
The BV metallurgical examination focussed on the material properties and while it did not review the design of the component, the report did note that “stresses are concentrated at the critical regions (e.g. radius relief groove, curves, thickness change)”.
Figure 12: Fracture faces of yoke
Source: Bureau Veritas including annotation
Fatigue cracking in yokes
Other research has been conducted into fatigue cracking and stresses in yokes in heavy haul services[7]. Cookson et al 2013, identified a propensity for fatigue cracking in yokes due to the vertical oriented bending of the draft system. This vertical oriented bending in yokes was considered most likely due to misalignment in the draft system. As such, identifying misalignment in the draft system was considered important as it could affect the overall service life of the yoke.
It was noted that PN’s Wagon Maintenance Manual (WMM) 07-02_09 (dated 7 May 2009) Couplers and Draft Gear Maintenance makes reference to checks for alignment when inspecting varous parts of the draft system, including; the knuckle, coupler shank, the yoke strap and drawgear carriers.
Wagon maintenance
The periodic maintenance program in place for PN’s grain wagons was developed to ensure reliable and safe operation of the wagons.
WMM 07-02_09 sets out the maintenance requirements for wagon couplers and draft gear. Section 1.9 deals with yokes and it includes the following:
Yoke straps are to be inspected for cracking
Where video/borescopes are available these shall be used to detect cracking in yoke straps.
If cracking is found (in the internal radius locations) the yoke shall be scrapped
Further, Section 1.9.4 makes specific reference to earlier design yoke straps:
‘Wagons manufactured before 1980 used lower grade steel in the draft gear castings and do not have a relief fillet in the rear top and bottom inside corners as identified in the picture below (see Figure 13) (i.e. sharp radius). Replace these when found.’
Additionally, while wagons with a draft capacity of less than 1.3 MN were excluded from this last requirement, NGKF 35898X had a draft capacity of 1.8 MN.
These requirements specified in PN’s WMM suggests the yoke on wagon NGKF 35898X should have been identified and replaced at the last maintenance intervention.
Figure 13: Yoke inspection areas
Source: PN WMM 07-02
It is worth noting the task requirements on wagon maintainers when required to inspect a unit train consist[8]. When a unit train consist is scheduled for maintenance, such as the B inspection[9], the consist is scheduled into a maintenance centre for usually 24 hours before being scheduled onto its next service.
In this time the wagon maintainer would need to attend to a list of items and either, test, service or lubricate, replace, adjust, gauge, measure or inspect each of these items on each wagon (for grain unit train consists, 40 wagons). The list of items to attend to on the B inspection sheet is 92 items.
Additionally, the location of the yoke in situ makes it challenging for the wagon maintainer to identify cracking in the yoke. As pictured in Figure 14, the wagon maintainer is required to inspect the highlighted areas. As Figure 15 shows, the access to view these areas while the yoke is in situ is difficult and potentially restricted by time constraints in the maintenance opportunity.
Figure 14: Yoke inspection areas
Source: PN WMM 07-02
Figure 15: Yoke in situ
Source: ATSB
Train handling
The yoke pin was found nearly 6 km before Parkville. There was a constant falling grade towards Parkville and the train driver was using dynamic braking to control train speed. Therefore, the wagons of the train generally remained in a compressed state ensuring the train did not part.
On account from the driver, there is a slight incline after Wingen just prior to the downhill into Parkville. At this point the driver allowed the train to roll freely through the section of track to maintain speed. This coincides with a releasing of the buffering or compressive forces in the draw gear. The yoke pin most likely fell from the train at this point which is where it was found during inspection post incident (see Figure 16).
The train was able to continue without a mechanical connection through to Parkville as compressive forces kept the train and brake air hoses together on the downhill into Parkville. However, as the train rolled into Parkville at approximately 60 km/hr, the driver reduced dynamic braking to account for the slight incline to the middle of the yard which is when the train parted, seperating the air hoses. This enabled the brakepipe to vent to atmosphere, triggering an emergency brake application which brought both parts of the train to a stop.
The train handling in the lead-up to the incident was not considered a contributing factor to the incident.
Figure 16: Yoke pin lying in the four foot
Source: ARTC including annotation
Previous incidents
PN reported that, over the five years preceding the incident, there were 14 incidents of trains parting due to defective yokes. These incidents occurred between 2011 and 2016. There were a further 13 wagons that had their yokes replaced during maintenance between 2011 and 2012. No older style yokes were detected after that date until this incident in 2016.
Pacific National investigation
PN conducted a safety investigation into the incident at Parkville. PN’s internal Investigation Report (Form PN-FOR-SAF) dated 10 May 2016 identified that the incident wagon (NGKF35898X)
‘underwent regular maintenance and inspection as per WMM 01-01b_05 Bulk Services Division Services Schedule of Inspections; however, the early model yoke strap on this wagon was not identified or replaced (as per WMM 07-02_09).’
Two of the preventative actions in the report were:
To develop and issue a Rolling Stock Notice (RSN) to all wagon maintenance and engineering teams to reiterate required inspections on early model yokes on wagons and replacement of specified parts.
Develop a safety alert detailing the contributing factors for derailment and distribute to NSW/Vic operations.
ATSB requested PN to demonstrate that the actions had been completed. PN notified the ATSB that the first two items had been closed out by issuing a single RSN (see Appendix A) to wagon maintainers.
From the evidence available, the following findings are made with respect to the derailment of grain train 5422N at Parkville, NSW on 6 April 2016. These findings should not be read as apportioning blame or liability to any particular organisation or individual.
Contributing factors
The yoke was an earlier design prone to failure through fatigue cracking.
The presence of the earlier design of yoke on wagon NGKF 35898X was not detected during preventative maintenance activities.
Other factors that increased risk
Nil
Safety issues and actions
Whether or not the ATSB identifies safety issues in the course of an investigation, relevant organisations may proactively initiate safety actions in order to reduce their safety risk. The ATSB has been advised of the following safety actions in response to this occurrence.
The presence of the earlier design of yoke on wagon NGKF 35898X was not detected during preventative maintenance activities.
Sources and submissions
Sources of information
The sources of information during the investigation included:
The Australian Rail Track Corporation Ltd.
Office of the National Rail Safety Regulator
Pacific National
Submissions
Nil
Appendices
Appendix A – Pacific National Rolling Stock Notice
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 4 April 2016 at approximately 0505 Australian Eastern Time (EST),[1] lead locomotive NR8 on train 7YN2 caught fire near the southern side of Cardiff railway station, New South Wales (NSW). The crew stopped the train 250 metres north of Cardiff station after they were alerted to the fire by flames emanating from the locomotive’s engine cab. The crew informed Network Control[2] they could not contain the fire and requested the assistance from NSW Fire and Rescue (NSWFR). At 0510, the crew applied the locomotive park brake and both crew members evacuated without injury. The crew further secured the train by applying the hand brakes on the first 10 wagons and isolated the other two locomotives. NSWFR attended at 0525 but could not commence extinguishing the fire due to the overhead electrical wiring remaining energised. At 0535 an engine explosion displaced a number of engine crank case inspection covers. At 0543 passenger service N116 was stopped from passing the burning locomotive and passengers were detrained to alternative transport. At 0635 the overhead wiring was de-energised and by 0830 NSWFR had extinguished the fire.
The ATSB found that two of the 10 oil filter tank cover bolts had failed beneath the retaining nut due to fatigue and overload fractures. There was evidence of fretting damage on the tank cover at the failed bolt positions. This was consistent with a loose bolted joint associated with in service movement of the assembly. The fire ignited when the displaced tank cover enabled high pressure oil to escape and come in contact with hot engine components (Figure 1).
Figure 1: Fire in lead locomotive NR8 passing through Cardiff Station
Source: Sydney Trains CCTV
The oil filter tank is a part of the locomotive engine’s lubricating oil system (Figure 2 and 3). The lubricating oil system capacity is approximately 1500 l. There was approximately 800 l of lubricating oil left in the system after the incident.
Figure 2: Schematic of engine oil lubricating system
Source: GE annotated by ATSB
Figure 3: NR8 engine removed from bay (L) and oil filter tank in situ (R)
Source: ATSB
Safety analysis
Fires can be unique in their ignition and progress, however, organisations need to prepare their staff to deal with the early stages of the fire so they can minimise escalation of the fire.
Failed components
An independent metallurgist established that the 19 mm diameter oil filter cover bolts (Figure 4) had suffered unidirectional fatigue failure through the root diameter of the thread. The fractures occurred directly below the nut adjacent to the cover’s nut seat. The nut seat exhibited fretting damage, which suggested there was movement across the bolted joint and most likely attributed to excessive torqueing of the bolts.
Figure 4: NR8 oil filter tank and cover
Source: ALS annotated by ATSB
The fracture surfaces of the broken bolts suggested the bolts had been moving transversely to the plane of the cover causing micro fractures at the root of the thread adjacent to the nut seat. Each movement of the bolt would have increased the size of the fatigue fractures and progressively reduce the bolt’s tensile strength.
Bolt eight exhibited a small central region ductile fracture surrounded by fatigue fractures emanating from the root of the thread. Bolt nine had a larger ductile fracture surrounded by a smaller area of fatigue fractures. The central region ductile fractures corresponded with the final failure of the bolts (Figure 5 and 6).
The area of the ductile fracture on bolt number eight was smaller than on bolt number nine suggesting bolt eight had fatigued more and failed first. Bolt nine most likely failed after bolt eight.
Figure 5: Bolt number 8 (Top) and 9 (Bottom) with cross sections
Source: ALS annotated by ATSB
Evaluation of the other eight bolts indicated similar fatigue fracturing in similar locations.
There was four engine crank case inspection covers displaced during the occurrence. This was likely due to a secondary fire and subsequent minor explosion in the crankcase.
Postfire response
The crew had on board one 4.5 kg fire extinguisher located in the cab of the locomotive and two larger 9 kg fire extinguishers located in the radiator compartment. All fire extinguishers were dry powder suitable for fuel, oil and electrical fires.
When the driver noticed the fire he immediately applied the emergency shutdown of the locomotives and called train control. Whilst on the radio to train control he made an assessment of the fire to determine if he was able to put it out but confirmed NSWFR would be needed. The driver then applied the park brake before exiting the locomotive with the second person and stood in a safe place away from the burning locomotive.
Approximately 15 minutes later NSWFR arrived at the nearby overhead bridge (Newcastle Street). Whilst the driver discussed the situation with NSWFR the second person applied hand brakes to 10 wagons to ensure the train was stabilized. The driver and second person then isolated the remaining two locomotives (NR90 and NR103). NSWFR commenced set up but could not start due to the live overhead lines.
The isolation of the overhead lines was confirmed at 0548 approximately 23 minutes after notification. It took a further 47 minutes before residual current was confirmed out of the overhead lines. NSWFR then commenced putting out the locomotive fire. All parties had ensured they could approach the burning locomotive only when it was safe to do so.
Network Control informed scheduled passenger service N116 of the locomotive fire at Cardiff at 0531 he informed the driver to approach at caution and provide advice on the situation on arrival. At 0543 N116 was on approach to the incident site when the site incident commander stopped the train and informed the driver of N116 that the train would need to be detrained as passing the site was not possible at that stage.
Findings
These findings should not be read as apportioning blame or liability to any particular organisation or individual.
Oil filter cover retaining bolts eight and nine had failed through the first thread beneath the nut by a unidirectional fatigue failure mechanism.
The pattern of fretting damage on the cover nut seat suggested relative movement was evident between bolts eight and nine and the oil filter cover.
Safety action
In response to the retaining bolt failures, Pacific National has advised the ATSB that they have completed the following actions:
Developed and implemented a bolt tightening and removal procedure
Reduced the bolt replacement frequency from 732 days to 366 days
Retrofitted a guard in NR locomotives to minimise the risk of oil escaping from the filter tank and contacting hot engine components.
Other finding
The postfire response from all parties was managed in accordance with procedures.
Safety message
Correct tensioning of bolts can be a critical aspect of ensuring equipment functions as designed. Operators should ensure their safety management systems provide appropriate procedures for bolt inspection and tensioning where there is a risk of equipment failure, especially when the consequence can escalate to fire.
The ATSB SafetyWatch highlights the broad safety concerns that come out of our investigation findings and from the occurrence data reported to us by industry.
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 17 April 2016, a Boeing 787-9, registered N36962, operated by United Airlines as flight UAL870, departed Sydney, New South Wales (NSW), for San Francisco, United States. On board were 4 flight crew, 11 cabin crew and 238 passengers. During the departure, cabin crew switched on the aft galley ovens (Figure 1) in preparation for meal services.
After the two ovens were switched on, there was a short burst of smoke, which set off a fire alarm in a nearby toilet for about one minute. One of the ovens displayed a “FAILURE” message. Several cabin crew detected a strong chemical odour and an electrical smell, as well as a blue haze. Other crew described it as an ozone smell. The oven interactive screen displayed a ‘Critical Error- Broken Fuse’ message.
The crew immediately pulled all relevant circuit breakers, and switched off all electrical sources to the aft galley. The inflight service manager (ISM) advised the captain. The ISM and a relief pilot from the cockpit arrived at the aft galley with fire extinguishers. By this stage, the smoke had dissipated, but the odour persisted. As it could not be confidently ascertained that the ovens were the sole source of the problem, the captain contacted the ground-based technical operations maintenance controller (TOMC) by satellite phone.
Figure 1: Rear section of a B787-9 depicting aft galley
Source: SeatGuru modified by the ATSB
The discussion with the TOMC involved all flight crew and the ISM. It was agreed that the safest option was to return the aircraft to Sydney. The captain advised ATC by a PAN[1] call. ATC initiated an INCERFA[2] phase. About 110 km east of Port Macquarie, NSW, the crew commenced a return to Sydney. As the aircraft was well in excess of its allowed landing weight, fuel was dumped during the descent.
The aircraft landed without incident in Sydney at 1258 Eastern Standard Time (EST) with emergency services in attendance.
Post-incident engineering report
A post-engineering inspection quarantined the suspect oven, and after an inspection, a fuse was replaced. After appropriate testing, the aircraft was released back to service.
Boeing and the oven manufacturer investigated the cause of the ‘Critical Error’ fault displayed on the oven screen (Figure 2).
The manufacturer individually tested all oven components. They reported that all individual components worked correctly, however, an additional measurement of the oven motor current detected that the motor did not run smoothly. The motor temperature was also above normal, most likely from insufficient airflow. This known fault had been rectified with a new oven software release.
Boeing reported that the oven manufacturer is working with United Airlines to update the software in all relevant ovens in their fleet.
The exact cause of the odour could not be determined.
A second similar occurrence
United Airlines have advised the ATSB of a second similar occurrence involving another B787 aircraft. On 2 June 2016, a United Airlines B787 aircraft, N35953 experienced an electrical/heat odour in the mid B galley. The flight crew dumped excess fuel and returned safely to Melbourne. On this occasion, no emergency was declared.
Maintenance were able to isolate one oven, and confirmed the error was a broken fuse. The oven was removed and replaced, and the aircraft returned to service.
Figure 2: Error message from oven on N35953
ATSB comment
As part of the investigation, the ATSB obtained reports from the flight crew and cabin crew on board during the incident.
It was evident that all emergency procedures were carried out efficiently and effectively. The captain involved all relevant crew members and the TOMC prior to making a decision to return the aircraft to Sydney.
Safety message
This incident highlights the correct management of an abnormal situation with effective crew coordination. Each crew member responded effectively and the situation was professionally managed by the captain.
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 afternoon of 13 January 2016, the roll-on/roll-off passenger ship Spirit of Tasmania II was loading cargo, vehicles and passengers at Station Pier, Melbourne. At 1752, strong wind gusts blew the ship off the wharf and all but two of the ship’s mooring lines (on the bow) parted. After breaking away, the stern swung around until the ship was 90 degrees to the wharf, parallel to nearby Port Melbourne Beach and in danger of grounding. While waiting for tugs to assist, the ship’s propulsion and thrusters were used to maintain its position and prevent grounding. By 1905, the ship was back alongside the wharf, assisted by two tugs.
The ship suffered minor damage to its lower bow ramp and bow doors. Shore infrastructure suffered extensive damage to the elevated roadway and ramp arrangement on the wharf and minor damage to wharf structures. No one was injured.
What the ATSB found
During the afternoon of 13 January, a band of severe thunderstorms passed across the location of Spirit of Tasmania II, with little warning. As the ship’s bridge was unattended throughout the port stay, none of its crew saw indicators of the approaching storm until just before the breakaway.
The ship’s crew responded swiftly. The bridge was manned and machinery was operational by the time the ship had turned 90 degrees to the wharf. The ship’s movement was then controlled using its thrusters and main propulsion until, with tug assistance, it was returned to the wharf.
What's been done as a result
The ship’s managers, TT-Line Company, advised the ATSB that it has implemented immediate changes to shipboard weather monitoring and notification arrangements along with changes to heavy weather and mooring procedures. These changes include: weather triggers for increased shipboard readiness; immediate notification of weather warnings; access to the Bureau of Meteorology (BoM) website from the bridge; changes to the wind speed alarm settings and; requiring all mooring lines to be held on the winch brakes.
TT-Line also engaged external marine consultants to complete extensive investigations and analyses into the mooring requirements and design for Station Pier. The consultants have completed mathematical modelling and incident replication simulations. Subsequent analyses will be used to identify and define operational parameters and recommend any alterations to berthing arrangements and infrastructure. The ATSB has issued one recommendation to TT-Line to complete safety action to adequately address the safety issue with respect to moorings.
The Victorian Ports Corporation (Melbourne) advised the ATSB that Melbourne vessel traffic service will broadcast BoM weather warnings on VHF channel 12. All masters of ships in port waters, including at berth or anchorage, are to ensure a listening watch is maintained at all times.
The BoM advised the ATSB that in addition to verifying the subscription service with the Victorian Ports Corporation (Melbourne) it continues to upgrade its marine weather services. This includes a one-stop webpage on its website for improved education, information and accessibility to marine and ocean services.
Safety message
All ships, especially those with high windage, are prone to breaking away from moorings during short-term events such as thunderstorms and squalls. The risks this presents to ships with large numbers of people on board mean that weather monitoring, mooring systems and procedures need to be regularly checked and verified for changing weather conditions.
The occurrence
At 0600 on 13 January 2016, the roll-on/roll-off passenger ship Spirit of Tasmania II (Figure 1) berthed at Station Pier in the Port of Melbourne. The ship had just completed its usual, scheduled Bass Strait transit from Devonport, Tasmania to Melbourne, Victoria.
Figure 1: Spirit of Tasmania II’s upper and lower bow ramps deployed at Station Pier
Source: Australian Pictorials
A ‘strong wind warning’[1] for Port Phillip was in effect and had been re-issued at 0527 and 1041 that morning. The warnings noted the risk of afternoon and evening thunderstorms with squalls to 45 knots.[2] The master was aware of these warnings, and also knew a ‘cold front’[3] was expected to pass across the port at about 2100 that evening. Departure was scheduled for 1930 and, as a precaution, the master had instructed that two of the ship’s four main engines be ready for manoeuvring at immediate notice from 1600.
After the ship had berthed, unloading of cargo, vehicles and passengers started and continued into the afternoon. As usual, Spirit of Tasmania II was to load cargo and vehicles via its bow, using a lower and an upper vehicle ramp (Figure 1). At about 1600, unloading was completed and the loading of cargo and vehicles started.
The master slept from 1330 onwards to rest for Spirit of Tasmania II’s evening sailing. At 1630, he awoke and attended to some administrative tasks. He also checked the Bureau of Meteorology (BoM) website and noted little change to the weather forecast. At about 1700, when he went to the navigation bridge (bridge), the weather situation appeared normal and as expected. At the time, the wind was about 20 knots from the west and it was cloudy and hot (44°C).
At about 1730, passengers began boarding – those with cars drove on board via the lower vehicle ramp while others boarded via the aft passenger access. The ship’s managers and master had agreed to board the waiting passengers about 30 minutes earlier than usual because it had been a hot day.
It was not until shortly after 1745 that the master returned to his cabin and again checked the BoM website (no internet access was available on the bridge). He saw that a ‘severe thunderstorm warning’[4] had been issued at 1737. The website indicated a series of thunderstorms approaching Station Pier from about 13 miles[5] to the west (Figure 2).
Figure 2: Map issued with the severe thunderstorm warning at 1737 on 13 January
Source: Bureau of Meteorology, annotations by ATSB
At about the same time, the second mate (the duty officer) was on the main vehicle deck attending to loading of cars into the main and lower decks. Vehicle traffic had been stopped on the landside of the ship’s bow ramp as traffic had banked up in the vehicle deck.
The second mate noticed a breeze coming in through the open pilot access door in the ship’s side. He called the chief mate via radio and informed him that the wind had picked up, suggesting that the change in weather may have arrived. He closed the passenger door/ramp and pilot access door.
Shortly after, as the master left his cabin to go to the bridge, he saw the wind speed indicator outside his cabin indicated gusts of up to 59 knots. He hurried to the bridge. At 1753, the wind peaked at 60.5 knots (112 km/h) from the southwest (Figure 3). By then, about 120 passengers had boarded and the ship’s stern had started to move off the wharf.
Figure 3: Wind speed and direction, and the ship's rate of turn
Source: Spirit of Tasmania II’s voyage data recorder
With the strong winds pushing the stern away from the wharf, the ship’s aft mooring lines failed in rapid succession. The ship pivoted about its bow and swung away towards the shore (Figure 4).
Figure 4: Section of navigational chart Aus 154 showing the ship’s breakaway
Source: Australian Hydrographic Service with annotations by ATSB (Inset: Google Earth)
On the bridge, Spirit of Tasmania II’s master saw the stern move off the wharf. He telephoned the engine control room but received no response. He then broadcast a message over the ship’s public address system asking an engineer to call the bridge and all crewmembers to go to their mooring stations. The senior engineer in the engine room heard the broadcast, returned to the engine control room, and called the bridge. The master asked for the bow and stern thrusters, followed by two main engines, to be started as quickly as possible.
By 1755, the ship was at an angle of approximately 45 degrees to the wharf and turning away from it (negative rate of turn in Figure 3) at 27°/min.
At 1756, an additional diesel generator was running and the thrusters were ready for use. The engineers then started necessary pumps and by 1758½, two main engines were also ready.
As the ship’s stern came off the wharf, all the aft mooring lines parted. At the bow, the breast line came under tension, but, along with one head line, did not part. The ship pivoted about the bow and its lower bow ramp slid across and dropped off the wharf leaving it hanging over the bulbous bow. As the ship continued to turn, the end of the lower bow ramp swung under the wharf, damaging two wooden piles below the wharf apron.
The upper vehicle ramp detached from the ship’s bow mount and jammed between the bow bulwarks.[6] As the ship continued to turn, the ramp twisted away from the bow and hung down from its elevated roadway base.
Spirit of Tasmania II’s purser and hotel staff were assisting passengers on board when the ship broke away. The purser became aware of the situation when she heard, and then saw, the aft mooring lines parting and the stern moving off the wharf. She followed the developments and response by listening to the ship’s internal radio traffic. She and her staff assisted passengers, distributed food, drinks and provided reassurance and information as it became available.
By 1759, Spirit of Tasmania II was lying at right angles to Station Pier, parallel to the nearby Port Melbourne Beach. By then, the stern thruster was being used to control the movement of the ship’s stern.
The Port of Melbourne’s vessel traffic service (VTS) had been aware of the strong wind warning and had two tugs standing by throughout the day. One tug was at Webb Dock (about 1 mile to the west of Station Pier) as a precaution to assist a high-sided ship berthed there. The duty VTS officer directed the tug’s master to proceed to the nearby Station Pier to assist Spirit of Tasmania II.
By 1800, the ship’s stern was being held off the beach using the main propulsion (twin screw, variable pitch) and the thrusters (bow and stern).
At 1809, the tug arrived off Spirit of Tasmania II and made fast to the ship’s stern. At 1826, the second standby tug also arrived and began assisting.
By 1900, the wind speed had dropped to about 10 knots and was blowing from the south. At 1905, the ship was again port side alongside Station Pier. It was held alongside using repaired mooring lines and spares, with the tugs assisting. The ship’s engines and thrusters were kept ready for use as required.
Later in the evening, passenger access was restored, and many were accommodated on board for the night. Access to unload vehicles and freight became available the following day.
When Spirit of Tasmania II had turned off the wharf, its starboard bow door had come to rest against the wharf structure. Weight had then come onto it and pushed it against the lower bow ramp, resulting in damage to a number of its door operating components. The hydraulic system used to operate the doors was not damaged. As the lower bow ramp moved off and under the wharf, its end fingers were damaged as they made contact with the wharf piles. The upper vehicle ramp suffered extensive damage (Figure 5) requiring it to be removed from the wharf and taken away for repairs.
Inspections and repairs as required by the Australian Maritime Safety Authority, the ship’s flag State authority, and its classification society were then undertaken. On 17 January, 4 days after the incident, Spirit of Tasmania II returned to service.
At the time of the incident, Spirit of Tasmania II was registered in Australia, and owned and managed by TT-Line Company, Australia (TT-Line). The ship was classed with the American Bureau of Shipping (ABS).
Spirit of Tasmania II’s standard mooring line pattern[7] had been developed for the Station Pier berth (Figure 6). This consisted of three head lines on bights[8] (numbered 7, 8 and 9 in Figure 6), a spring line (10) and a breast line (11) forward, and three stern lines (1, 2 and 3), two breast lines (4 and 5) and a spring (6) line aft. All mooring lines were 64 mm diameter, mixed synthetic fibre rope,[9] each with a breaking load of 82 t.
Figure 6: Spirit of Tasmania II’s standard mooring line pattern for Station Pier
Source: TT-Line Company with annotations by ATSB
In Melbourne, the ship loads cargo and vehicles via its bow, using a lower and an upper vehicle ramp. The lower ramp is part of the ship’s equipment and, when deployed, extends out from the main vehicle deck, through the bow doors and onto the wharf. The upper ramp, also known as the skybridge, extends from a raised roadway above the wharf onto the ship’s bow. The skybridge attaches to the ship’s bow via a ball hitch arrangement that accommodates movement of the ship. This provides vehicular access into the upper vehicle decks through the foredeck cargo door.
The ship had a crew of 70 Australians. The master held an Australian master’s certificate of competency and a pilotage exemption for the Port of Melbourne (which includes Port Phillip). He first went to sea as a cadet in 1981, and obtained his master’s certificate of competency in 1994. He had sailed on board Spirit of Tasmania II and its sister ship, Spirit of Tasmania I, since 2002 and first sailed as master of Spirit of Tasmania II in 2004. He had joined the ship about 3 weeks before the incident.
Moorings
A ship’s moorings are intended to keep it safely alongside the berth in the prevailing and expected weather conditions, up to certain limits. Mooring equipment is not designed to be used in extreme conditions. A number of industry organisations publish a range of useful guidance material to assist with effective mooring.
Equipment design
A ship’s mooring equipment, including related machinery and fittings, and its positioning, is determined at the ship design stage. The size, strength and other specifications of various equipment (including mooring ropes, brakes and winches) are based on the ‘equipment number’ or EN. The EN is based on a ship’s size (displacement, beam and height above the waterline)[10] so that a new ship is outfitted with the appropriate equipment. If circumstances during the ship’s service life change, for example its trading pattern, the mooring equipment, machinery and fittings and how these are arranged on deck may need to be modified.
Other than the EN, there are few other design criteria, rules, regulations or class requirements for mooring equipment. Mooring equipment arrangements are not part of class requirements, other than the sizing of the equipment and the strength of the supporting structure. Class does require a minimum number and size of mooring lines, based upon the ship’s EN, to be used. Mooring equipment must be properly maintained to preserve its design specifications.
Influencing factors
Many of the dynamic factors influencing a ship’s mooring arrangement are weather-related. These include wind, sea and swell along with local factors such as currents, tides and tidal streams. Port and harbour specific factors include berths, channels, other infrastructure and traffic. Many factors interact with and/or influence other factors, for example strong winds can affect currents or streams.
Wind
Wind is one of the most significant factors influencing a berthed ship. The wind force acting on the ship depends on the wind speed and the surface area of its hull exposed to wind (windage). When the wind blows across the berth and towards the water, it will tend to push the ship away from the berth and its moorings will need to withstand the forces imposed.
A simple formula[11] that gives an approximation of wind loading is:
F = 1.5 x 10-5 * A * V2
where
F is the wind force in tonnes A is the exposed area in m² V is wind speed in knots.
Since the wind force varies as the square of the wind speed, small increases in speed translate to large increases in force. In strong winds, gusting amplifies these forces significantly. Strong winds frequently contribute to breakaways.
Sea, swell, currents and tides
The sea and swell acting on a ship’s underwater hull, when sufficiently strong, can cause it to move and impose forces on its moorings. Currents or tidal streams also act on the underwater hull and create lateral forces. The ship’s moorings need to withstand all of these forces.
The height of tide is another important factor as the ship’s hull rises or falls relative to the berth and its mooring lines tighten or slacken. The lines need to be appropriately tended to keep the ship safely alongside.
Port, berth and local factors
The design and location of a port is intended to provide a sheltered area for ships. Some ports have a naturally sheltered harbour, whereas others might need a man-made breakwater at their entrance. Similarly, wharves and jetties are designed to provide appropriate berths for ships. The number and position of bollards and hooks at a berth is an important factor.
Factors such as the proximity of a berth to a shipping channel are important considerations. Passing ships in the channel can interact with a berthed ship and cause it to surge and part its moorings. Ship speeds in channels are limited to eliminate such interaction.
Industry guidance
The Oil Companies International Marine Forum (OCIMF) provides comprehensive industry guidance on mooring in publications such as its Mooring Equipment Guidelines and Effective Mooring. The Nautical Institute provides guidance in two volumes of its Mooring and Anchoring of Ships publication. Together, these publications are the most significant body of research and information on the subject. They are regularly referred to within and outside the oil sector and are accepted as best practice for effective mooring.
The following points summarise some key considerations for effective mooring line patterns described in the publications referred to above:
optimum load sharing between mooring lines is achieved when lines are loaded to the same percentage of each line’s breaking strength (rather than each line carrying an equal share of the load)
lines should be of the same material and diameter (lines of differing materials and diameters have different elasticities and will not share additional load evenly)
lines should be of similar lengths because, as the ship moves off the wharf, the lines will be stretched the same amount which will result in uneven loading in lines of different lengths
the steeper the angles a line forms with the pier surface and with the parallel side of the ship, the less effective it is in holding the ship against the pier
mooring winches should be left on the brake once the ship is secured alongside because the brake holding capacity is significantly more than that available in self-tensioning mode[12]
if self-tensioning mode is used, it is best used for lines at 90 degrees to the ship’s axis (breast lines)
sharp bends in a line under load decrease its strength and may cause premature damage or failure
while running a line as a bight increases its effective holding capacity, if the line parts then two components of the mooring line pattern are lost in a single event
regular inspection and tending of the mooring lines is required to maintain the mooring effectiveness and counter changes in forces including those due to environmental conditions, tide and draught changes.
The mooring line pattern of a ship will vary significantly with its size and type, port and berth infrastructure, weather and dynamic factors, and effective mooring depends on a range of complementary measures. The pattern should allow for the maximum static load on the moorings with an adequate safety factor to account for expected dynamic factors.
However, certain dynamic situations cannot be easily accounted for by design features or by precautionary measures. For example, when the forces acting on the moorings change suddenly and last longer than it takes the mooring line pattern to respond, a failure becomes more likely. A typical situation is the parting of a mooring line due to a sudden gust in a period of strong winds or surging due to interaction with a large ship passing close at high speed.
The mooring equipment arrangements and mooring line pattern are just two components of effective mooring. Other important considerations include regular tending of mooring lines and effective monitoring of the weather to ensure necessary precautionary measures can be taken in a timely manner. The management of both the ship and the port have a role in effective mooring.
TT-Line Company
The TT-Line Company (TT-Line) has been operating passenger and vehicle ferry services across Bass Strait since 1985. The sister ships Spirit of Tasmania I and Spirit of Tasmania II began service in 2002. Both ships maintain year round, daily, one-way sailings with daily return sailings during the peak summer period. Each ship can carry 1,400 passengers and 500 vehicles. In the 2014-15 financial year, the company’s ships carried 384,501 passengers.
Port of Melbourne
The Port of Melbourne is Australia’s largest container and automotive port. Victorian Ports Corporation (Melbourne) manages the port,[13] which is located at the northern end of Port Phillip (Figure 7). The port has 36 commercial berths for all types of cargo. In the 2014-15 financial year, the port recorded 3,023 ship visits and handled 87 million tonnes including over 2.58 million containers and 350,000 new motor vehicles.
The Port of Melbourne vessel traffic service (VTS) operates 24-hours a day, 7-days a week, to coordinate the port’s maritime operations.
Figure 7: Melbourne and Port Phillip
Source: Port Phillip Sea Pilots, annotations by ATSB
Station Pier is a historic pier in Port Phillip. Opened in 1854, the pier is Melbourne's primary passenger terminal, servicing interstate ferries and cruise ships. To the east of the pier is Port Melbourne Beach, a popular public beach for residents and visitors.
Weather
The Bureau of Meteorology (BoM) provides weather forecasts, warnings and observations for marine users in coastal and local water areas and high seas off Australia (see Appendix A). Forecasts for wind speed and direction, and sea and swell height are routinely issued every 12 hours. When a wind warning is in force, these forecasts are updated every 6 hours. Wind warnings are issued whenever averaged 10-minute wind speeds are expected to exceed the thresholds for the following categories: strong wind (26 to 33 knots), gale (34 to 47 knots), storm force (48 to 63 knots) and hurricane force (more than 64 knots). Initial warnings aim to provide 24‑hour lead times and are updated every 6 hours.
Weather warnings are broadcast on high frequency radio each hour. The warnings are also broadcast via the maritime SafetyNet satellite communication network when issued and updated at 0400 and 1600. These messages are automatically received and printed on board ships to which SOLAS[14] applies. All current warnings are posted on the BoM website.
In the absence of severe weather, such as thunderstorms or squalls,[15] it is normal for wind speed to vary by up to 40 per cent over a 10 minute period. The preamble in all marine forecasts cautions that wind gusts can be 40 per cent higher than the mean wind speed predicted. Stronger gusts are also likely near showers and frontal systems. During thunderstorms or squalls, significantly stronger gusts (higher than the 40 per cent caution) can occur.
In addition to official warnings and marine forecasts, BoM provides a range of weather monitoring tools on its website (see Appendix B). These include tools for accessing weather forecasts as well as weather observations (current and past weather), such as MetEye, weather watch radar and wind forecasting tools. The information displayed by all the tools is to some extent historical as it is based on average data recorded over a period of time and not instantaneous readings. The BoM cautions against using the tools for real-time monitoring and decision-making. These tools cannot be used to accurately monitor short-term weather events such as thunderstorms, squalls and gusts in real time. The intensity of these short duration weather events is also difficult to display historically.
Port Phillip
The weather in Port Phillip is regularly affected by fronts associated with east moving depressions (low pressure weather systems) often centred well south of the coast. On 35 to 40 occasions each year, these depressions create winds of gale force or greater strength offshore. Such weather is more than twice as likely during winter as in summer. While the wind in Port Phillip is less intense than offshore, it can still be significant. Thunderstorms and squalls can also produce strong local winds in Port Phillip and are more frequent in summer (December to February).
In addition to local waters forecasts and wind warnings for Port Phillip, severe thunderstorm warnings for the Melbourne region are particularly relevant to users and the Port of Melbourne. These warnings provide short-term advance warning of severe thunderstorms and are updated every 3 hours. Typically, such warnings are issued on 10 days per year. Squall warnings not associated with thunderstorms are issued on a further 5 or 6 days a year. Sudden wind gusts over 48 knots are experienced often in Port Phillip and a number of weather related incidents have occurred over the years.
Previous incidents
Amongst previous weather related incidents, the two breakaways described below have some similarities to Spirit of Tasmania II’s 2016 breakaway.
2002 - Spirit of Tasmania II
On 13 October 2002, in westerly winds of about 53 knots (98 km/h), Spirit of Tasmania II broke away from Station Pier and turned parallel to the nearby Port Melbourne Beach. At the time, the ship’s two stern lines were made fast to a single wharf bollard. This bollard shifted under the strain causing uneven loading of the mooring lines, which then parted in rapid succession.
Safety action taken by TT-Line after this incident included:
fitting a wind alarm in the alleyway adjacent to the master’s and deck officers’ cabin
the ship’s starboard stern fairlead array was altered to allow extra mooring lines to be run
a new heavy bollard was fitted on the wharf for running of ships’ stern lines.
The mooring line pattern in use on 13 January 2016 was the same pattern that was implemented after the 2002 incident.
Further, this 2002 incident led to a number of procedural changes related to precautions to consider if adverse weather was experienced while berthed. These include:
monitoring the wind indicator and alarm outside the master’s cabin
manning the bridge and monitoring weather reports and warnings
placing the main engines and thrusters on short notice
closely monitoring mooring lines and equipment
stopping cargo and passenger operations
retracting ramps and closing hull doors
preparing for emergency letting go
engaging tugs.
2009 - Leyte Spirit
On 21 August 2009, a severe squall with winds of up to 68 knots (126 km/h) passed over the Port of Melbourne causing the crude oil tanker Leyte Spirit to breakaway from its berth.
A safety investigation into the incident[16] found that the squall line developed very rapidly and, while it was detected by BoM and a warning was issued, neither the terminal nor the tanker received the warning. No appropriate systems were in place to obtain and share local weather information and the investigation recommended that port users develop procedures for obtaining local weather information.
Subsequently, the terminal updated its procedures to ensure that vessels are aware of local weather warnings. The Port of Melbourne Corporation Harbour Master’s Directions[17] were updated to assist ship masters with obtaining local weather information.
Between 1751 and 1753 on 13 January 2016, the south-westerly wind at Station Pier, Melbourne, increased from about 25 knots to 60 knots. As the wind speed was peaking, Spirit of Tasmania II’s aft mooring lines parted in quick succession and its stern broke away from the Station Pier wharf. As the stern swung away from the wharf, the originally slack forward breast line (11) came under tension and, along with one head line on the bight (9), did not part (Figure 6).
By 1759, the ship lay at right angles to the wharf and parallel to Port Melbourne Beach (Figure 8). While waiting for tugs to assist, the ship’s propulsion was used to maintain its position and prevent grounding. By 1905, the ship was back alongside Station Pier with the two tugs assisting. No one was injured and damage was limited to the ship’s bow ramp and doors.
The sudden and significant increase in wind speed contributed directly to Spirit of Tasmania II’s breakaway. The more than doubling of wind speed increased the wind loading on the ship’s side exponentially (nearly six times). As the buildings on the pier provided least protection from the wind to the shipside area aft, the aft mooring lines took the load first.
As the wind speed was peaking, the ship’s stern moved more than 4 m off the wharf stretching the aft moorings lines. Soon after 1753, the shortest of the stern lines parted. As the ship moved further off the wharf, the load increased on the other stern lines and they quickly parted. The two breast lines were unable to pay out quickly enough and parted. As the ship pivoted about its bow, the forward mooring lines overloaded and parted leaving just two lines intact (Table 1).
Figure 8: Spirit of Tasmania II adjacent to Port Melbourne Beach
Source: George Donikian
Table 1: Sequence and times of lines parting
Weather on 13 January
The weather conditions experienced in Port Phillip on 13 January were consistent with the typical Australian summer weather patterns. The weather systems, including those associated with short-term events such as thunderstorms, regularly develop in the area and were not unusual or extraordinary.
Forecasts
The Port Phillip local waters forecast issued at 1641 on 12 January indicated the possibility of afternoon thunderstorms with squalls up to 45 knots on the following day. The local waters forecast issued at 0527 on 13 January reiterated that earlier forecast, and stated:
Weather Situation
A high pressure centre lies over the Tasman Sea while a strong cold front approaches from the west during Wednesday. The cold front is expected at Port Phillip and Western Port around 2100hrs.
Forecast for Wednesday 13 January until midnight
Strong Wind Warning for Wednesday for Port Phillip
Winds: Variable around 10 knots tending northerly 15 to 20 knots during the morning. Winds shifting west to northwesterly 15 to 25 knots during the afternoon ahead of a southwesterly change of 20 to 30 knots during the evening. Squalls to 45 knots possible near the change and with thunderstorms. Seas: Around 1 metre. Weather: Sunny morning. 40% chance of showers during this afternoon and evening. The chance of a thunderstorm during this afternoon and evening.
The next local waters forecast at 1041 confirmed the ‘strong cold front’ was expected at about 2100 that evening and repeated the strong wind (26 to 33 knots) warning. Significantly, both the 0527 and 1041 local waters forecasts included the possibility of squalls up to 45 knots associated with afternoon and evening thunderstorms.
Severe thunderstorms were detected in western Victoria that afternoon. As the storms moved east, severe thunderstorm warnings for damaging winds for the state were issued from 1436 onwards. Each warning included a map highlighting the warning area. The 1508 warning stated that a detailed severe thunderstorm warning for the Melbourne area would be issued should thunderstorms develop there. At 1603, a warning indicated that Melbourne may be affected, including the possibility of damaging winds (defined by BoM as exceeding 49 knots or 90 km/h).
The local waters forecast at 1640 retained information about the cold front and squalls associated with thunderstorms. At 1714, a severe thunderstorm warning was issued for the Melbourne area, which indicated that two thunderstorm cells with damaging winds located southwest of Melbourne were moving southeast (Figure 9). The next warning was due to be issued an hour later, at 1815.
Figure 9: Maps issued with the 1714 and 1737 severe thunderstorm warnings
Source: Bureau of Meteorology, with annotations by ATSB
However, at 1737, a further severe thunderstorm warning was issued after thunderstorms were detected west of Melbourne. This band of thunderstorms with winds up to 54 knots (100 km/h) was moving east toward Melbourne and forecast to affect the city by 1805. The thunderstorm passed across Station Pier at 1753, bringing with it a sudden increase in wind speed. By 1830, the storm had passed and, a couple of hours later, the thunderstorm warnings were cancelled.
Therefore, the weather experienced at Station Pier when Spirit of Tasmania II broke away on 13 January had been generally forecast, well in advance, in local waters forecasts. However, the thunderstorm that actually passed over Station Pier at 1753 was not specifically predicted until 1737 (that is, 15 minutes before the breakaway) shortly after the band of storms was detected.[18]
Ship’s preparedness
Spirit of Tasmania II’s master had monitored successive Port Phillip local weather forecasts for the period covering the ship’s stay at Station Pier. He was aware of the strong wind warning and the ‘strong cold front’ expected in Port Phillip at 2100 on 13 January. As a precaution, he required the engine room and machinery to be at ‘short notice’[19] from 1600, well before the cold front’s arrival.
Once the ship berthed, its bridge was not manned and weather forecasts and warnings available via VHF radio and other equipment were not continuously monitored. The master’s cabin had internet access from where he could access the BoM website. However, he was sleeping between 1330 and 1630 and remained unaware of the thunderstorm warnings issued since 1436. After he awoke, he checked the local waters forecast and found it unchanged.
At about 1700, while on the bridge, he saw that the wind and weather were also as he expected. There was no internet access on the bridge so he could not check the BoM website and remained unaware of the severe thunderstorm warning issued at 1714. After returning to his cabin at 1745, he checked the BoM website at about 1750 and saw the severe thunderstorm warning issued at 1737. By then, it was too late for the propulsion to be readied as the thunderstorm struck a couple of minutes later at 1753.
The master had use of the ship’s thrusters and main engines about 5 minutes after it broke away. He was able to use the ship’s propulsion to partially recover from the situation. Had he been able to use its propulsion earlier when the thunderstorm struck, it is likely that he would have used it to try to prevent the breakaway.[20]
While the local waters forecasts predicted thunderstorms in the afternoon and evening, the evidence suggests that the master expected the weather to deteriorate when the cold front arrived in the evening. He had no means to automatically receive thunderstorm or squall warnings other than to continually check BoM’s website.
TT-Line expected masters to take appropriate adverse weather precautions based on their significant experience on board company ships regularly calling at the same ports. However, there were no specific trigger points for taking any particular precaution and reliance was on an individual master’s discretion and decision-making. On this occasion, the master did not become aware of the thunderstorms which developed that afternoon and precautions such as manning the bridge, monitoring the moorings or engaging tugs were therefore not taken.
The wind monitor and alarm outside the master’s cabin was of little use in providing advance warning for the thunderstorm. The alarm was set to activate (audible alarm) when a wind speed of 18 m/s (33 knots) or more persisted for at least 10 minutes. In other words, it was set to provide a warning of a steady gale force wind. The winds experienced in the thunderstorm did not activate the alarm. When the master was going to the bridge, he saw that the wind speed indicator had registered 59 knots.
The precautions taken on board Spirit of Tasmania II on 13 January provided little defence against the sudden, storm force winds experienced during the thunderstorm. Shipboard procedures for adverse weather when alongside did not take into account all the necessary factors required to provide effective defences against significant, short-term weather events such as thunderstorms.
Port’s response
The Port of Melbourne’s vessel traffic service (VTS) provides essential and timely information to masters to assist in safe and efficient traffic movement in the port. The weather is an important part of such information and VTS provides current and forecast weather reports on request via VHF channel 12. The port also requires all ships to monitor weather conditions and obtain BoM forecasts, including weather reports issued by Coast Radio Melbourne on VHF channels 16/67.
The masters of berthed ships are also required to attend to moorings to ensure their ship is appropriately and effectively secured and notify VTS if mooring lines part. High-sided ships such as car carriers are required to have a tug standing by in certain conditions depending on the berth (for example, a car carrier at Webb Dock must have a tug standing by when there is a steady wind of 30 knots or more from the south-west to west). However, there were no other requirements for berthed ships when strong winds were expected.
At the time of the incident, VTS received BoM marine weather forecasts and warnings, and severe thunderstorm warnings via email subscription. However, on 13 January, the severe thunderstorm warnings were not received due to an incorrect BoM system configuration. The harbour master was trialling a third party weather monitoring and alert system and received the warnings. He telephoned VTS at 1622 and found it had not received the state-wide severe thunderstorm warning issued by BoM at 1603. He read out the warning and then forwarded a copy to VTS.
At 1721, a Port Phillip pilot advised VTS that there were winds of about 70 knots in Geelong.[21] At 1726, VTS contacted the harbour tug company and asked for tugs to be readied, one of which was to attend a car carrier berthed at Webb Dock. During the following 30 minutes, VTS called a number of port users (terminals, ship agents and some ships) and warned that winds over 50 knots were approaching from the west. However, neither TT-Line nor Spirit of Tasmania II was amongst the parties that VTS notified.
The VTS procedures provided for limited assistance to be given to the masters of berthed ships in the event of a weather event. The responsibility remained with the ships’ masters to monitor and respond to the weather. As a consequence, on 13 January, VTS’s response to the sudden weather event was only partially effective in dealing with the situation in Port Phillip.
Spirit of Tasmania II’s moorings
Spirit of Tasmania II’s mooring line pattern on 13 January was effectively the same as the standard pattern for that berth (Figure 6). The only difference was that a forward headspring[22] had not been run that day, and the forward breast line was slack at the time of the breakaway. Therefore, the ship was effectively being held alongside by three mooring lines on the bight forward and six lines aft (Figure 10).
The ship’s planned maintenance system required mooring lines to be inspected every month. The maintenance records indicated that all aft lines were in good or new condition when inspected on 2 January, a few days before the breakaway. The records indicated that regular maintenance had been carried out on the mooring winches. The design winch brake holding capacity was 69 t. No defects were recorded at the previous 3 monthly mooring equipment inspection, completed in October 2015 and, while the brakes had never been render (slip and pay out) tested, there was no requirement for such testing.
Figure 10: Spirit of Tasmania II's mooring lines as run on 13 January (lines highlighted)
Source: Google Earth, with annotations by ATSB
At the time of the breakaway, all stern lines were held on the winch brake. The winches holding both aft breast lines were set to self-tension. The self-tension setting was reported to have been between 80 and 100 per cent of the winches’ rated pull[23] of 15 t.
The CCTV footage from 1752 shows Spirit of Tasmania II’s stern moving off the wharf and its aft mooring lines stretching due to the wind loading (Figure 11). By 1753, as the wind was peaking at about 60 knots, the ship’s stern was over 4 m off the wharf with the lines considerably stretched as the winch brakes held. Moments later, the shortest stern line parted when it could not stretch any further. As the load transferred to other stern lines, they parted in quick succession.
Figure 11: CCTV images showing the ship moving off the wharf and the first line parting
Source: TT-Line Company, with annotations by ATSB
As the ship’s stern was breaking away, the winches holding the breast lines payed out. However, as the ship’s movement gathered speed, those lines parted. Two of the three forward head lines also parted during the breakaway.
Based on the pattern of the aft mooring lines (not including the spring) their combined static load capacity was approximately 200 t, as set-up.[24] The forces from the wind loading on Spirit of Tasmania II cannot be calculated exactly because of the buildings and structures above and below Station Pier. However, if these and other complexities are ignored, a force of approximately 190 t could have been acting on the ship when the wind speed peaked.
In theory, therefore, the aft moorings should have prevented the breakaway. The moorings did hold until the wind speed reached 60 knots. Possible reasons for the failure of the mooring lines include:
unequal tension in mooring lines (before the sudden increase in wind speed)
not all lines were held on winch brakes (some winches on self-tension)
unknown mooring line defects
sudden dynamic forces (six-fold increase in wind loading)
dynamic loading (ship’s momentum as it rapidly moved off the wharf).
It is possible that the breakaway might have been prevented had the precautions for adverse weather been more carefully considered. For example, had the breast lines been held on winch brakes, the increased static load capacity of the aft mooring line pattern may have prevented the first stern line parting. Similarly, had the ship’s propulsion been ready for immediate use, the stern thruster (maximum 15 t transverse thrust) could have been used to counteract some wind loading to prevent the first line parting.
While the standard mooring line pattern had been successfully used for many years, the breakaway indicated that the risk could have been further reduced to be better prepared for such unusual circumstances. The number, type and strength of mooring lines, winch brake testing, use of self-tension winches, bollard location and other items to improve leads for mooring lines and line tending practices are amongst the areas where enhancements may reduce risk.
From the evidence available, the following findings are made with respect to the breakaway of Spirit of Tasmania II that occurred at Station Pier, Melbourne on 13 January 2016. 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
At 1753 on 13 January, the wind speed suddenly increased from about 25 to 60 knots during a thunderstorm resulting in a six-fold increase in the wind loading force acting on Spirit of Tasmania II.
The exponential increase in the wind loading force pushed the ship’s stern off the wharf, stretching its aft mooring lines.
Soon after 1753, the shortest of the three stern lines (all held on winch brakes) stretched beyond its limit and parted. As the load transferred to the remaining stern lines, they parted in quick succession.
The two breast lines aft (held on self-tension winches) payed out before parting, allowing the stern to breakaway.
While the weather experienced, in general, was forecast well in advance, the thunderstorm was not specifically predicted until 15 minutes before the breakaway leaving little time for the ship’s crew to respond.
The adverse weather procedures for TT-LineCompany ships when alongside did not take into account all the necessary factors to provide effective defences against significant, short-term weather events such as thunderstorms and squalls. [Safety issue]
Other factors that increased risk
The Port of Melbourne vessel traffic service (VTS) procedures for adverse weather were not comprehensive and, hence, its response on 13 January was only partially effective. One important consequence was that VTS’s advance warning of storm force winds did not reach all relevant parties, including Spirit of Tasmania II’s master. [Safety issue]
While TT-Line Company’s standard mooring line patternfor ships at Station Pier had been successfully used for many years, the breakaway indicated the risk could have been further reduced to better prepare for such unusual circumstances. [Safety issue]
Safety issues and actions
The safety issues identified during this investigation are listed in the Findings and Safety issues and actions sections of this report. The Australian Transport Safety Bureau (ATSB) expects that all safety issues identified by the investigation should be addressed by the relevant organisation(s). In addressing those issues, the ATSB prefers to encourage relevant organisation(s) to proactively initiate safety action, rather than to issue formal safety recommendations or safety advisory notices.
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: The adverse weather procedures for TT-Line Company ships when alongside did not take into account all the necessary factors to provide effective defences against significant, short-term weather events such as thunderstorms and squalls.
Safety issue description: The Victorian Ports Corporation (Melbourne) advised the ATSB that it had issued Victorian Notices to Mariners number 032-2016 advising that Melbourne vessel traffic service will broadcast Bureau of Meteorology weather warnings on VHF channel 12. All masters of ships in port waters, including at berth or anchorage, are to ensure a listening watch is maintained at all times.
A Port Information Notice (PIN number 08/2016) was also issued for passenger ships at Station Pier to ensure adequate mooring lines are used throughout the ships’ stay. Masters are reminded that the berth is exposed to strong winds and inclement weather. They are to keep a listening watch on VHF channel 12.
Safety issue description: While TT-Line Company’s standard mooring line pattern for ships at Station Pier had been successfully used for many years, the breakaway indicated the risk could have been further reduced to better prepare for such unusual circumstances.
Appendices
Appendix A – Bureau of Meteorology weather service and terms
The Bureau of Meteorology (BoM) provides the maritime community with weather forecasts, warnings and observations for coastal waters areas around Australia.
Marine and local waters forecasts
Forecasts for wind speed and direction, and sea and swell heights are issued routinely every 12 hours. When a marine wind warning is in force, these forecasts are updated routinely every 6 hours. These forecasts provide information on mean wind, sea height, swell height and direction and weather conditions:
local waters - areas such as bays, harbours and inland waters on which frequent boating activity occurs
coastal waters – areas within 60 miles of the coast.
The following terminology is used to describe time periods in these forecasts:
Early in the morning - expected to occur before 0600
In the morning - expected to occur between 0700 and 1000
Middle of the day - expected to occur between 1100 and 1300
During early afternoon - expected to occur between 1400 and 1500
In the afternoon - expected to occur between 1600 and 1800
During the evening - expected to occur between 1900 and 2000
Later in the evening - expected to occur after 2100.
All BoM marine forecasts and warnings predict wind speed as the average (or mean) speed expected over any given 10 minute period at a height of 10 m above the ground. Hence, marine forecasts provide a summary of the average wind speed expected within the forecast area, not the maximum wind gust expected.
Wind warnings are only issued when mean winds (winds speed averaged over 10 minutes) are expected to exceed 25 knots for an extended period of time across more than 10 per cent of the marine area. A wind warning is not issued for individual wind gusts, which typically last seconds.
In the absence of severe weather, such as thunderstorms or squalls, it is normal for wind speed to vary by up to 40 per cent over a 10 minute period. Hence, a statement is added to BoM marine forecasts to remind users that wind gusts at any time can be 40 per cent higher than the mean wind predicted in BoM marine forecasts and warnings. Wind gusts can be much higher than 40 per cent during severe weather, such as thunderstorms or squalls.
Mariners should refer to the weather section of the coastal or local waters forecast to determine the likelihood of thunderstorms occurring. They should also be aware that there is a risk of damaging wind gusts occurring with any thunderstorm and should actively monitor weather conditions when thunderstorms are forecast.
For operations near the shore, BoM’s severe thunderstorm warning service should also be used. These warnings provide short-term advanced notice of thunderstorms that are likely to be accompanied with wind gusts in excess of 48 knots, as well as other hazardous conditions that can accompany thunderstorms.
Recipients are further cautioned that maximum wave heights may be up to twice the height of those forecast (average). More information is available from the BoM we
bsite.
Marine weather warnings
Marine weather warnings are issued whenever strong winds, gales, storm force or hurricane force winds are expected. The following warnings are provided:
coastal waters wind warnings
ocean wind warnings – issued to ships at sea whenever gale, storm or hurricane force winds are expected
severe weather warnings – provided for potentially hazardous or dangerous weather that is not directly related to severe thunderstorms, tropical cyclones or bushfires.
Gust speed can be 40 per cent greater than the predicted wind speed.
Marine wind warnings aim to provide a 24-hour lead-time and are normally reviewed every 6 hours and issued every 12 hours.
Wind speed warnings
These warnings are issued whenever strong winds, gale, storm or hurricane force winds are expected. They provide around 42 to 24 hours’ notice and are updated every 6 hours. The wind warning statement is included in all the coastal waters and local waters forecasts affected by the warning. A state-wide marine wind warning summary is also available and lists all the coastal waters and local waters affected by warnings for that day and the following day (listed on the BoM’s National warnings summary webpage).
These warnings are based on expected 10 minute mean winds (not maximum wind gusts) and are only issued if mean winds are expected to exceed thresholds for extended periods. The following 10 minute mean thresholds are used:
strong wind warning: 26 to 33 knots, force 6 to 7[25]
gale warning: 34 to 47 knots, force 8 to 9
storm force wind warning: 48 to 63 knots, force 10 to 11
hurricane force wind warning: 64 knots or more, force 12.
Wind direction is given using the eight compass points for forecasts and 16 points for observations and is the direction the wind is coming from.
Severe thunderstorm and severe weather warnings
Severe thunderstorm warnings provide short-term advance warning of the likelihood of severe thunderstorms impacting on the region and are updated every three hours. A severe thunderstorm can produce any of the following:
Severe weather warnings are issued for the following conditions:
sustained winds of gale force (34 knots or more) wind gusts of 48 knots or more
very heavy rain abnormally high tides
unusually large surf waves expected to cause dangerous conditions on the coast.
The BoM’s forecasters determine which warnings to issue to best inform the public about the hazards – for example, a severe weather warning may not be issued if a severe thunderstorm warning has already been issued.
While both the severe thunderstorm and severe weather warnings are land-based products, people undertaking marine operations near the shore should also consider these warnings.
Terms used
Wind speed is the average speed of the wind over a 10-minute period at a height of 10 m above the surface level.
Gusts are increases in wind speed lasting for just a few seconds. Wind gusts are typically 40 per cent stronger than the average wind speeds provided in marine forecasts. However, thunderstorms and squalls may produce much higher gusts.
A squall is an abrupt and large increase in wind speed that usually only lasts for minutes then diminishes rather suddenly. The gusts in a squall may exceed 40 or 50 knots.
More information is available from the BoM website.
Appendix B – Bureau of Meteorology weather monitoring tools
The Bureau of Meteorology (BoM) provides a great deal of information and weather monitoring tools to users via its website www.bom.gov.au. In addition to its official warnings and marine forecasts, the BoM provides the following weather monitoring and forecast tools.
Australian weather watch radar
The BoM provides a system of dedicated and part-time weather watch radars with images available online.[26] The images are updated about every 10 minutes for the dedicated system. The primary image display is of rainfall rates.
There are fifteen levels of rainfall intensity shown on the images - each level provides an approximate indication of the rainfall rate in millimetres per hour. Rain radar is not an indicator or predictor of the wind speed although there is some correlation between the severity or concentration of the displayed rain pattern and associated weather events (such as localised storms) and wind strength.
Thunderstorm precipitation cells can appear as isolated cells or in clusters or lines. Fast moving cells, rapidly growing cells, a bow in the direction of the movement of a line of cells and/or a long-lived cell moving in a markedly different direction to others may indicate the potential for severe weather (large hail, damaging winds and/or very heavy rain).
Where available, the individual weather radar is able to display Doppler wind data. Unlike the rain radar images, the Doppler wind data can be difficult to understand and interpret. Due to the complexities of interpretation, the BoM warn that care should be taken when using this data.
The information displayed on radar images is always historical being at least 10 minutes old and is based on average rather than instantaneous readings.
MetEye™
MetEye™ is an online[27] weather-mapping tool which displays forecasts, in three-hourly blocks, for weather features including wind speed and direction. MetEye™ allows animation of a range of official BoM forecast grids including rainfall, temperature and wind. These forecasts are updated twice daily. The system also allows information from other sources such as latest weather information including rain radar and wind speed and direction to be overlayed.
The wind forecast maps show the averaged 10 minute wind speed and direction for the time selected, based on a 10 m standard height. As wind speeds vary with gusts and lulls during a 10 minute period so the wind gusts experienced may be 40 percent stronger than the average speeds presented on the map.
MetEye™ is not a suitable real-time weather tracking or monitoring tool, especially for short-term weather events such as squalls and thunderstorms, as it is based upon forecasts from 10 minute average readings taken as much as 3 hours before the current time. Data refresh times vary and the most recent information is at least 6 minutes old. For this reason, care must be taken when using the tool for weather-based decisions.
Marine wind forecast
The BoM also provides computer generated marine wind forecast maps. The maps show average wind speeds and should be used in conjunction with official BoM marine weather forecasts as additional information and forecaster expertise are used to generate the official forecasts.
Accessing marine forecast and warning services
All BoM forecasts and warnings are available on the BoM website. In addition, there are a number of communication methods mariners can use to check BoM marine services, forecasts, warnings and observations:
Email subscription service – interested parties can subscribe to and receive a range of BoM services as they are released, including forecasts and warnings.
MarineLite (marine weather services (lite): includes high seas, coastal and local waters forecasts and warnings in text-only format, making them faster, more affordable and more accessible for slow data links.
VHF voice radio: inshore, broadcasts at scheduled times, provided by local marine organisations.
HF voice radio: off-shore, broadcast continuously 24/7, provided by BoM.
Inmarsat: global satellite communications for high seas warnings and forecasts, some coastal forecasts, provided by BoM for Australian high seas and coastal areas.
Telephone weather services: pre-recorded coastal and local waters forecasts and warnings for each State/Territory.
Marine weather knowledge centre – this web-based service provides a one-stop webpage for marine weather education information. The site includes a range of educational material to support the improved understanding of, and accessibility to, the existing information on BoM's Marine & Oceans webpage.
Mobile phone application, BoM Weather – users can view weather information (forecasts, warnings, rain radar, wind) for any location.
Note that severe thunderstorm warnings are not provided on HF radio or Inmarsat.
On 14 January 2016, investigators from the Australian Transport Safety Bureau (ATSB) attended Spirit of Tasmania II while the ship was in Melbourne, Victoria. The master and directly involved crewmembers were interviewed and each provided their account of the occurrence. Photographs of the ship and copies of available, relevant documents were obtained.
References
American Bureau of Shipping (ABS) 2015, Rules for Building and Classing Steel Vessels 2016, ABS, USA.
Australian Maritime Safety Authority (AMSA) 2016, Maritime Safety Awareness Bulletin, Issue 2 – Thinking – mooring safety, AMSA, Canberra. Available at www.amsa.gov.au
Babicz, J 2015, WartsilaEncyclopedia of Ship Technology, 2nd edn, Wartsila Corporation, Helsinki. Available at www.shippingencyclopedia.com
British Standards Institution (BSI) 2014, Maritime Works. Code of practice for design of fendering and mooring systems, BS 6349-4:2014, BSI, London.
Clark, IC 2009, Mooring and Anchoring Ships, Volume 1: Principles and Practice, The Nautical Institute, London.
Haddara, M & Soares, CG 1999, ‘Wind loads on marine structures’, Marine Structures, 12(3), pp.199-209.
International Association of Classification Societies (IACS) 2005, Recommendation No.10 Equipment, IACS. Available at www.iacs.org.uk
International Association of Classification Societies (IACS) 2014, Requirements concerning Mooring Anchoring and Towing, IACS. Available at www.iacs.org.uk
International Maritime Organisation (IMO), 1974, The International Convention for the Safety of Life at Sea, 1974, as amended (SOLAS), IMO, London.
International Maritime Organisation (IMO) 2005, MSC Circular 1175: Guidance on shipboard towing and mooring equipment, IMO, London.
Oil Companies International Marine Forum (OCIMF) 1989, EffectiveMooring, OCIMF, Bermuda.
Oil Companies International Marine Forum (OCIMF) 2008, Mooring Equipment Guidelines, 3rdedn, OCIMF, Bermuda.
Oil Companies International Marine Forum (OCIMF) 2010, Estimating the Environmental Loads on Anchoring Systems, OCIMF, Bermuda.
Port of Melbourne Corporation (PoMC) 2015, Port Information Guide, PoMC, Melbourne.
Port of Melbourne Corporation (PoMC) 2015, Harbour Master’s Directions, PoMC, Melbourne.
Thoresen, CA 2014, Port Designer’s Handbook, 3rdedn, Institute of Civil Engineers, London.
Turk, A & Prpic-Orsic, J 2009, ‘Estimation of Extreme Wind Loads on Marine Objects’, BrodoGradnja, 60(2), pp. 147-156.
UK P&I Club 2002, Technical bulletin Number 1, 2002 Mooring Practices, UK P&I Club, London.
Witherby Seamanship International 2007, Tanker Jetty Safety, Witherbys, London. Available through www.witherbyseamanship.com
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 TT-Line Company, the master and chief mate of Spirit of Tasmania II, the Australian Maritime Safety Authority, Victorian Ports Corporation (Melbourne), American Bureau of Shipping and the Bureau of Meteorology.
Submissions were received from TT-Line Company, the master and chief mate of Spirit of Tasmania II, the Australian Maritime Safety Authority, Victorian Ports Corporation (Melbourne), American Bureau of Shipping and the Bureau of Meteorology. 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: 18/04/2016
Safety summary
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 ongoing 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 report.
What happened
On the afternoon of 13 January 2016, the roll-on roll-off ship Spirit of Tasmania II was loading passengers and vehicles in Melbourne. At 1752, strong wind gusts blew the ship off the wharf and all but one of the ship’s mooring lines (on the bow) parted. As it broke away, the stern swung around until the ship was 90 degrees to the wharf and parallel to the nearby public beach, in danger of grounding. Action was taken to arrest the swing and the ship was returned to the wharf without touching bottom.
Wharf cargo and vehicle loading infrastructure was seriously damaged. The ship suffered minor bow damage. No one was injured.
What the ATSB has found so far
Based on the preliminary information that ATSB obtained, it was apparent that a band of severe thunderstorms passed across the area, including the location of Spirit of Tasmania II, with comparatively little notice. As the ship’s bridge was unattended throughout the port stay, none of the crew saw indicators of an approaching storm until immediately before the breakaway.
The ship’s crew responded swiftly. The bridge was manned and machinery was operational by the time the ship had turned 90 degrees to the wharf. The ship’s movement was then controlled using its thrusters and propulsion until it could be turned, with the assistance of a harbour tug, away from the beach and returned to the wharf.
Investigation direction
The investigation is ongoing and will focus on weather events in the Port of Melbourne area, and how the port and port users prepare for such events. This will include:
The ship’s managers’ (TT Line Company) preparations and procedures
Spirit of Tasmania II preparations and procedures for port stays and weather events, including mooring arrangements and equipment
Port of Melbourne procedures and actions
Bureau of Meteorology weather forecasting and warnings
distribution of weather information and warnings to and amongst port users.
The occurrence
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 ongoing 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 report.
At 0600[1] on 13 January 2016, the roll-on roll-off passenger ship Spirit of Tasmania II (Figure 1) berthed at Station Pier in the Port of Melbourne. The ship had just completed its usual, scheduled Bass Strait transit from Devonport, Tasmania to Melbourne, Victoria.
Figure 1: Spirit of Tasmania II alongside Station Pier after the incident
Source: ATSB
A ‘strong wind warning’[2] for Port Phillip was in effect and had been re-issued at 0527 that morning. The master was aware of this warning, and also knew a cold front was expected to pass across the port at about 2100 that evening. As a precaution, he had instructed that two of the ship’s four main engines be ready for manoeuvring at immediate notice from 1600. Departure was scheduled for 1930.
After the ship had berthed, unloading of cargo and passengers started and continued into the afternoon. As usual, Spirit of Tasmania II was to load freight and vehicles via its bow, using a lower and an upper vehicle ramp. At about 1600, unloading was completed and the loading of freight and vehicles started.
At 1630, the master, who had slept from 1330 onwards to rest for Spirit of Tasmania II’s evening sailing, awoke and attended to some administrative tasks. He also checked the Bureau of Meteorology (BoM) website and noted little change to the weather forecast. At about 1700, when he went to the navigation bridge, the situation appeared normal and as expected. At the time, the wind was 20 knots[3] from the west and the weather was clear.
At about 1730, passengers began boarding – those with cars driving on board via the lower vehicle ramp while others boarded via the aft passenger access. The ship’s managers and master had agreed to board the waiting passengers about 30 minutes earlier than usual because it had been a hot day (44°C).
At about 1750, the master returned to his cabin and again checked the BoM website. He saw that a ‘severe thunderstorm warning’[4] had been issued at 1737. The website indicated a series of thunderstorms approaching Station Pier from the southwest. The thunderstorms were about 13 miles off (Figure 2).
Figure 2: Thunderstorm map at 1737 on 13 January
Source: Bureau of Meteorology with annotations by ATSB
On the main vehicle deck, the duty officer, the second mate, was attending to loading of cars into the main and lower decks. Vehicle traffic had been halted on the landside of the ship’s bow ramp as it had banked up in the vehicle deck.
The second mate noticed a breeze coming in through the open ship’s side, pilot access door. He called the chief mate via radio and noted that the wind had picked up, suggesting that the change in weather may have arrived. He closed the passenger door/ramp and pilot access door.
Shortly after, as the master left his cabin to go to the bridge, he saw the wind speed indicator outside his cabin indicated gusts of up to 59 knots. He hurried to the bridge.
At 1753, the wind at Spirit of Tasmania II’s location was nearly 60 knots from the southwest (Figure 3). By then, about 120 passengers had boarded and the ship’s stern had moved off the wharf. With the violent storm force wind[5] pushing the stern off, the ship’s aft mooring lines failed in rapid succession. As the stern broke away from the wharf, the ship swung away and towards the shore (Figure 4).
Figure 3: Wind speed and direction, and the ship's rate of turn
Source: Spirit of Tasmania II voyage data recorder
On the bridge, Spirit of Tasmania II’s master saw its stern move off the wharf. He telephoned the engine control room but received no response. He then broadcast a message over the ship’s public address system asking an engineer to call the bridge and all crew members to go to their mooring stations. The senior engineer in the engine room heard the broadcast, returned to the engine control room, and called the bridge. The master asked for the bow and stern thrusters, followed by two main engines to be started as quickly as possible.
By 1755, the ship was at an angle of approximately 45 degrees to the wharf and turning away from it (negative rate of turn in Figure 3) at 27°/min.
At 1756, an additional diesel generator was running and the thrusters were available for use. The engineers then started necessary pumps and by 1758½, the two main engines were also ready.
Spirit of Tasmania II’s purser and hotel staff were assisting passengers on board when the ship broke away. The purser became aware of the situation when she heard, and then saw, the aft mooring lines parting and the stern moving off the wharf. She followed the developments and response by listening to the ship’s internal radio traffic. She and her staff assisted passengers, distributed food, drinks and provided reassurance and information as it became available.
Figure 4: Section of navigational chart AUS154 showing the ship’s breakaway
Source: Australian Hydrographic Service with annotations by ATSB (Inset: Google Earth)
As the ship’s stern came off the wharf, all the aft mooring lines and all except one of the forward lines parted. The ship pivoted about the bow, its bow ramp slid across and dropped off the wharf leaving it hanging over the bulbous bow. As the ship continued to turn, the end of the bow ramp swung under the wharf, damaging two wooden piles below the wharf apron.
The upper vehicle ramp detached from the ship’s bow and jammed between the bow bulwarks. As the ship continued to turn, this ramp twisted away from the bow and hung down from its elevated roadway base.
By 1759, Spirit of Tasmania II was lying at right angles to Station Pier, parallel to the nearby public beach. By then, the stern thruster was being used to control the movement of the ship’s stern.
The Port of Melbourne’s vessel traffic service (VTS) had been aware of the strong wind warning issued by BoM and had had two tugs on standby throughout the day. One tug was at nearby Webb Dock as a precaution to assist a car carrier there. The duty VTS officer directed the tug’s master to proceed to Station Pier to assist Spirit of Tasmania II.
By 1800, the ship’s stern was being held off the beach using two main engines and the bow and stern thrusters. At 1805, the other two main engines were also running and available for use.
At 1809, the tug arrived off Spirit of Tasmania II and was made fast to the ship’s stern. Another tug (the second one on standby) arrived at 1826 and began assisting. By 1900, the wind speed, which had peaked at 1752, was around 10 knots from the south.
By 1905, Spirit of Tasmania II had again been made fast, with its port side alongside Station Pier.
Later in the evening, passenger were allowed access onto and off the ship. Many passengers were accommodated on board for the night while damage assessments and investigations started. Access to unload vehicles and freight became available the following day.
When Spirit of Tasmania II had turned off the wharf, its starboard bow door had come to rest against the wharf structure. Weight had then came onto it and pushed it against the bow ramp, resulting in damage to a number of its door operating components. The hydraulic system used to operate the doors and ramp was not damaged. The upper vehicle ramp suffered extensive damage (Figure 5) requiring it to be removed from the wharf and taken away for repairs.
Figure 5: Photograph showing damaged upper vehicle ramp near Spirit of Tasmania II’s bow
Inspections and repairs were undertaken to the satisfaction of the ship’s flag State and classification society. On 17 January, 4 days after the incident, Spirit of Tasmania II returned to service.
At the time of the incident, Spirit of Tasmania II was registered in Australia, and owned and managed by TT Line Company, Australia (TT Line). The ship was classed with the American Bureau of Shipping (ABS).
Spirit of Tasmania II’s standard mooring arrangement had been developed for the Station Pier berth. This consisted of three head lines, a spring line and a breast line at the bow, and three stern lines, two spring lines and a breast line at the stern (Figure 6). All mooring ropes were 64 mm diameter KapaFlex 24 (24 strand, 40 per cent polyester, 60 per cent polypropylene/polyethylene) mixed rope, each with a breaking load of 82 t.
Figure 6: Spirit of Tasmania II’s mooring arrangement
Source: TT Line Company with annotations by ATSB
In Melbourne, the ship loads freight and vehicles via its bow, using a lower and an upper vehicle ramp. The lower ramp is part of the ship’s equipment and, when deployed, extends out from the main vehicle deck, through the bow doors and onto the wharf. The upper ramp, also known as the sky bridge, extends from a raised roadway above the wharf onto the ship’s bow. The sky bridge attaches to the ship’s bow via a ball hitch arrangement that accommodates movement of the ship. This provides vehicular access into the upper vehicle decks through the foredeck cargo door.
The ship had a crew of 70 Australians. The master held an Australian master class 1 certificate of competency and a pilotage exemption for the Port of Melbourne (which includes Port Phillip Bay). He first went to sea as a cadet in 1981, and had sailed as master since 1994. He had sailed on board Spirit of Tasmania II and its sister ship Spirit of Tasmania I since 2002. He first sailed as master of Spirit of Tasmania II in 2004. He joined the ship about 3 weeks before the incident.
TT Line Company
The TT line Company (TT Line) has been operating a passenger and vehicle ferry service across Bass Strait since 1985. The sister ships Spirit of Tasmania I and Spirit of Tasmania II began service in 2002. The ships maintain year round daily sailings with double sailings (two Bass Strait crossings by each ship, one in either direction, each day) during the peak summer period. Each ship can carry 1,400 passengers and 500 vehicles. In the 2014-15 financial year, the company’s ships carried 384,501 passengers and 102,309 containers.
Weather
The Australian Government’s Bureau of Meteorology (BoM) provides weather forecasts, warnings and observations for marine users in coastal and local water areas and high seas off Australia. Weather warnings are issued whenever strong winds, gales, storm-force or hurricane-force winds are expected. The initial warnings aim to provide 24-hour lead times, and are updated every 6 hours.
The warnings are broadcast via VHF radio channels 16 and 67 at 0718 and 1918 every day and are available on the BoM website. The warnings are also transmitted at 0400 and 1600 daily as enhanced group calling (EGC) messages via the maritime satellite communication network. Normally these messages are automatically received and printed on board ships.
On 13 January, BoM issued local waters forecasts for Port Phillip at 1041 and 1640. These forecasts included a strong wind warning for Port Phillip with squalls up to 45 knots possible.
At 1715, a severe thunderstorm warning was issued for the Melbourne area. This warning indicated that severe thunderstorms with damaging winds had been detected southwest of Melbourne and were heading to the southeast.
At 1737, the severe thunderstorm warning was re-issued for Melbourne with the storms now moving in an easterly direction. This warning indicated that a band of severe thunderstorms with winds up to 100 km/hr (about 51 knots) was moving toward Melbourne from the southwest. The storms were expected to affect Melbourne City by 1805. The warning included a map showing the band of storms and its expected movement (Figure 2).
The preamble in all BoM warnings cautions that wind gust speeds can be 40 per cent higher than the average wind speed, and stronger gusts are likely near showers, thunderstorms and frontal systems.
Port of Melbourne
The Port of Melbourne is Australia’s largest container and automotive port. The Port of Melbourne Corporation manages the port, which is located at the north end of Port Phillip Bay (Figure 7).
The port has 36 commercial berths for all types of cargo. In the 2014-15 financial year, the port recorded 3,023 ship visits and handled 87 million tonnes including over 2.58 million containers and 350,000 new motor vehicles.
Melbourne vessel traffic service (VTS) operates 24-hours a day, seven days a week, to coordinate the port’s maritime operations.
Figure 7: Port Phillip Bay, Victoria
Source: Port of Melbourne Corporation and ATSB
Investigation direction
The investigation is ongoing and will focus on weather events in the Port of Melbourne area, and how the port and port users prepare for such events. This will include:
The ship’s managers’ (TT line Company) preparations and procedures
Spirit of Tasmania II preparations and procedures for port stays and weather events, including mooring arrangements and equipment
Port of Melbourne procedures and actions
Bureau of Meteorology weather forecasting and warnings
distribution of weather information and warnings to and amongst port users.
Sources and submissions
Sources of information
On 14 January 2016, investigators from the Australian Transport Safety Bureau (ATSB) attended Spirit of Tasmania II while the ship was in Melbourne, Victoria. The master and directly involved crew members were interviewed and each provided their account of the occurrence. Photographs of the ship and copies of available, relevant documents were obtained.
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 the afternoon of 10 October 2014, the pilot of an amateur-built One Design DR-107 aircraft, registered VH-EGT, was performing low-level aerobatic manoeuvres. The manoeuvres were being performed to the east of Goolwa Airport, South Australia.
Witnesses described the aircraft performing a series of similar manoeuvres. Each involved a vertical climb and tumbling manoeuvre followed by a vertical dive and a low altitude recovery.
Witnesses reported that, during recovery from the last vertical dive, the aircraft collided with terrain. The aircraft was destroyed by the impact and the pilot was fatally injured.
What the ATSB found
The ATSB found no evidence of pilot incapacitation or a mechanical fault with the aircraft that could have contributed to the accident. There was insufficient evidence to determine why the recovery was not accomplished above the pilot’s minimum authorised aerobatics height.
The Civil Aviation Safety Authority (CASA) recommends that pilots performing low-level aerobatics undertake regular peer reviews due to the high level of skill and fine safety margins involved. The ATSB found no evidence of the pilot undertaking a peer review of their aerobatic performance in the 15 months prior to the accident.
Finally, the ATSB identified a safety issue that CASA does not require builders of amateur-built experimental aircraft to produce a flight manual, or equivalent, for their aircraft following flight testing. Without a flight manual, the builder, subsequent owners and other pilots do not have reference to the operational and performance data necessary to safely operate the aircraft.
What’s been done as a result
In response to the identified safety issue, the ATSB has issued a safety recommendation to CASA to take action to require builders of amateur-built experimental aircraft to produce a flight manual, or equivalent, for their aircraft following flight testing.
Safety message
This accident highlights the risks inherent in performing low-level aerobatics. Applying the recommendations in CASA civil aviation advisory publication CAAP 155-1(0) Aerobatics will reduce these risks. Specifically, pilots are encouraged to always maintain minimum approved heights above the ground when performing aerobatics and to engage in regular peer reviews.
Owners of amateur built experimental aircraft are also encouraged to ensure a comprehensive and accurate flight manual, or equivalent, is available for reference by themselves, subsequent owners and other pilots who may fly the aircraft.
Amateur-built One Design DR-107 aircraft, registered VH-EGT
Source: FlightAware
The occurrence
On the afternoon of 10 October 2014, the pilot of an amateur-built One Design DR-107 (DR-107) aircraft, registered VH-EGT (EGT), was performing low-level aerobatic manoeuvres to the east of Goolwa Airport, South Australia. The aerobatics were observed by a number of witnesses and described as consisting of a series of repeated manoeuvres. Specifically, the aircraft was seen to conduct a number of vertical climbs and tumbling manoeuvres, followed by a vertical dive and a low altitude recovery.
Witnesses observed that as the aircraft started to recover from a vertical dive it collided with terrain. The two witnesses furthest from the aircraft reported that the aircraft was rolling or spiralling while in the final vertical dive. By contrast, the two closest witnesses reported the aircraft was not rolling or spiralling during the final vertical dive.
The collision occurred at about 1430 Central Daylight-saving Time[1] in a paddock to the east of Goolwa Airport (Figure 1). Emergency services received a telephone call from a witness to the accident at 1431. Police, fire and ambulance personnel arrived at the accident site by 1443. The aircraft was destroyed by the impact and the pilot was fatally injured.
Figure 1: Goolwa Airport showing the location of the aircraft wreckage and witness locations. The runway is oriented basically north/south
The pilot held a Private Pilot (Aeroplane) Licence, issued in March 1983, with the appropriate aircraft endorsements to operate a DR-107-type aircraft. The pilot also held a valid and unrestricted Class 2 Aviation Medical Certificate, issued by the Civil Aviation Safety Authority (CASA).
The pilot’s last recorded biennial aeroplane flight review was on 4 September 2013. Entries in the pilot’s logbook recorded a total flying experience of 993 hours to 20 September 2013. There were no further entries in the pilot’s logbook after this date. The pilot had logged a total of 201.8 hours in EGT to 20 September 2013. From that date, entries in EGT’s maintenance release indicated the pilot flew the aircraft for a further 16.8 hours. It could not be determined if the pilot flew any aircraft other than EGT after 20 September 2013.
In November 1990 the pilot was assessed as competent, by a CASA Approved Testing Officer, to recover from spins in a Cessna 152-type aircraft and to perform basic aerobatic manoeuvres. The manoeuvres included loops, aileron rolls, slow rolls, barrel rolls and stall turns.
In order for CASA to grant a low-level aerobatics approval, pilots were required to demonstrate proficiency at progressively lower levels. The pilot was granted progressively lower low-level aerobatics approvals as follows:
in August 2009, the pilot was found competent to recover from inverted spins and to perform low-level aerobatics down to 500 ft above ground level[2]
in August 2011, the pilot was found competent to perform low-level aerobatics down to 330 ft and to perform non-aerobatic manoeuvres down to 100 ft.
In August 2013 CASA renewed the pilot’s low-level aerobatics approval for a further 2 years. The pilot’s continued competence to perform low-level aerobatics was not re-assessed prior to this renewal.
The pilot exceeded the recommended minimum recent experience for low-level aerobatics contained in Civil Aviation Advisory Publication (CAAP) 155-1(0) Aerobatics.
Aircraft information
General
The aircraft was a single seat, low wing, fixed gear, amateur-built[3] aircraft designed for competition aerobatics (Figure 2). Entries in the aircraft’s logbook indicated that the pilot commenced construction of the aircraft as an amateur builder in October 2003. The aircraft was completed in March 2008. A CASA authorised person issued a special certificate of airworthiness in the experimental category on 13 March 2008.
Figure 2: Amateur-built One Design DR-107 aircraft, registered VH-EGT
Source: FlightAware
The last entry in the aircraft’s maintenance records was the removal, by the pilot,[4] of the propeller and engine in September 2013. The pilot removed the engine and propeller from the aircraft for overhaul following a propeller overspeed. There was no record of the engine and propeller overhaul or subsequent installation in the aircraft. There was also no record of the last annual inspection performed on the aircraft.
On 14 January 2014, the pilot issued a maintenance release that was valid for 12 months. This allowed the aircraft to be operated privately under the day visual flight rules.[5] The aircraft flew for 16.8 hours between 14 January 2014 and the accident. No defects or unserviceable equipment endorsements were recorded on the maintenance release.
Aircraft weight and balance
The pilot, as builder of the aircraft, determined the aircraft’s empty weight and balance limits and produced a weighing summary document in March 2008. The empty weight was recorded to be 475 kg. The ATSB found no record of a maximum take-off weight (MTOW) in the aircraft’s records or the CASA aircraft file. No flight manual or placards relating to the aircraft’s weight and balance were found.
The kit supplier of the plans and building materials for the aircraft specified an empty weight of 322 kg and a MTOW of 517 kg. The aircraft’s weighing summary contained an aerobatic weight of 610 kg, which is 93 kg above the kit supplier’s listed MTOW. CASA allowed builders of amateur-built experimental aircraft to nominate their own MTOW. However, builders are required to demonstrate that their aircraft are safe to fly at their nominated MTOW during flight testing. Logbook entries indicated that the aircraft was test flown at approximately 610 kg on 1 and 2 June 2008.
The aircraft designer determined that the aircraft was capable of withstanding a flight load factor of plus or minus 10 g[6] at a weight of 454 kg. Operations at weights above this required a corresponding decrease in the maximum flight load factor. This included a corresponding reduction in the aircraft’s maximum manoeuvring speed (VA).[7] The ATSB determined that the manoeuvring speed on the aircraft’s airspeed indicator was marked appropriately for a 610 kg aerobatic weight. The VA marking on the airspeed indicator, in the absence of a flight manual, indicated the application of a reduced flight load factor limit.
The ATSB surveyed DR-107 owners on the Australian civil aircraft register to place the aircraft’s MTOW in context with other aircraft of the same type. Reported empty weights varied from 408 kg to 493 kg. MTOWs varied from 550 kg to 669 kg. One responder stated that their aircraft did not have a MTOW.
Flight manuals
The ATSB found no evidence that a flight manual or equivalent placarding was produced for the aircraft. A flight manual documents emergency procedures, systems information, operational and performance data necessary to safely operate an aircraft. For certified aircraft,[8] a flight manual is produced by the aircraft manufacturer for use by any pilot who flies the aircraft. For an amateur-built experimental aircraft, the builder of the aircraft is considered the manufacturer.
As each amateur-built aircraft is unique, CASA requires the builder to test their aircraft following construction. The purpose of flight testing is to determine that the aircraft is safe to fly and to determine the aircraft’s flight limits and performance characteristics. CASA recommends, but does not require, builders of amateur-built experimental aircraft to produce a flight manual for their aircraft following flight testing.
The ATSB’s survey of DR-107 owners indicated that half of the responders did not have a flight manual. The owners without a flight manual were not the builders of their aircraft, having purchased their aircraft from the builder or a subsequent owner.
Meteorological information
The Bureau of Meteorology did not provide observations or forecasts for Goolwa Airport. The area forecast[9] covering Goolwa Airport indicated that a trough would pass over the airport from the south-west at around the time of the accident. Low altitude winds were forecast to change from the north-west to the south-west as the trough passed.
Weather observations from nearby Victor Harbour and Hindmarsh Island indicated that the trough passed Goolwa at least 2 hours before the accident.
Witnesses at Goolwa Airport reported that the weather was fine and sunny with good visibility at the time of the accident. Witnesses also reported a ‘strong wind’ coming from the south-west. Due to low terrain to the south-west of Goolwa Airport, the presence of mechanical turbulence was considered unlikely.
Wreckage and impact information
On-site examination
The accident site was in a flat, recently harvested paddock adjacent to Goolwa Airport. The aircraft collided with terrain approximately 400 m east of the northern end of runway 01/19.[10] The wreckage trail was approximately 45 m long on a bearing of 115°. The length of the wreckage trail, combined with the initial ground impact mark and damage to the aircraft, indicated an impact at relatively high vertical and horizontal speed. Ground impact marks and aircraft damage further indicated that the aircraft collided with terrain in a wings level, slightly nose down pitch attitude (Figure 3).
Figure 3: Initial ground scar
Source: ATSB
Fuel soaked soil was identified under the wreckage, indicating that the aircraft’s fuel tank contained fuel prior to its disruption during the impact sequence. No evidence was found of any fault with the aircraft that could have contributed to the accident.
Propeller ground impact marks, blade dispersion and damage was consistent with the engine operating under power at the time of the accident. Witness reports of engine noise were consistent with the engine operating normally up to the collision with terrain. There was no evidence of an in-flight fire or break-up.
The aircraft was not fitted with a fixed emergency locator transmitter, nor was it required to be by regulation.
Medical and pathological information
The forensic pathologist who conducted the post-mortem examination concluded that the pilot succumbed to injuries sustained during the impact sequence. No abnormalities were identified that could have led to pilot incapacitation.
Toxicology results did not identify any substances that could have impaired the pilot’s performance.
Operational information
Aerobatic manoeuvres
Witnesses reported that the pilot was performing low-level aerobatic manoeuvres on the day of the accident, including vertical dives. Vertical dives meet the definition of an aerobatic manoeuvre contained in CAAP 155-1(0). There was insufficient evidence to determine the height at which the pilot was recovering from the vertical dives.
The ATSB was unable to determine the reason why the pilot was performing low-level aerobatics. However, the pilot had previously performed air show aerobatic routines and may have been practicing for an upcoming performance.
An experienced aerobatic pilot pointed to the possibility that the aircraft was in a spin,[11] which may have become an inverted spin during the final descent. While there was insufficient evidence to confirm that proposition, such a development would have required additional time, and therefore height, to recover the aircraft to level flight.
The ATSB obtained video evidence of the pilot performing aerobatic manoeuvres at Goolwa Airport significantly below 330 ft 1 week prior to the accident. This was below the height that the pilot was permitted to engage in aerobatic flight.
Peer reviews
Due to the ‘high level of skill and fine safety margins’ in low-level aerobatics, CAAP 155-1(0) part 7.28.1 strongly suggested pilots undertake regular peer reviews of their aerobatic performance. In this respect, Part 7.28.2 of the CAAP stated:
The peer review process is intended to provide an independent assessment by a similarly qualified person or persons on the way the pilot conducts the activity and to identify any incorrect techniques or practices that the pilot may have developed over time. It is not intended to be a flight test for the renewal of the permission, but an opportunity for constructive discussion with other practitioners with a view to enhancing the safety of a pilot’s performance.
CAAP 155-1(0) recommended a maximum of 15 months between reviews. The ATSB was unable to find any evidence of the pilot undertaking a peer review of their aerobatic performance in the 15 months before the accident.
The fatal/serious injury accident rate across the period of the study was significantly higher for amateur-built aircraft (average 1.27 per 10,000 hours) than it was for similar factory-built aircraft (average 0.22). The fatal and serious injury accident rate was more than 5.5 times higher for amateur-built aircraft compared to factory-built during private operations.
Similar to the total accident rate, the fatal/serious injury accident has reduced from 1988-1999 to 1999-2010, but the reduction has been significantly greater for amateur-built aircraft. In the second half of the period of study from 1999-2010, the fatal/serious injury accident rate was more than 3.5 times higher for amateur-built aircraft.
Those results were consistent with the findings of the United States National Transportation Safety Board (NTSB) safety study NTSB/SS-12/01 The Safety of Experimental Amateur-Built Aircraftavailable at NTSB website. The abstract of that study noted that:
Experimental amateur-built (E-AB) aircraft represent nearly 10 percent of the U.S. general aviation fleet, but these aircraft accounted for approximately 15 percent of the total—and 21 percent of the fatal—U.S. general aviation accidents in 2011…
The NTSB study also stated that:
Areas identified for safety improvement include expanding the documentation requirements for initial aircraft airworthiness certification, verifying the completion of Phase I flight testing, improving pilots’ access to transition training and supporting efforts to facilitate that training, encouraging the use of recorded data during flight testing, ensuring that buyers of used E-AB aircraft receive necessary performance documentation, and improving aircraft identification in registry records.
As a result of their safety study, the NTSB made a number of recommendations to the United States Federal Aviation Administration (FAA) that were aimed at improving the safety of amateur-built aircraft. These included recommendations that the FAA:
Revise 14 Code of Federal Regulations 21.193, Federal Aviation Administration Order 8130.2G, and related guidance or regulations, as necessary, to require applicants for an airworthiness certificate for experimental, operating amateur-built aircraft to submit for Federal Aviation Administration acceptance a flight test plan that will (1) ensure the aircraft has been adequately tested and has been determined to be safe to fly within the aircraft’s flight envelope and (2) produce flight test data to develop an accurate and complete aircraft flight manual and to establish emergency procedures and make a copy of this flight test plan part of the aircraft’s certification file. (A-12-29)
…
Revise Federal Aviation Administration Order 8130.2G, and related guidance or regulations, as necessary, to require the review and acceptance of the completed test plan documents and aircraft flight manual (or its equivalent) that documents the aircraft’s performance data and operating envelope, and that establishes emergency procedures, prior to the issuance of Phase II operating limitations. (A-12-32)
The FAA responded to these NTSB recommendations on 24 September 2012 and advised that they were ‘creating a cross organizational Amateur-Built Safety Team to review the current guidance and policy for amateur-built certification and operation.’ At the time of writing, no further safety action had been reported to the NTSB.
Related occurrences
A review of the ATSB occurrence database identified three potentially similar accidents that occurred during aerobatic manoeuvres. Of these, two involved amateur-built aircraft. One of the occurrences was preceded by a loss of engine power during take-off. There was insufficient information available on the circumstances of the other two occurrences to determine if they were substantially similar to this accident.
While performing aerobatic manoeuvres the pilot did not fully recover the aircraft from a vertical dive before colliding with terrain. The ATSB did not find any evidence of pilot incapacitation or a fault with the aircraft that could have contributed to the accident. The weather conditions were also considered unlikely to have influenced the development of the accident. Additionally, the pilot was qualified to perform low-level aerobatics down to 330 ft and the aircraft type was appropriate for the aerobatic manoeuvres being performed that day.
This analysis will consider the possible reasons why aerobatic flight was continued below 330 ft. In addition, the safety benefit of aerobatic peer reviews and provision of aircraft flight manuals for amateur built experimental aircraft will be discussed.
The occurrence
The accident site ground impact marks and aircraft damage indicated that the aircraft was in a slightly nose low, wings level attitude at impact. Additionally, the ATSB determined that the aircraft collided with terrain with a high vertical and horizontal speed. This evidence is consistent with witness reports indicating that the aircraft appeared to be pulling out of a dive when it collided with terrain.
There was insufficient evidence to determine why the recovery was not accomplished above the pilot’s minimum aerobatics height of 330 ft. It is possible that the pilot either intentionally or inadvertently delayed the recovery of the aircraft during the vertical dive.
Misjudgement of the height that recovery was initiated
The ATSB was unable to determine the intended lowest height of the aerobatics on the day of the accident. However, evidence was provided to the ATSB that the pilot performed aerobatic manoeuvres significantly below 330 ft 1 week prior to the accident.
It is possible that the pilot was completing the aerobatic manoeuvres below 330 ft on the day of the accident. If this occurred, a misjudgement of the recovery initiation height may have resulted in insufficient remaining height above terrain for the pilot to recover the aircraft from the vertical dive before impacting terrain.
Inadvertent late initiation of the recovery
Raising the aircraft’s pitch attitude from vertical nose-down to close to horizontal while maintaining the wings level required active inputs by the pilot and flight control authority. Consequently, pilot incapacitation or a fault with the aircraft’s flight controls were considered unlikely. Momentary incapacitation of the pilot or an intermittent aircraft fault that distracted the pilot and delayed initiation of the recovery; however, could not be ruled out.
Witness descriptions of the aircraft rolling or spiralling were consistent with the aircraft being in a spin during the final descent. However, the two closest witnesses described the aircraft descending vertically without spinning. Additionally, observations of the attempted recovery and accident site ground impact marks indicated that the aircraft was not in a spin when it collided with terrain. If the aircraft had inadvertently entered an inverted spin at some stage during the vertical dive, additional height would have been required to recover the aircraft to level flight. In that case, the possibility that there was insufficient height available to fully recover the aircraft could not be ruled out.
Aerobatics peer review
The ATSB found no evidence of the pilot undertaking a peer review of their aerobatic performance in the 15 months prior to the accident. A peer review, as suggested by Civil Aviation Advisory Publication 155-1(0) Aerobatics has the potential to help a pilot maintain safety margins in low-level aerobatic routines, and may have assisted the pilot avoid inadvertently breaching their minimum approved aerobatics height. However, there was insufficient evidence to determine if the non-completion of the peer review influenced the development of the accident.
Aircraft flight manual
The Civil Aviation Safety Authority did not require amateur-built experimental aircraft to have a flight manual or equivalent placards. The ATSB found no evidence that a flight manual or equivalent placarding was produced for the aircraft following flight testing. The lack of a flight manual was unlikely to have influenced this accident due to the pilot's familiarity with the aircraft. This familiarity was a result of their experience building, test flying and operating the aircraft.
However, as evidenced by the recommendations made to the United States Federal Aviation Administration by the National Transportation Safety Board, not having a flight manual increases the risk associated with amateur-built experimental aircraft operations. Without a flight manual the builder, other pilots and especially subsequent owners do not have reference to operational and performance data necessary to safely operate the aircraft. Given that accidents involving amateur-built aircraft occur at a significantly higher rate than comparable factory-built aircraft, a requirement to document important operational information would be a valuable safety enhancement.
Findings
From the evidence available, the following findings are made with respect to the collision with terrain involving One Design DR-107 aircraft, registered VH-EGT, which occurred near Goolwa Airport, South Australia on 10 October 2014. These findings should not be read as apportioning blame or liability to any particular organisation or individual.
Safety issues, or system problems, are highlighted in bold to emphasise their importance. A safety issue is an event or condition that increases safety risk and (a) can reasonably be regarded as having the potential to adversely affect the safety of future operations, and (b) is a characteristic of an organisation or a system, rather than a characteristic of a specific individual, or characteristic of an operating environment at a specific point in time.
Contributing factors
The aircraft collided with terrain while the pilot was attempting to recover from an aerobatic manoeuvre at low level.
Other factors that increased risk
Although suggested by Civil Aviation Advisory Publication 155-1(0) Aerobatics, the pilot probably did not undertake a peer review of their aerobatic performance in the preceding 15 months to the accident.
The Civil Aviation Safety Authority did not require builders of amateur-built experimental aircraft to produce a flight manual, or equivalent, for their aircraft following flight testing. Without a flight manual the builder, other pilots and subsequent owners do not have reference to operational and performance data necessary to safely operate the aircraft. [Safety issue]
Sources and submissions
Sources of information
The sources of information during the investigation included:
an experienced aerobatic pilot
the Bureau of Meteorology
the Civil Aviation Safety Authority (CASA)
the South Australian Police and Forensic Science SA
the One Design DR-107 designer
a number of other One Design DR-107 aircraft owners.
Submissions
Under Part 4, Division 2 (Investigation Reports), Section 26 of the Transport Safety Investigation Act 2003 (the Act), the 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 experienced aerobatic pilot, the Sport Aircraft Association of Australia and CASA.
Submissions were received from the experienced aerobatic pilot, the Sport Aircraft Association Australia and CASA. 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.
On 28 March 2016, at about 1306 Eastern Daylight-saving Time (EDT), a Schemp-Hirth Janus glider, registered VH-GWQ (GWQ) launched from Porepunkah Airfield, Victoria, for a pleasure flight. On board were two pilots. The pilot seated in the rear seat was the pilot in command for the flight. The glider tracked over Simmons Gap, to a ridge about 3 km north-west of Mount Beauty Airport (Figure 1). The pilots could hear and see other gliders being towed onto the ridge. They joined a thermal[1] and climbed in tight orbits (‘thermalling’) in a clockwise direction.
Figure 1: Relative tracks of gliders VH-GWQ and VH-CQP and positions at 1355:02
Source: Gliding Federation of Australia
At about 1335, the pilot of a Rolladen-Schneider LS3-A glider, registered VH-CQP (CQP), launched from Mount Beauty Airport, Victoria, for a pleasure flight. At about 1355, the glider was 3 to 4 km north-west of the airfield and descending through about 4,000 ft, when the pilot heard an alarm sounding, but did not identify it as issuing from the FLARM collision avoidance system (see FLARM below) fitted to the glider. The glider was tracking to the north, and the pilot reported that they had been keeping a lookout for other gliders but were not aware of any in the vicinity at the time.
The pilot tried to identify the source of the alarm inside the cockpit, which diverted their attention from looking outside. As the pilot became stressed by the noise, particularly as it became ‘quite shrill’, the cockpit fogged up, further reducing the pilot’s ability to see outside.
At that time, GWQ was thermalling and in a right bank at about 40–45°, and had completed four orbits. The front seat pilot sighted a glider approaching from the opposite direction at about the same altitude. They assumed that the glider would join the thermal behind them, in the same direction, and on the opposite side of the orbit, in accordance with normal procedures. The front seat pilot asked the rear seat pilot whether they could see the glider, who responded ‘no’. The FLARM fitted to their glider indicated that there was another glider in close proximity and the rear seat pilot looked outside to see where it was.
The front seat pilot assessed that the approaching glider was not going to manoeuvre to join the thermal or to avoid a collision, so took control of the glider and pushed the stick forwards to descend rapidly. The other glider (CQP) passed overhead.
The pilot of CQP sighted a glider pass below, and estimated there was less than 100 ft vertical separation. Both gliders continued their flight for about another hour after which GWQ landed at Porepunkah and CPQ landed at Mount Beauty without further incident.
Flight data
According to the flight data recorded by the gliders’ flight logger, at 1354:58, CQP was at 3,606 ft and GWQ at 3,523 ft. Four seconds later as the gliders’ paths crossed, CQP was at 3,605 ft and GWQ had descended to 3,458 ft.
FLARM
FLARM is a collision avoidance system that shows other similarly equipped aircraft in the vicinity. The display shows the approximate direction of detected traffic and whether it is above, below or at about the same level (Figure 2).
Each FLARM device determines its position and altitude with a highly sensitive state of the art GPS receiver. Based on speed, acceleration, heading, track, turn radius, wind, altitude, vertical speed, configured aircraft type, and other parameters, a very precise projected flight path can be calculated. The flight path is encoded and sent over an encrypted radio channel to all nearby aircraft at least once per second.
At the same time, the FLARM device receives the same encoded flight path from all surrounding aircraft. Using a combination of own and received flight paths, an intelligent motion prediction algorithm calculates a collision risk for each received aircraft based on an integrated risk model. The FLARM device communicates this, together with the direction and altitude difference to the intruding aircraft, to the connected FLARM display. The pilots are then given visual and aural warnings and can take resolutive action.
Pilot comments
Pilot of VH-CQP
The pilot of CQP reported that they had flown gliders fitted with FLARM for 7–8 years and had never heard it make a noise before. This may have been because they had never been close enough to another glider to trigger the alarm before. They were briefed and had a briefing note circulated by the gliding club when they were first installed. The pilot did not think there were any other gliders in the vicinity, and did not associate the alarm with FLARM.
The pilot had a VHF radio with the local area frequency selected, but did not make or hear any broadcasts regarding GWQ.
Pilots of VH-GWQ
The pilot in the front seat of GWQ reported that there were some radio broadcasts at the time, mainly from the glider tug pilots in the circuit at Mount Beauty and Porepunkah. They had not made any broadcasts, and had not heard any from CQP.
The pilot in the rear seat commented that the head and shoulders of the pilot in the front seat obscured their vision immediately ahead at the same level. When the FLARM sounded, rather than looking at the display, they looked outside for the other glider.
The pilot in the rear seat further reported that the FLARM unit in CQP had recently been upgraded to a PowerFlarm. This may have included a new display, and also may have been indicating ADS-B transmissions. Changes to display and aural warnings of the FLARM fitted to CQP may have been confusing for the pilot of CQP.
Safety message
The glider pilots reported that see and avoid was the usual means of maintaining separation from other gliders. It was not uncommon to be in close proximity to other gliders, particularly when thermalling. They did not normally broadcast their position or intentions when thermalling, and expected other glider pilots to adhere to standard procedures.
Avoidance systems such as FLARM can enhace safety in non-controlled airspace by detecting conflicting aircraft also fitted with a compatible system. These assist in alerting pilots to the presence of other aircraft and directing them where to look. The ATSB report Limitations of the See-and-Avoid Principle outlines the major factors that limit the effectiveness of un-alerted see-and-avoid. Insufficient communication between pilots operating in the same area is the most common cause of safety incidents near non-controlled aerodromes.
It is essential that when equipment is installed in an aircraft, pilots have an understanding of its operation and are familiar with its characteristics.
The following publications provide valuable and relevant references for glider pilots:
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 3 April 2016, the pilot of a de Havilland DH-82 aeroplane, registered VH-BJE, conducted a 30‑minute scenic flight from Redcliffe Airport, Queensland, with one passenger on board. The weather was fine, with wind was from the east-northeast at less than 8 kt, no precipitation, and the runway was dry.
On returning to Redcliffe, the pilot elected to join the circuit on a mid-field crosswind leg for runway 07. The pilot reported that the approach was normal.
As the aircraft landed, the pilot reported the tail was slightly higher than normal, but the aircraft’s speed was normal. The aircraft wheels touched down at the pilot’s aiming point, about half way along the grass strip to the right of the sealed runway. The pilot reported that the wheels seemed to dig in. The aircraft nose pitched down, the propeller struck the grass runway, and the aircraft rolled over forwards, coming to rest inverted (Figure 1).
The pilot and passenger were uninjured. The aircraft sustained substantial damage.
Figure 1: Accident site showing damage to VH-BJE
Source: Ron Ennis – modified by the ATSB
Pilot comments and experience
At the time of the accident, the pilot had a total of 259.3 hours of aeronautical experience, including 7.9 hours on the aircraft type. The pilot held tailwheel and aerobatic endorsements, obtained in an American Champion/Bellanca Citabria aircraft.
The pilot commented that the ground was a bit soft where the wheels had touched down, but that they had landed there twice previously that day without incident. In future, the pilot would land with a slightly higher nose attitude.
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 7 April 2016, the pilots of two Robinson R22 helicopters flew from Mossman, Queensland to various fishing locations to the north with a passenger in each helicopter. Late in the afternoon, the pilots commenced the direct return flight to Mossman. However, the pilots encountered weather and winds that slowed their progress and required them to refuel at Cooktown.
The pilots departed Cooktown at last light intending to track via the coast to Mossman. As the flights progressed, the light available from the sun continued to decrease and there was no moon. There were also patches of cloud and rain in the general area.
Shortly after passing Cape Tribulation, in dark night conditions, one of the helicopters, registered VH-YLY (YLY), collided with the sea. The passenger was injured in the accident but was able to reach the shore and notify emergency services. Unaware of the accident, the occupants of the other helicopter continued to Mossman.
A search was initiated and the missing helicopter was located on 9 April 2016 in about 400 m offshore in about 10 m of water. The pilot was not located.
What the ATSB found
The ATSB found that the pilot of YLY, who was only qualified to operate in day-VFR conditions, departed on a night flight and continued towards the destination in deteriorating visibility until inadvertently allowing the helicopter to descend into water.
The ATSB also identified the following other factors that collectively increased risk:
an unapproved modification attached to the skids of YLY
exceedance of weight and balance limitations
non-carriage of life jackets
incomplete operational information
overdue calibration checks of the helicopter pitot-static system and altimeter.
Safety message
To avoid the usually fatal consequences of losing visual reference, day-VFR pilots need to plan to arrive at their destination at least 10 minutes before last light and to have a realistic ‘plan B’ to use when it becomes apparent that the intended flight cannot be completed in daylight. A further consideration for pilot decision-making about flying conditions is the degree to which passengers are also exposed to risk.
Key messages from the ATSB Avoidable Accidents series report No.7 highlight that some nights and some terrain are darker than others, and inadvertently flying into instrument meteorological conditions is also harder to avoid at night. Pilots need to be mindful of similar messages provided in pilot operating handbooks that refer to risks associated with loss of visibility and night flight in bad weather.
The occurrence
On the morning of 7 April 2016, the pilot of a Robinson Helicopter Company R22 helicopter, registered VH-YLY (YLY) flew from Mareeba to Mossman, Queensland, to join the pilot of another R22 for a fishing trip (Figure 1). Both pilots were qualified to operate the helicopters on private‑category operations by day under the visual flight rules[1].
The two pilots departed Mossman at about 0800 EST[2] with a passenger in each of the two‑seat helicopters. The pilots tracked north to Cooktown to refuel then continued northward to Pipon Island, landing at various coastal locations so they and the passengers could do some fishing. At one of the landing sites the pilots were able to fill the helicopter fuel tanks from a drum of fuel.
By the time the helicopters arrived at Pipon Island, the occupants had caught between 20 and 30 kg of fish, which were carried in a non-standard container attached to the left skid of YLY. After spending some time at Pipon Island, the pilots departed separately somewhere between 1600 and 1700 to return to Mossman. The pilots had used GPS route information and local weather conditions, rather than the required aviation Area Forecast,[3] to decide that they had sufficient fuel and daylight to make it to Mossman, a direct-track distance of 148 NM (274 km).
Figure 1: Location of Pipon Island with Cooktown and Mossman destinations showing position of VH-YLY accident site south of Cape Tribulation. The inset shows Cape Tribulation relative to far north Queensland
Source: Google Earth and modified by ATSB
During the return flight, squalls and headwinds of about 20 kt were encountered that slowed progress and necessitated a landing at Cooktown to refuel. Relative to the other helicopter, YLY was slower likely due to the aerodynamic drag of the skid‑mounted container, the additional weight of the fish, and doors-off operation.
Recorded fuel transaction information showed the time of fuel uplift at Cooktown occurred at 1836. Given last light for Cooktown was calculated to be 1838 (based on ideal conditions), night conditions existed by the time the pilots were ready to depart. Neither pilot was qualified to fly at night or in low visibility conditions that would require instrument flying. Additionally, the helicopters were not equipped with an artificial horizon instrument and lacked other equipment required for flight at night under Australian regulations.
The passenger in YLY reported he was concerned about the available light and weather conditions and queried the pilot’s intention to continue the flight after Cooktown. Without elaborating, the pilot indicated he intended to continue and the flight (direct-track distance of 60 NM) would only take 45 minutes. None of the pilots or the passengers reported any specific reason to arrive at Mossman that evening. The pilots did not discuss staying in Cooktown.
Both helicopters departed Cooktown at about 1840. A witness who observed the departure reported it was getting dark, which was consistent with it being after last light and no moonlight (local moon-set was 1811). Due to the slower speed of YLY, the other helicopter drew ahead. Every 10 minutes or so, the pilot of the lead helicopter, who was the more experienced pilot, checked in with the pilot of YLY by radio.
According to the pilot in the lead helicopter, and as had been discussed, they flew close to the coast intending to keep an outline of the mountains as a visual reference. In ‘bad conditions’, the proposed method was to ensure the altimeter was accurate relative to sea level and use it to fly not below 250 ft above the water. If the conditions got ‘really bad’, in that it got too dark and the outline of the mountains or sight of the water was being affected, the plan discussed between the pilots was to land on a beach. The pilot of the lead helicopter advised they had done this on previous occasions due to bad weather, though only in daylight.
The coastal route between Cooktown and Mossman has few settlements and from halfway, most of the coastline is part of the Cape Tribulation section of the Daintree National Park. It is a remote area and, as such, there is little or no ground lighting.
Initially, the weather conditions were reported to be clear but about 15-20 minutes into the flight, the pilot of the lead helicopter encountered small squalls with a bit of cloud. After passing Cape Tribulation, the pilot of the lead helicopter radioed the pilot of YLY who advised he had just passed the cape and did not report any difficulties.
The passenger in YLY recalled that at one stage the helicopter descended close to the water before the pilot corrected and climbed. He further stated that visibility decreased until it was dark and the pilot dimmed the cockpit lights to reduce glare off the windscreen. The passenger recalled passing the Cape Tribulation campgrounds and was aware that there were people on the beach as indicated by campfires and a spotlight being waved at the passing helicopters.
To the passenger, it got ‘really dark’ and concerning so he suggested landing on the beach. The pilot did not respond verbally but it appeared to the passenger that they might be descending to land. Suddenly the passenger saw the ocean more clearly followed almost immediately by a ‘massive bang’ as the helicopter impacted the water.
When the passenger regained consciousness, he was strapped into the helicopter on the ocean floor. He was able to release himself, reach the surface, and tread water in a heavy swell. Injuries to a leg and arm restricted swimming but the tide carried him to the beach where he was able to make contact with campers and notify the authorities. Campers had heard an impact and already reported it to police.
About 10 minutes after the radio call near Cape Tribulation, the pilot in the lead helicopter tried unsuccessfully to contact the pilot of YLY. He recalled that, at the time, there was light cloud, mist, showers, and strong winds. The lead pilot said he was concerned about YLY and wanted to turn back but was prevented by limited time, strong winds, and marginal visibility. So the pilot continued to Mossman.
A search was mounted but the missing pilot was not found.
Queensland Police with the assistance of specialist divers located the empty wreckage about 400 m seaward of Noah Beach. The wreckage was not recovered but divers examined the wreckage and recorded underwater video imagery, which was provided to the ATSB (Figure 2).
Figure 2: Underwater image of VH-YLY showing significant damage to the nose and belly panel of the helicopter particularly to the right forward section and loss of the cockpit upper portion
Extensive damage to the right side of the helicopter, including the pilot’s seat belt fitting found torn from its mount, indicated a significant right‑side impact with the water. The main rotor blades and transmission were present and similarly damaged. The tailboom was not identified in the footage and it likely became detached during the accident sequence. The damage to the helicopter and rotor system was consistent with powered flight into the water.
A maintenance release had been issued for YLY authorising VFR Day operation only. A check calibration of the aircraft pitot-static system and altimeter was due in October 2015 and had yet to be certified as complete. As such, the accuracy of the airspeed indicator, vertical speed indicator and altimeter was not assured. A review of the helicopter logbooks did not find any reference to installation of the container to the left skid.
Based on the reported loading of YLY, on departure from Cooktown for Mossman the helicopter was estimated to be at least 35 kg over the maximum gross weight limitation. Although there was no weight and balance data for the skid-mounted container, the longitudinal and lateral centre of gravity were estimated to be outside limits on departure or as fuel was consumed.
Although some of the flying that day included flight over water, the pilots and passengers did not wear life jackets. This was contrary to the regulatory requirement for the occupants of single‑engine aircraft being operated beyond gliding/autorotation distance of land and while below 2,000 ft.
According to the applicable area forecast required for flight planning, in the area of operation east of the ranges and coast, there would be isolated showers with associated low cloud and reduced visibility. The wind was expected to be from the east to south-east at 15 to 20 kt. From 2300, isolated areas of low cloud were expected east of the ranges and coast. This was broadly consistent with the aerodrome forecasts for Cooktown and Cairns, except that the showers and low cloud were due at Cairns from 1700.
The closest official weather observation site to the accident location was at Low Isles, 15 NM (28 km) to the south. At the approximate time of the accident, the wind was from the south-east at 20 kt. No precipitation was recorded during the hour before and after the accident. The extent of cloud coverage was not measured at the site.
The recorded imagery from the Cairns weather radar showed a small area of light rain inland near Cape Tribulation and patches of light to moderate rain off the coast no closer than 15 NM (28 km).
For aviation purposes, night is defined as the period of darkness from last light (end of evening civil twilight) to first light (beginning of morning civil twilight). At last light, in ideal conditions, there will be enough light from the sun for large objects to be seen but no detail. As time passes, light from the sun further diminishes to reach a point where it is insufficient to allow a horizon to be seen at sea level. This point (end of evening nautical twilight) at Cape Tribulation was calculated to be 1919 but high terrain to the west would tend to make it effectively earlier.
The time of the accident was estimated to be 1930, which was about 10 minutes after end of evening nautical twilight.
Safety analysis
The pilots of the two helicopters conducted a series of flights to various fishing locations without incident. However, on the return trip from Pipon Island to Mossman late in the afternoon, the progress was slower than the pilots expected in part due to the prevailing weather conditions. The pilots diverted to Cooktown to refuel and landed about 10 minutes before last light.
As the pilots were only qualified to fly during daylight hours in VFR conditions, there was insufficient time for the pilots to conduct the flight to Mossman within the regulatory requirements and with an appropriate level of risk. Despite that, the pilots departed Cooktown just after last light for the nominally 45-minute flight to Mossman.
Influences on pilot decision-making are complex and the decisions by both pilots to depart Cooktown after last light are not analysed in depth. In simple terms, the pilots were motivated to reach the intended destination and either did not perceive the high level of risk, or had a high tolerance of that risk.
It is noted that both pilots had reportedly completed other flights in low visibility conditions, including landing on a beach when the conditions required. This may have provided a level of confidence in their ability to conduct the flights at night, which would be misplaced if it was based on their experience in daylight. In response to a draft of this report, the surviving pilot advised that he did not believe the flight was high-risk or that their ‘plan B’ was unrealistic.
As the flights progressed down the coast, the natural light was reducing. By 1919 (local end of nautical twilight), at low level over water and without moonlight or significant artificial lighting, the horizon would technically not be visible. Other factors adversely affecting visibility in the southward direction of flight were the high terrain to the west and potentially cloud and/or rain. In response to a draft of this report, the surviving pilot reiterated that he was able to see the horizon at all times and had sufficient light.
By the time YLY passed Cape Tribulation, the pilot was operating in dark night conditions with little to no terrestrial lighting. When the passenger expressed concerns to the pilot about variations in the helicopter’s height relative to the water, there was no apparent consideration of landing. It is likely that the prospect of landing in an unfamiliar area in darkness and windy conditions was not a favourable option for a pilot operating with minimal visual references. At this point, YLY was over halfway to the destination and turning back to Cooktown in the prevailing conditions was unlikely to be viewed as a realistic option. The pilot may also have been influenced to continue by the progress made by the more experienced pilot ahead of him.
It is not clear how the pilot of YLY was maintaining control of the helicopter. Without an artificial horizon instrument and the training to use it, the pilot was reliant on external reference to a natural horizon or surface features. As the horizon became less distinct during the flight, the pilot presumably maintained visual reference to the coastline and surface of the water. When the pilot dimmed the instrument panel lights to reduce internal glare and reflections, this was apparently necessary to assist with external reference.
Although the method proposed by the pilot of the lead helicopter to maintain height at low level was to use the altimeter, it would have been difficult for the pilot of YLY to monitor the instruments (including the vertical speed indicator) with dimmed cockpit lighting as well as scanning the surrounding surface features. This would be made more difficult by the windy conditions and inherent instability of helicopters.
Shortly after passing Cape Tribulation, the helicopter collided with water in a right skid-low attitude. Given the passenger account that the helicopter was operating normally and the accident occurred without warning, the pilot was probably unaware of the proximity of the water and/or was in the process of losing control of the helicopter. In the dark night conditions, the pilot could have also encountered an area of cloud and/or rain that wasn’t evident in the weather radar image.
The ATSB found that the pilot of YLY, who was only qualified to operate in day-VFR conditions, departed on a night flight and continued towards the destination in deteriorating visibility until inadvertently allowing the helicopter to descend into water.
It is concerning that some day-VFR helicopter pilots continue to operate illegally and unsafely by flying at night. Although pilots have done this with mixed outcomes, as the two pilots involved in this occurrence experienced, the risks of such operation are extremely high.
This was demonstrated on 24 April 2011, when an R44 registered VH-RUR, collided with the sea while on approach at a reduced airspeed to a helicopter landing site. The helicopter was being flown after last light by a pilot who was not approved, nor was the helicopter equipped to fly at night. The ATSB found that the flight was commenced with insufficient daylight to complete the planned flight under the day visual flight rules. Although the pilot survived, they had no visual reference with the sea surface immediately before the helicopter inadvertently descended into the water (ATSB Transport Safety Report AO-2011-051).
In similar circumstances to this occurrence, on 25 August 2014, the pilots of two R22 helicopters were delayed and were operating in dark night conditions to reach the planned destination. One of the pilots inadvertently allowed the helicopter to descend into terrain and was fatally injured (ATSB Transport Safety Report AO-2014-144).
The ATSB Avoidable Accidents series report No. 7 refers to accidents and the risks associated with visual flight at night. One of that report’s key messages was that some nights and some terrain are darker than others, and inadvertently flying into instrument meteorological conditions is also harder to avoid at night.
The helicopter manufacturer, Robinson Helicopter Company (RHC), included safety notices in the R22 Pilot’s Operating Handbook that target specific risks relating to loss of visibility (RHC Safety Notice SN-18) and night flight in marginal weather conditions (RHC Safety Notice SN-26). Copies of the safety notices referenced in this section are available on the RHC website.
To avoid the often fatal consequences of losing visual reference, day-VFR pilots need to plan to arrive at their destination at least 10 minutes before last light and to have a realistic ‘plan B’ to use when it becomes apparent that the intended flight cannot be completed in daylight. A further consideration for pilot decision-making about flying conditions is the degree to which passengers are exposed to risk.
In addition to the night flying risks, the ATSB identified the following other aspects of the operation that were not considered to be contributing safety factors but collectively increased risk:
The unapproved skid-mounted container increased the aerodynamic drag of the helicopter, increased capacity for overloading, affected lateral balance, and had implications for skid down-load and vibration-related loading (RHC Safety Notice SN-13 discusses the risk of attaching items to the helicopter skids).
The weight and balance of the helicopter during parts of the flight exceeded the limitations with associated adverse effects on handling, performance and component fatigue (RHC Safety Notice SN-37).
For over-water operations, non-carriage of life jackets compromised survivability following an accident involving a collision with water (such as this occurrence) or a ditching.
The pilots made operational decisions without taking aviation forecasts and reports or the availability of celestial or terrestrial lighting into account.
Operation of YLY with overdue calibration checks of the pitot-static system and altimeter increased the risk of unreliable instrument indications that could be problematic in low visibility conditions.
Findings
From the evidence available, the following finding is made with respect to the collision with water of a Robinson Helicopter Company R22 helicopter, registered VH-YLY that occurred near Cape Tribulation, Queensland, on 7 April 2016. These findings should not be read as apportioning blame or liability to any particular organisation or individual.
Contributing factor
The pilot of YLY, who was only qualified to operate in day-VFR conditions, departed on a night flight and continued towards the destination in deteriorating visibility until inadvertently allowing the helicopter to descend into water.
The sources of information during the investigation included the:
surviving occupants of the two helicopters
Cairns Forensic Crash Unit, Queensland Police
the Bureau of Meteorology
the Civil Aviation Safety Authority (CASA)
Airservices Australia
Geoscience Australia
References
ATSB (2013), Avoidable Accidents No. 7. Visual flight at night accidents: What you can’t see can still hurt you, ATSB, Canberra, Australia.
Robinson Helicopter Company (Rev. 1994), Safety Notice SN-13 Do not attach items to the skids.
Robinson Helicopter Company (Rev. 1994), Safety Notice SN-18 Loss of visibility can be fatal.
Robinson Helicopter Company (Rev. 1994), Safety Notice SN-26 Night flight plus bad weather can be deadly.
Robinson Helicopter Company (2001), Safety Notice SN-37 Exceeding approved limitations can be fatal.
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 other R22 pilot and CASA.
Submissions were received from the other R22 pilot and CASA. 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.
