On 15 September 2017, the ATSB commenced an investigation into an aircraft loading event involving Saab 340, VH-ZRB, Sydney Airport, New South Wales on 14 September 2017.

The flight was originally planned to be conducted in a different aircraft, and the captain had advised the trim staff that 239 kg of freight could be loaded. There were subsequently two aircraft changes prior to departure. After the flight crew had been assigned VH-ZRB to conduct the flight, the captain advised the trim staff that the 239 kg of freight was not to be loaded on the aircraft due to weight limitations. The trim staff advised the baggage staff of the change.

There was subsequently a misunderstanding between the baggage staff and the loader, resulting in the loader believing that a barrow containing passenger baggage and freight was approved to be loaded on to VH-ZRB. Subsequent communications between the loader and the trim staff were not conducted in accordance with the operator’s loading and freight confirmation procedure, and the error was not detected until after the aircraft had departed. Subsequent calculations revealed that the aircraft departed Sydney about 77 kg over the aircraft’s maximum take-off weight.

The ATSB obtained the operator’s investigation report into the occurrence and related material, and reviewed previous loading occurrences involving the operator. Based on this information, the ATSB considered it was very unlikely that further investigation would identify any systemic safety issues. The ATSB has discontinued the investigation, but will continue to monitor loading occurrences involving all operators.

What happened

At about 1330 Eastern Standard Time (EST), on 4 September 2017, the flight crew of a PNG Air GIE Avions de Transport Regional ATR 72-212A aircraft, registered P2-ATR, prepared to operate passenger charter flight CG950 from Cairns, Queensland, to Lihir Island Airport,^[1] Papua New Guinea. On board the aircraft were the captain, first officer, three cabin crew and 49 passengers.

While preparing for the flight, the flight crew noted the weather conditions included no significant cloud below 6,500 ft and visibility in excess of 10 km. The flight crew then contacted air traffic control (ATC) to obtain an airways clearance. The airways clearance included a departure from runway 33 via the CAIRNS TWO standard instrument departure (SID). This departure required the flight to maintain the runway track of 330° magnetic (M) until the aircraft climbed to an altitude of 500 ft above mean sea level (AMSL) and had passed the departure end of the runway. The SID then required the flight crew to turn the aircraft to a heading assigned by ATC which would be between 335°M and 070°M (Figure 1).

Figure 1: Cairns Two standard instrument departure

The figure shows the CAIRNS TWO departure including the direction to turn to an ATC assigned heading after passing the departure end of the runway (DER) and upon reaching 500ft. Source: Airservices Australia, annotated by ATSB.

After receiving the airways clearance, the flight crew identified the assigned heading segment of the SID. The captain noted that he had not previously departed using a SID with an assigned heading segment. The captain, acting as pilot flying,^[2] elected to depart initially using the lateral navigation mode of the flight management system (FMS). Once the aircraft had climbed through 500 ft AMSL and had past the departure end of the runway, the captain planned to instruct the first officer to select the automatic flight control system (AFCS)^[3] heading mode with the ATC assigned heading selected. The captain would then turn the aircraft to the heading assigned by ATC.

At 1344, the flight crew taxied the aircraft to runway 33 and reported being ready for departure. At 1345:07, ATC provided a take-off clearance to the flight crew with an assigned heading of 335°M. This was just a five degree right turn from the runway heading of 330°M. The captain selected the heading bug to 335 and entered the assigned heading using the scratch pad function of the control display unit of the FMS for later reference.

The take-off was conducted normally and as the aircraft climbed through 500 ft, the FMS, in lateral navigation mode, directed a right turn past the assigned heading. The captain followed the FMS direction and turned the aircraft right. As the aircraft turned to 335°M the FMS continued to command a right turn. The captain followed the FMS direction and continued the turn past the assigned heading (Figure 2).

Figure 2: Overview of departure

An overview of the initial departure showing both the assigned flight path and actual flown flight path. Source: Google earth, annotated by ATSB

As the aircraft climbed through about 700 ft, the captain instructed the first officer to select the heading mode of the AFCS with the assigned heading of 335°M selected. Recorded flight data showed that by this time, the aircraft had turned to a heading of 013°M. The AFCS then directed a left turn toward the assigned heading of 335°M. Both flight crew members immediately identified that they had turned past the assigned heading. The captain then began a left turn back to heading 335°M. At about the same time, ATC contacted the flight crew to confirm the assigned heading of 335°M. The flight crew responded confirming the assigned heading and continued the left turn to this heading.

At 1346:24, ATC contacted the flight crew again to confirm that operations were normal, and the flight crew confirmed that they were. ATC then cleared the flight to continue the departure visually on heading 335°M, without reference to the SID.

The flight continued to Lihir Island without further incident. No persons were injured, and the aircraft was not damaged during the incident.

Flight management system

The lateral navigation mode of the FMS provided flight path guidance along a series of pre-programmed waypoints. The aircraft manufacturer advised that when the FMS was programmed with the CAIRNS TWO SID from runway 33 and operated in lateral navigation mode, the system was programmed to maintain the runway track of 330°M until the aircraft climbed to 500 ft. For further track guidance, the continuing flight path needed to be managed by the flight crew.

In the absence of a manually programmed flight path, as in this incident, the SID navigation database within the FMS was programmed to command a turn to a heading of 025°M. This turn was presented to the flight crew on the aircraft’s navigation display prior to departure.

The heading mode of the AFCS provided means for the flight crew to manually select a desired, or ATC instructed, heading. Once activated, the heading mode provided flight director guidance to turn the aircraft to the selected heading.

Operator SID training

Cairns was the only airport in the operator’s network at which SIDs were used.^[4]

The company conducted regular simulator training for flight crews at six monthly intervals. The training included departures involving SIDs. However, this training used SIDs that required navigating along a path of pre-programmed waypoints. In these cases, the FMS provided automated guidance.

The training did not include SIDs that involved an assigned heading component.

Captain comments

The captain provided the following comments:

• The captain had not previously encountered a SID with an assigned heading component.
• In future he will use the heading mode of the FMS for SIDs with an assigned heading component.

First officer comments

The first officer provided the following comments:

• The flight crew were not expecting a 5-degree heading change. When assigned the heading, they did not recognise that the heading change was so minor.
• The first officer had previously flown ATR72-212A aircraft for an Australian domestic operator and had flown SIDs extensively. In that role, she always conducted SIDs in heading mode.

Safety analysis

When preparing for the departure, the flight crew elected to use the lateral navigation mode of the FMS to provide flight path guidance instead of the more appropriate heading mode.

After take-off, the FMS directed a turn past the assigned heading of 335°M to a heading of 025°M. The flight crew initially followed this guidance, incorrectly turning the aircraft to a heading of 013°M before selecting heading mode and detecting the error. At about the same time, ATC observed the aircraft turning beyond the assigned heading and contacted the flight crew. The error was then managed by both ATC and the flight crew and the flight continued without further incident.

Findings

These findings should not be read as apportioning blame or liability to any particular organisation or individual.

• The selected FMS mode was inappropriate for the assigned departure and provided flight path guidance not aligned with the SID. The flight crew followed the guidance provided by the FMS before the error was identified and corrected.

Safety action

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

Aircraft operator

As a result of this occurrence, the aircraft operator has advised the ATSB that they are taking the following safety actions:

Flight crew education and training

• Aircrew notices were circulated to all flight crew, providing education on SIDs and operations at Cairns.
• The simulator training program has been changed to include a greater focus on SIDs and Cairns operations.

Aircraft manufacturer

As a result of this occurrence, the aircraft manufacturer, in coordination with the navigation database provider, has taken the following safety action:

Change to SID navigation database

• To avoid a sharp right turn during a take-off using an incorrect AFCS mode, the runway 33 CAIRNS TWO SID navigation database commanded heading has been changed from 025°M to 335°M.

Safety message

This incident highlights the importance of effective use of flight management systems. Modern flight management systems, when used appropriately, can greatly reduce flight crew workload and provide increased levels of safety and efficiency. However, to effectively manage a modern aircraft, flight crews need to ensure they have a thorough understanding of the relevant flight management systems.

Also underlined is the need to effectively cross check flight instrumentation indications to ensure that the aircraft follows the intended flight path. The United States Federal Aviation Administration publication: Advanced Avionics Handbook, Chapter four Automated Flight Control provides the following guidance for effective use of the flight director:

The convenience of flight director cues can invite fixation or over reliance on the part of the pilot. As with all automated systems, you must remain aware of the overall situation. Never assume that flight director cues are following a route or course that is free from error. Rather, be sure to include navigation instruments and sources in your scan. Remember, the equipment will usually perform exactly as programmed. Always compare the displays to ensure that all indications agree. If in doubt, fly the aircraft to remain on cleared track and altitude, and reduce automation to as minimal as possible during the problem processing period. The first priority for a pilot always is to fly the aircraft.

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

__________

1. Lihir Island Airport is also known as Londolovit or Kunaye.
2. Pilot Flying (PF) and Pilot Monitoring (PM): procedurally assigned roles with specifically assigned duties at specific stages of a flight. The PF does most of the flying, except in defined circumstances; such as planning for descent, approach and landing. The PM carries out support duties and monitors the PF’s actions and the aircraft’s flight path.
3. The automatic flight control system uses information provided by the flight crew and the FMS along with other aircraft sensors and instrumentation systems to provide autopilot and flight director functions.
4. The operator’s home airport, Port Moresby International Airport, also provided SIDs. However, the operator and both flight crew reported that these were not used operationally.

On 2 May 2017, the pilot of a Messerschmitt-Bölkow-Blohm BK117-A3 helicopter, registered ZK‑IED, was tasked to relocate power poles from the northern side of the inlet to the southern side at Porirua Harbour, Pauatahanui Arm, New Zealand. The pilot reported that while in the cruise at about 300 ft, the helicopter started an un-commanded yaw, at which point the pilot released the sling load and entered into an autorotation onto the water. The helicopter was substantially damaged, and the pilot was uninjured.

On 3 July 2017, the New Zealand Transport Accident Investigation Commission (TAIC) requested Australian Transport Safety Bureau (ATSB) technical assistance with data recovery from a global positioning system (GPS) device that was onboard the helicopter at the time of the ditching. In accordance with paragraph 5.23 of Annex 13 to the Convention on International Civil Aviation Aircraft Accident and Incident Investigation, the ATSB appointed an accredited representative (ATSB investigator) to the TAIC investigation. To facilitate this request, the ATSB initiated an external investigation under the provisions of the Transport Safety Investigation Act 2003.

The ATSB successfully recovered the data from the GPS and provided a copy to the TAIC on 4 September 2017.

TAIC is responsible for the release of the final investigation report into this accident. Any enquiries in respect of the ongoing TAIC investigation or release of the investigation report should, in the first instance, be directed to the:

Transport Accident Investigation Commission
Level 9, 114 The Terrace
PO Box 10323
Wellington, 6143, New Zealand

_______________
The information contained in this web update is released in accordance with section 25 of the Transport Safety Investigation Act 2003 and is derived from the initial investigation of the occurrence. Readers are cautioned that new evidence will become available as the investigation progresses that will enhance the ATSB's understanding of the accident as outlined in this web update. As such, no analysis or findings are included in this update.

Safety summary

What happened

On the evening of 8 September 2017, an instructor and student from the Australian International Aviation College were preparing to conduct night training circuits at Port Macquarie Airport, New South Wales, in a Diamond DA40 NG aircraft, registered VH-YPQ. As it was the student’s first time conducting night circuits, once lined up on the runway, the instructor took the aircraft controls for the take-off.

After take-off, as the aircraft climbed, the instructor heard and felt the engine and propeller surging. Propeller speed and engine power fluctuations occurred from about 200 ft above the runway and increased in amplitude as the aircraft climbed to about 400 ft. Recorded data showed that the engine was producing full power despite the fluctuations. The instructor interpreted the fluctuations as a partial engine power loss and commenced a left turn, aiming to return and land on the runway in the opposite direction to the take-off. The instructor had considered landing straight ahead but assessed that there was power available to turn and that they would be unable to see and avoid trees or to be sure to land in a suitable clearing ahead.

In the 10 seconds that the instructor was assessing and making decisions about a perceived partial power loss, the airspeed reduced from 75 to 69 kt due to the aircraft’s nose-up pitch attitude. Then, at the same time as commencing the turn back towards the runway, the instructor reduced engine power to 30 per cent, while maintaining a nose-up attitude, and the airspeed reduced rapidly.

During the turn, the aircraft aerodynamically stalled resulting in a loss of control. Although the aircraft pitched down and the instructor subsequently increased the power, control was not regained. The aircraft descended and collided with trees, coming to rest inverted. The student and instructor were seriously injured, and the aircraft was destroyed.

What the ATSB found

After reducing the power, the instructor did not maintain adequate airspeed during the turn. This resulted in an aerodynamic stall, loss of control and collision with terrain.

The aircraft manufacturer could not determine the reason for the engine speed fluctuations. Propeller speed fluctuations had occurred in other aircraft, and either resolved without pilot input or by moving the power lever.

Although not contributing to this occurrence, in the course of the investigation it was found that engine cylinder heads for the aircraft type were cracking prior to reaching their service life.

What's been done as a result

After the accident, the Australian International Aviation College:

• added a requirement to the take-off safety briefing to include setting partial power safety speed in the event of a partial power loss
• conducted partial engine failure after take-off training for instructors and students, comprising pre-flight planning and self-briefing, ground training, and flight training
• performed flight simulator tests for partial engine failure after take-off conditions in each single engine aircraft model operated by the flying school to assess the power required to maintain altitude in the event of a partial power loss.

Safety message

In this accident, the instructor perceived there was a partial power loss. The ATSB research report Avoidable Accidents No. 3 – Managing partial power loss after take-off in single-engine aircraft provides information to assist pilots maintain aircraft control in the event of an emergency or abnormal situation after take-off. The report prescribed initial actions to be considered including:

• Lower the nose to maintain the glide speed of the aircraft. If turning is conducted, keep in mind an increased bank angle will increase the stall speed of the aircraft.
• Maintain glide speed and assess whether the aircraft is maintaining, gaining or losing height to gauge current aircraft performance.
• Fly the aircraft to make a landing, given the aircraft’s height and performance, and the pre-planned routes for the scenario.

The occurrence

On the evening of 8 September 2017, an instructor and student from the Australian International Aviation College (AIAC) planned to conduct night circuits at Port Macquarie Airport, New South Wales, in a Diamond DA40 NG aircraft, registered VH-YPQ. The student taxied the aircraft to runway 21, then, as it would be the student’s first time conducting circuits at night, the instructor took over the controls for the take-off.

It was dark and the moon had just risen above the horizon when the aircraft commenced the take-off roll at 1957:12 Eastern Standard Time^[1] (Figure 1). During the initial climb, about 20 seconds after lift-off, while still above the runway and passing 200 ft, the instructor noticed changes in the engine sound, felt acceleration changes and saw fluctuating indications on engine load and propeller RPM gauges, despite maintaining the power lever in the fully forward maximum power position.

The fluctuations increased over the next 20 seconds as the aircraft climbed. During that time, the aircraft pitched up to about 8 degrees, consistent with the normal attitude for initial climb and to achieve the best rate of climb speed, and the vertical speed increased to more than 900 ft per minute. The instructor assessed that the fluctuations were due to an engine problem and considered the options for landing.

The instructor considered landing straight ahead, however, while there were two fields that may have been suitable for a landing, they were not visible at night and the instructor was concerned the aircraft may land beyond the field in trees. While still above the runway and climbing through about 400 ft, and as the engine was still producing power, the instructor elected to conduct a left turn, aiming to land on the reciprocal runway, 03. The left turn commenced at 1957:58 and 1 second later, the instructor moved the power lever aft, reducing the engine load to about 30 per cent, where it remained for the next 10 seconds.

At 1958:01, the instructor broadcast on the common traffic advisory frequency^[2] that they had engine problems and would land on ‘runway 21,’ although actually intending to land on runway 03.

The instructor recalled concentrating on trying to maintain adequate speed—not ‘nosing up too much getting closer to the stall speed or nosing down and not being able to make the runway.’ As the aircraft turned and the runway came into sight, the instructor assessed that the aircraft was not going to make it back as the runway was too far away and the airspeed was too slow. The instructor also recalled looking down and all that could be seen was the trees.

The instructor made changes with the power lever to see if the fluctuation issue improved. Engine data showed the power lever position moving to maximum power for 3 seconds, back to 25 per cent engine load for less than 2 seconds then to full power for the final 4 seconds. The instructor also noticed signs of an impending aerodynamic stall^[3]—buffeting and sloppy controls, followed by a left wing drop. The aircraft entered uncontrolled flight, descended rapidly and impacted trees and terrain.

The aircraft collided with trees about 18 seconds after the left turn commenced, 325 m abeam the runway 03 threshold, and came to rest inverted (Figure 1). Both occupants were seriously injured and the aircraft was destroyed.

Figure 1: Aircraft flight path and key events

Source: Google Earth, annotated by the ATSB based GPS, engine control unit and radio recordings, and accident site assessment

__________

1. Eastern Standard Time (EST): Coordinated Universal Time (UTC) + 10 hours.
2. Common Traffic Advisory Frequency (CTAF): A designated frequency on which pilots make positional broadcasts when operating in the vicinity of a non-controlled aerodrome or within a Broadcast Area.
3. Aerodynamic stall: occurs when airflow separates from the wing’s upper surface and becomes turbulent. A stall occurs at high angles of attack, typically 16˚ to 18˚, and results in reduced lift.

Context

Operational information

Aircraft operator

The Australian International Aviation College (AIAC) was a flight training and charter organisation based in Port Macquarie, New South Wales. It conducted flight training for up to 90 Australian and international students, from initial training through to commercial pilot licence, and single and multi-engine aeroplane, instructor and instrument flight ratings. The AIAC operated a fleet of Diamond aircraft and an approved flight training device (simulator).

Instructor qualifications and experience

The instructor:

• held a commercial aeroplane pilot licence issued in April 2014
• held a valid Class 1 medical certificate issued in March 2017 with the restriction that distance vision correction must be worn
• held instrument and instructor ratings
• met the recency requirements for the planned night circuits
• satisfactorily completed flight crew emergency procedures training at AIAC in February 2017
• completed a flight instructor standardisation and proficiency check in January 2017 including attaining a Grade 2 instructor rating
• obtained an instrument rating training endorsement for single-engine aircraft below 5,700 kg in June 2017
• had accrued 1,160.9 hours total aeronautical experience, including 86.2 hours in command at night (73.7 in single-engine aircraft and 12.5 in multi-engine aircraft) and 680.7 hours instructing in single-engine aircraft
• logged 124 hours in the DA40 NG and DA42 NG (diesel-engine) aircraft in 2017.

Student pilot experience

The student pilot commenced flight training at AIAC in March 2017 and completed a recreational pilot licence test on 24 May 2017. The student had accrued a total of 85.4 hours flying time, of which 19.8 were as pilot in command.

Weather and environmental information

The weather was fine with light winds and little to no cloud. The time of the occurrence was past astronomical twilight and the nearly full moon was just above the horizon. The township of Port Macquarie provided some light and a horizon reference to the east but the ground near the airport was dark apart from a few scattered dwellings. The instructor reported that the moon and town lights were below their field of vision during the initial stage of the flight.

The Bureau of Meteorology recorded the wind strength and direction at Port Macquarie Airport at 1-minute intervals. Nil wind was recorded at the aerodrome for the 20-minute period encompassing the short duration of the accident flight. Although the aircraft took off with nil wind at runway level, according to GPS data, it encountered an increasing tailwind of up to 8 kt during the climb and turn. As the aircraft climbed, the temperature increased from 12.5 °C at runway level to 16.5 °C at the maximum height reached.

Aircraft information

General information

The Diamond Aircraft Industries DA40 NG aircraft is a four-seat, low-wing, T-tail aircraft constructed from composite materials. The aircraft is factory-fitted with a turbo charged four-cylinder diesel Austro E4 engine, operated on aviation turbine fuel, and a three-bladed wooden composite variable-pitch MT-Propeller MTV-6-R propeller.

VH-YPQ

The aircraft serial number 40.N292 was manufactured in 2015 and placed on the Australian register in early 2016 as VH-YPQ (Figure 2).

The aircraft had a current certificate of airworthiness and maintenance release^[4] with no outstanding maintenance or defects notated. The aircraft had a total time in service of about 906 flight hours before the accident flight. The logbook statement indicated that the aircraft was maintained in accordance with the Diamond Aircraft DA40 NG maintenance schedule. It was equipped and certified for operation under the instrument flight rules^[5] and was maintained and operated in the Charter Class B category.

Figure 2: VH-YPQ

Source: Simon Coats

System description and information

Engine and propeller control

The constant speed propeller has a governor, which changes the blade pitch to maintain a constant RPM, regardless of the amount of engine torque, airspeed or altitude. The aircraft’s maximum take-off propeller speed was 2,300 RPM.

The engine and propeller were controlled by a single electronic engine control unit (ECU) with dual-redundant hardware that performed continual self-testing. A single power lever provided command input via two separate channels to the ECU and the ECU controlled the engine fuel injection to match the power lever command. For redundancy, there was a cockpit switch for selecting ECU channels in case of a fault.

Integrated instrument and avionics system

VH-YPQ was factory-fitted with a Garmin G1000 integrated avionics system, which consolidated all communication, navigation, surveillance, automatic flight control system, primary flight instrumentation, engine indication, and annunciation systems on two liquid crystal display units (DU) and an audio panel. The two DUs consisted of a primary flight display on the left (student pilot side), and the multi-function display (MFD) on the right (instructor side). The audio panel was located between the two display units. The aircraft was not fitted with the optional terrain awareness and warning system.

Recording capability

The G1000 avionics system was capable of storing up to 60 flight and engine parameters on a data memory card. Data was logged to a new file that was created each time the MFD was powered on. All parameters were recorded at 1-second intervals. The electronic ECU was capable of storing significant amount of engine data and had a backup battery for redundancy.

Aircraft weight and balance

The aircraft was assessed as being within the weight and balance limits throughout the accident flight, with a take-off weight of 1,143 kg.

Relevant speeds

Best glide speed is used to achieve the greatest distance for the height in case of engine failure. At any airspeed faster or slower than the best glide speed, the aircraft will travel less distance over the ground.

According to the Airplane Flight Manual, the best glide speed (‘airspeed for best glide angle’) for the aircraft with flaps up was 88 kt, and ‘airspeed for emergency landing’ with engine off and flaps in the take-off position was 78 kt.

The best rate of climb speed (VY) was 72 kt with take-off flaps set. Therefore, in the event of a power loss when in the climb at VY, the pilot would need to lower the aircraft nose to achieve the best glide speed, and the airspeed for emergency landing.

Accident site examination

The aircraft initially impacted trees causing significant structural damage to the wings, after which it impacted terrain inverted and at a nose-down angle of about 25°. The cockpit was partially collapsed, with the airframe resting on the instrument panel and seat backs. Both wings and the tail detached during the impact sequence (Figure 3).  

Figure 3: Wreckage of VH-YPQ, which was partially disturbed during rescue of the occupants

Source: ATSB

Examination of the aircraft wreckage found no pre-existing airframe issues. The flap actuator was found in the take-off position. The aircraft’s fuel tanks were breached and there was evidence of fuel spillage. A small quantity of fuel was drained from the tanks and matched the characteristics of aviation turbine fuel. It tested negative to water content. Previous fuel records and the engine data both indicated that there was sufficient fuel on board the aircraft for the intended flight.

The engine was examined externally with the cowls removed and there was no noted fluid leakage or pre-impact defects identified. During removal of the engine and control parts, no defects were identified. The propeller blades were fragmented, consistent with the engine driving the propeller when the aircraft impacted with terrain. The propeller and propeller control unit were removed from the aircraft for function testing.

The G1000 avionics unit data memory card and electronic ECU memory module were removed from the aircraft and transported to the ATSB technical facilities for examination and download.

Component testing

Propeller hub and propeller control

The propeller hub and controller were sent to the propeller manufacturer for testing. No faults were identified during function testing or visual examination.

Manufacturer data analysis

The aircraft and engine manufacturers analysed the recorded data and could not determine the cause of the propeller speed fluctuations. The engine manufacturer reported that the oscillation was still in a normal range and did neither create an ECU warning nor a reduction of power. There are lots of possible reasons for RPM oscillation but in this case they were not engine related.

They advised that at the propeller speeds that were recorded during the last 40 seconds before the accident, the propeller should have been sitting on the fine pitch limit stops. Whether the propeller was at full fine pitch during the fluctuations could not be determined as the actual propeller blade pitch angle was not sensed or recorded. Fuel injection quantity and timing were not recorded on the ECU, but fuel quantity and fuel flow were recorded on the Garmin system. In the recorded Garmin data, the fuel flow rate increased and decreased in step with the propeller RPM and in response to the changes in power lever position.

Assessment of possible causes of RPM fluctuations

Figure 4 is an extract from the aircraft maintenance manual, which provided troubleshooting for fluctuating propeller RPM from a list of possible causes.

Figure 4: Extract from the aircraft maintenance manual

Source: Diamond Aircraft

From the listed possible causes, the post-accident inspection found the following.

• The engine gearbox oil level was not measured, but the ATSB assessed its quantity as unremarkable. There was no sign of an oil leak on or around the engine that would have depleted the oil quantity.
• No defects or metal debris were identified in the gearbox oil filter or magnetic chip detector.
• No faults were identified with the electrical wiring harness.
• The propeller governor was satisfactorily tested by the propeller manufacturer and all measurements and adjustments were found to be within tolerances.

Recorded data

Recorded data from the G1000 memory card and ECU memory module was analysed. The data from the two sources correlated to within about 2 seconds. Selected data is displayed in the graph in Figure 5. In this graph, the altitude, airspeed, ground speed, pitch and roll were retrieved from the G1000 and the propeller speed (RPM), power lever position and engine load were from the ECU. The altitude is accurate to within 30 ft.

Propeller speed fluctuations

According to the data, the aircraft lifted off at 1957:26. Twenty seconds later, the aircraft climbed through 200 ft at an airspeed of 74 kt. The propeller RPM began fluctuating about 10 RPM per second above and below 2,250 RPM. The power lever remained constant in the maximum power position and the engine load made small fluctuations around 98.5 per cent power, consistent with the RPM fluctuations.

The RPM and engine load fluctuations increased in amplitude as the aircraft climbed. The engine manufacturer advised that fluctuations up to ± 20 RPM were acceptable, although no normal operating range of fluctuations was defined. That ‘acceptable’ change in amplitude was first exceeded at 1957:47, 218 ft above the runway.

The fluctuations continued to increase in amplitude over the next 10 seconds to a maximum variation of 73 RPM in 1 second (which equated to about 3 per cent of an average 2,250 RPM) as the aircraft climbed to 361 ft.

Figure 5: Plot of selected engine and aircraft GPS data parameters

Image shows increasing fluctuations in propeller RPM and engine load, followed by a sequence of power lever movements with corresponding RPM and load variations.

Source: Austro Engine and Garmin 1000 data analysed by ATSB

Manoeuvring

As the aircraft climbed through 300 ft above the runway, the airspeed reached 75 kt, which was the maximum achieved on the flight. Over the next 10 seconds the airspeed decreased to 69 kt as the pilot maintained a pitch-up attitude of about 8 degrees, and the aircraft climbed at a rate of up to 938 ft per minute. After reaching a peak pitch-up of 8.3 degrees and still climbing at 928 ft per minute, the pilot started to reduce the aircraft’s pitch attitude.

At 1957:58, as the aircraft was climbing at 904 ft per minute through 389 ft, the airspeed had reduced to 69 kt and a left bank (roll) commenced. At that time, the aircraft was still pitched up about 7 degrees. One second later, the power lever was moved to a lower power setting that corresponded with an engine load reduction to about 30 per cent and the propeller speed decreased to about 1,750 RPM.

Over the next 4 seconds, the aircraft continued to climb albeit at a reducing rate, the airspeed continued to decrease, and the angle of bank increased. At 1958:05, the maximum altitude of about 430 ft was reached, airspeed was 59 kt and the aircraft nose then pitched down. The aircraft then started to descend. The last Garmin data recorded was at 1958:07 with the airspeed at 59 kt (which it had been for 4 seconds), a 29-degree angle of bank and a propeller speed of 1,720 RPM.

The ECU data indicated that at 1958:09, the power lever was moved forward for 3 seconds, back for 2 seconds then forward for the final 4 seconds and the engine load increased and decreased correspondingly. The last recorded ECU data at 1958:17 was consistent with the aircraft colliding with terrain.

Uncontrolled descent

The G1000 memory card was missing approximately the last 12 seconds of data. This was likely due to the unit’s power being cut abnormally during the accident sequence, instead of a normal system shutdown. Consequently, the aircraft’s final descent and flight path were not recorded. Based on alignment between the accident site location, direction of travel and the final recorded position, the left turn and steep descent continued until impact. The aircraft descended from 428 ft in 10 seconds, which was an average descent rate of 2,568 ft per minute.

The stalling speed with power off (VS) for the aircraft weight and take-off flap was 58 kt indicated airspeed (KIAS) at 0° angle of bank, and 62 KIAS at 30° angle of bank. The last recorded airspeed was 59 KIAS when the bank angle was about 30°, below the power-off stalling speed.

No ECU faults were recorded.

Aircraft performance testing

Following the accident, the AIAC conducted in-flight performance testing of partial engine failure scenarios. With an aircraft weight of 1,156 kg (similar to the accident flight), a DA40 NG aircraft maintained level performance (occasional 100 ft per minute climb) and airspeed at a safe margin above the stalling speed, with 30 per cent power, flaps in the take-off position and glide attitude set, including in a turn.

Regulatory requirements

The Civil Aviation Safety Regulations (CASR) Part 61 Manual of Standards (MOS) detailed competency standards for all flight crew qualifications as well as proficiency checks, flight reviews and flight test standards.

Competencies specified for private and commercial aeroplane pilot licence and single-engine aeroplane class ratings included managing a simulated engine failure after take-off in the circuit area, but not partial power loss. Underpinning knowledge of the competencies included ‘engine failure scenarios and procedures for partial and complete power loss.’

Competencies specified for the night visual flight rules (NVFR) rating included skills and knowledge required for managing emergency situations at night including that ‘(in simulated conditions) aircraft control is maintained.’ The Airservices Australia Aeronautical Information Publication En Route 1.1 did not permit simulated engine failures to be conducted below 1,500 ft at night in the circuit area.

The Part 61 MOS Unit A5 – Advanced stalling, included that a pilot is required to demonstrate recovery from a stall with full or partial loss of engine power. The Civil Aviation Safety Authority’s (CASA’s) Flight instructor manual stated:

Before carrying out any advanced stalling exercise it is important that sufficient height is gained to ensure recovery by 3,000 feet above ground level…

CASA advised that competency in this standard ‘should mitigate [against the] risk of partial power loss in the take-off.’

Operator procedures

Managing partial power loss after take-off

The AIAC syllabus of training was in line with the Part 61 MOS requirements and did not reference partial engine failure after take-off or in the circuit. However, the AIAC standard operating procedures included expanded procedures for engine failure immediately after take-off with insufficient/partial engine power. It listed the following considerations in managing a partial power loss after take-off:

• When faced with a partial power loss, the pilot in command should not try to diagnose the engine problem at the expense of maintaining aircraft control;
• Climbing at the aircraft’s best rate of climb speed will maximize options if a partial power loss or total power loss occurs;
• Lower the nose to maintain the glide speed of the aircraft (if unable to climb);
• If a partial power loss has occurred with remaining runway, the earlier a decision is made to cut remaining engine power, the greater landing distance is available; the immediate extension of landing flaps is recommended as this will also help to reduce the aircraft groundspeed prior to ground contact;
• Conduct the Phase One checks as outlined in the QRH. However, this should only be done if there is sufficient time;
• Maintain best glide speed and assess whether the aircraft is maintaining, gaining or losing height to determine current aircraft performance. This will assist in making decisions in the available options for landing.
• Fly the aircraft to make a landing, given the aircraft’s height and performance, and the pre-planned routes for the situation. If any turning is conducted, be mindful that an increased bank angle will increase the stall speed of the aircraft. Keeping the aircraft in balance will also minimize rate of descent in any turn.
• Re-assess landing options throughout any manoeuvers. Be decisive but be prepared to modify the plan if required.
• Maintain glide speed up to the point of landing flare; this will ensure that when flaring there is enough energy to arrest the vertical descent rate.

Circuits under the night visual flight rules

At the time of the accident, the AIAC NVFR training syllabus included 5 hours of night circuits. Prior to commencing night circuits, students were given a briefing by an instructor. The briefing included human factors, illusions that can occur at night, lighting and flight with reference to instruments. It did not include emergencies, but students received a briefing on circuit emergencies earlier in their training. Additionally, at night, company pilots were required to conduct the climb after take-off at the best rate of climb speed.

Quick reference handbook

Emergency and non-normal procedures were detailed in the AIAC Quick Reference Handbook (QRH). These were derived from the emergency procedures detailed in the Airplane Flight Manual. The QRH included ‘Phase 1 checks’ that were printed in bold type. The Phase 1 checks were ‘recall items’ required to be committed to memory. The other, ‘Phase 2’ checks were not memory recall items and the QRH checklist was required to be used when carrying out these checks.

Two QRH procedures were relevant to the accident flight: the checklist for partial or full engine failure after take-off and the ‘defective propeller RPM regulating system’ checklist. The first of these is depicted in Figure 6. Nearly all the items were in bold and were therefore memory recall items. The first (memory recall) item was to achieve an airspeed of 88 kt with flaps up, or 77 kt with flaps in the take-off position. The second item, if time permitted, was to check the power lever was at maximum.

Figure 6: QRH checklist for partial or full engine failure after take-off

Source: AIAC

In this occurrence, the instructor assessed the speed fluctuations as symptomatic of a partial power loss, although the engine was still producing maximum power. The instructor reported confirming the power lever was in the full power position.

The engine manufacturer reported that the defective propeller RPM regulating system procedure had been in the airplane flight manual of the DA40 D aircraft, which had a different engine, since 2003. The procedure was introduced due to issues with the propeller control system of that engine type, but was still a valid procedure to be followed in case of oscillating RPM in the DA40 NG aircraft and in the DA40 NG flight manual.

The QRH defective propeller RPM regulating system checklist carried the following warning:

IN CASE OF DEFECTIVE RPM REGULATING SYSTEM, REDUCED ENGINE PERFORMANCE SHOULD BE ANTICIPATED.

For oscillating RPM, the first action, was:

POWER setting…………..CHANGE

If problem does not clear;

VOTER Switch……………Swap between ECU A and ECU B

If problem does not clear;

VOTER Switch……………AUTO

Land at nearest suitable airport

Revise ENGINE FAILURE DURING FLIGHT checklist (Page-322).

END OF CHECKLIST

The instruction to change the power setting was not a memory recall item. The AIAC Operations manual section QRH procedures stated that ‘the QRH should not be referred to unless flight crew workload is sufficiently low to operate the aircraft while the checks are conducted,’ and that, in visual meteorological conditions, the pilot was not to refer to the QRH (for non-recall items), unless the aircraft was at or above minimum/lowest safe altitude, and/or other safe conditions could be met.

The aircraft manufacturer advised that changing the power lever angle in accordance with the procedure should fix a problem such as that encountered in the accident occurrence, because ‘changing the operating state of the engine will typically stop periodic interactions of systems and thus stop RPM fluctuations.’

The AIAC head of operations considered that the engine fluctuations were ‘very small’ and should have been able to be remedied by changing power setting/moving the power lever slowly back and forth (then changing the ECU switches if necessary), in accordance with the QRH checklist. However, the instructor had not been aware of similar issues within the operator’s aircraft fleet before the accident and reported that similar events had not been discussed or trained for by the operator to a point of being memorable.

Decision making

Pilots operate in a safety-critical environment and need to be trained and supported to make the best possible decisions in challenging conditions. Orasanu (2010)^[6] stated that

in many high-risk consequential environments, time for making a decision is limited, information is incomplete, conditions change dynamically, and goals shift, rendering analytic decision-making impractical, if not impossible.

Orasanu detailed ways in which expert knowledge contributes to cockpit decision-making. These included quick and accurate interpretation of a problem and performing rehearsed responses. Where a pilot made an error or did not select the best solution, it may be due to an incorrect interpretation of the situation, or choosing an inappropriate course of action.

The instructor had not previously experienced propeller speed fluctuations and interpreted the problem as a partial power loss. The ATSB Avoidable Accidents No. 3 publication, Managing partial power loss after take-off in single-engine aircraft, stated that while following a complete engine failure, a forced landing was inevitable, a partial power loss required the pilot to make a decision whether to continue flight or land immediately. Research for the publication found that in 145 of 160 occurrences where a pilot turned back to the runway following a partial power loss, the aircraft made it to within the aerodrome grounds. However, the increase in stall speed during the turn and the associated potential for a loss of control meant that the consequences of a mishandled turn back were more serious than a controlled forced landing.

The publication suggested four main considerations when assessing if a turn back to the aerodrome is possible. These were:

• height available

• remaining engine power available – do you have enough power to climb?

• increased stall speed associated with any increase in angle of bank increasing the risk of an aerodynamic stall

• level of confidence in the remaining engine power – but assume the engine may fail at any moment.

Take-off safety briefing

The ATSB Avoidable Accidents publication also stated that pilots should self-brief prior to each and every take-off. The take-off brief ‘serves as a reminder of your planned actions in the event of an emergency such as a partial power loss.’ Planning actions under non-stressful and controlled circumstances prepares pilots for a quick response, reducing mental workload and mitigating some effects of decision making under stress, such as reduced short-term memory, if an emergency situation does eventuate.

The AIAC operations manual specified that a take-off safety briefing must be given by the pilot flying^[7] at the completion of pre-take-off checks.

The company take-off safety briefing was:

1. Any emergencies prior to VR^[8] I will reject the take-off.

2. Any emergencies at or after VR I will reject and land on remaining runway or clearway.

3. Any emergencies airborne with no runway remaining I will pick a landing area 30° either side of the nose and conduct a forced landing.

4. Special considerations – unique airport information, weather conditions, terrain or obstacles on departure, other known risks and intentions.

The take-off safety briefing did not specify what events constituted an emergency.

The instructor recited their normal take-off safety briefing as:

If there was anything on the ground, power to idle, brakes to come to a stop. Airborne with runway remaining, nose down land on the remaining runway. Airborne with insufficient runway remaining, pick an area either side of the nose, nose down, maintain glide speed, consider flap and then shutdown checks.

Related occurrences

Aircraft operator fleet events

After the accident involving VH-YPQ, the pilot of a twin-engine Diamond DA42 aircraft with a similar engine type to VH-YPQ and operated by AIAC, experienced a similar propeller RPM fluctuation event, where it varied about 50 RPM. In that event, the wind was gusting to 25 kt. During the take-off, there were some RPM changes due to wind gusting, but the pilot reported that was normal. Passing about 400 ft during the climb, fluctuation noise caught the pilot’s attention. The pilot reduced the power to 85 per cent and after 5 to 6 seconds returned it to full power, and the fluctuations, noise and vibration ceased. The reasons for that fluctuation could not be determined from post-event examinations.

The aircraft maintainer provided extracted data of five similar fluctuation events from the aircraft operator’s fleet. The last 3 hours of engine operating data were routinely downloaded from the ECU at each 100-hour maintenance event. Within that data, which represented 3 per cent of an aircraft’s operating time, the maintainer found that oscillations/fluctuations in RPM were present ‘quite a lot of the time.’ Of the five sample events provided to the ATSB, four had fluctuations of greater magnitude than the accident flight. Only one of the five sample events was reported to the maintainer by the flight crew, and they reported ‘hunting’ of power and propeller RPM and that the aircraft was operating in turbulent conditions at the time.

The aircraft maintainer advised that the level of oscillation recorded in the ECU data for the accident flight was ‘quite common’ for the aircraft type (in the AIAC fleet) and believed they were generally introduced by an outside influence such as slight turbulence. The aircraft operator also reported that it was ‘common to observe and hear changes in propeller pitch on DA40 NG aircraft when flying in gusting wind conditions.’

The aircraft manufacturer advised that they were not aware of RPM fluctuations caused by windshear, gusts or turbulence, and that these would normally be eliminated by the governor. The AIAC chief engineer reported that in other previous occurrences of propeller fluctuations involving AIAC aircraft, the fluctuations had reached a maximum amplitude then reduced to near zero without pilot input.

Other aircraft events

In addition to the two events involving AIAC aircraft, the engine manufacturer reported that it was aware of 16 other events on aircraft fitted with the AE300 engine worldwide, involving fluctuations in propeller speed and load, none of which resulted in an accident. The engine manufacturer reported that in all cases, fluctuations ceased with a change in power lever position. Fifteen of these 16 occurrences were found to be the result of either a fuel pressure issue, propeller imbalance, or faulty electrical connection in the propeller governor and the cause of one was not found.

One additional event occurred in 2019, which was traced to excessive play in the alternator rotor. This was fixed by replacement of the alternator.

The reason for the fluctuations was found in all but three of the total of 19 reported occurrences.

Cylinder head cracking

On 25 May 2018, another DA40 NG aircraft in the AIAC fleet, VH-YPJ, had an occurrence where the engine coolant system over-pressurised and the coolant leaked. Inspection of that aircraft found cracking in six of the eight pairs of cylinder valve openings on the cylinder head.

Inspections were then carried out on the remaining aircraft in the operator’s fleet of five DA40 NG aircraft. The findings of the inspections are summarised in Table 1. The time between overhaul for the cylinder head was 1,800 hours.

Table 1: Cylinder head cracking in AIAC DA40 NG fleet

Source: AIAC

After the accident and under the supervision of the ATSB, the operator examined the cylinder head fitted to VH-YPQ. Hairline cracks were identified in the cylinder head in the same location as VH-YPJ (Figure 7), however those cracks were assessed as insufficient in size to create engine problems at the time of the accident.

Figure 7: Cylinder cracks in VH-YPQ and another aircraft

Source: AIAC and ATSB

__________

4. Maintenance release: an official document, issued by an authorised person as described in Regulations, which is required to be carried on an aircraft as an ongoing record of its time in service (TIS) and airworthiness status. Subject to conditions, a maintenance release is valid for a set period, nominally 100 hours TIS or 12 months from issue.
5. Instrument flight rules (IFR): a set of regulations that permit the pilot to operate an aircraft to operate in instrument meteorological conditions (IMC), which have much lower weather minimums than visual flight rules (VFR). Procedures and training are significantly more complex as a pilot must demonstrate competency in IMC conditions while controlling the aircraft solely by reference to instruments. IFR-capable aircraft have greater equipment and maintenance requirements.
6. Orasanu, J., 2010, Flight Crew Decision-Making, in Kanki, B., Helmreich, R. and Anca, J., Crew Resource Management, Elsevier, San Diego, USA
7. Pilot Flying (PF) and Pilot Monitoring (PM): procedurally assigned roles with specifically assigned duties at specific stages of a flight. The PF does most of the flying, except in defined circumstances; such as planning for descent, approach and landing. The PM carries out support duties and monitors the PF’s actions and the aircraft’s flight path.
8. VR: Rotate speed, is the speed at which the pilot begins to apply control inputs to make the aircraft nose pitch up, after which it leaves the ground.

Safety analysis

Development of the accident

The aircraft experienced propeller speed fluctuations shortly after take-off, which although relatively small, the instructor had not experienced previously. The feel, sound and fluctuating engine indications were interpreted by the instructor as a partial power loss. The instructor considered landing ahead as for a complete engine failure. However, due to darkness, obstacles in the path, the aircraft position and the perceived power available, the instructor decided to turn back to the runway.

During the 10 seconds that the instructor was assessing and decision-making prior to commencing the turn, the airspeed decreased 6 knots due to the aircraft’s nose-up pitch attitude. At the same time as commencing the turn, the instructor reduced engine power, while maintaining a nose-up attitude.

Following the power reduction, the combination of nose-up pitch attitude and increasing angle of bank, resulted in an aerodynamic stall. The instructor recognised the signs of the impending stall and knew it was necessary to lower the aircraft nose to recover but did not do so due to the perceived proximity of the trees below. This resulted in the loss of aircraft control and collision with terrain.

In-flight performance testing of a DA40 NG aircraft demonstrated that 30 per cent power was sufficient to maintain level flight including in a turn. Therefore, in this occurrence, had adequate airspeed above the stalling speed been achieved and maintained by lowering the aircraft’s nose, it was likely aircraft control would have been retained, even following the power reduction commanded by the instructor.

Managing a perceived partial power loss

Although the engine did not sustain a partial power loss, the instructor perceived that there was an impending engine issue due to the noise and vibrations brought on by the propeller fluctuations. Therefore, the instructor managed the issue by dealing with it as a partial power loss. However, the instructor could not recall reducing the power at the same time as commencing the turn, or why.

Managing partial power loss after take-off in single-engine aircraft was not included in the units of competency within the Civil Aviation Safety Regulations Part 61 Manual of Standards (MOS) or the aircraft operator’s training syllabus, other than as ‘underpinning knowledge’. The instructor had previously demonstrated competence in managing simulated complete engine failure after take-off in single-engine and multi-engine aircraft and in managing simulated partial engine failure in multi-engine aircraft. The instructor was also required to have demonstrated competence in recovery from a stall with full or partial loss of engine power in accordance with the MOS.

Training for complete engine failure after take-off is straightforward and the trained response is primarily to lower the aircraft nose to achieve a safe airspeed and land ahead. Training for partial power loss is more complex as there can be significant variation in the power loss presentation. This includes anything from almost full to almost no power available, and the situation may resolve, worsen, or both. Following a partial power loss, a pilot will need to make an assessment of the power available, aircraft performance, suitable landing areas and other factors such obstacles and terrain. It may be that turning back to land on a runway is achievable. In any event, the first memory item stipulated in the aircraft’s quick reference handbook in case of full or partial engine power loss, was to achieve the stated (best glide) airspeed, which is essential to avoid a loss of control.

Research conducted for the ATSB publication Avoidable Accidents No. 3 – Managing partial power loss after take-off in single-engine aircraft, found proportionally more fatal accidents occurred following a partial power loss than a complete engine failure. In particular, loss of aircraft control occurred more often following a partial power loss, either during a turn back to the runway or as a result of pilot inaction to prevent airspeed decay. The consequences of these loss of control occurrences were more serious than landing ahead following a complete engine failure. The research identified that pilots generally used a take-off safety briefing that primed them for actions in case of complete engine failure after take-off but not for partial power loss.

The instructor’s take-off safety briefing was consistent with that used across the general aviation industry and included instructions (paraphrased):

In case of engine failure when airborne with no runway remaining, lower the nose to achieve best glide speed and land ahead.

As lowering the aircraft nose when close to the ground is counterintuitive, such priming may help the pilot resist pitching the nose up, thereby avoiding a stall and loss of control. However, in this occurrence, the instructor commenced the turn before lowering the aircraft nose to achieve a safe speed, and although knowing it was necessary to lower the nose, did not want to direct the aircraft towards the trees below.

In the Avoidable accidents publication, the ATSB assessed that including consideration of partial power loss in the take-off safety briefing by reminding the pilot to lower the nose to achieve best glide speed before assessing performance and decision-making, may improve outcomes following (actual or perceived) partial power loss after take-off.

Cylinder head cracking

While not contributing to this accident, engine cylinder head cracking in the aircraft operator’s fleet of DA40 NG aircraft had occurred before the service life of 1,800 hours was exceeded. A crack in an engine cylinder head may develop to a size that results in loss of compression or power, foreign object damage or engine failure.

Findings

From the evidence available, the following findings are made with respect to the collision with terrain involving Diamond DA40 aircraft, registered VH-YPQ, 1 km south of Port Macquarie Airport, New South Wales, on 8 September 2017. These findings should not be read as apportioning blame or liability to any particular organisation or individual.

Contributing factors

• In response to increasing propeller speed fluctuations which were interpreted as a partial power loss, the instructor decided to attempt to turn back to land at the aerodrome. The instructor reduced power to 30 per cent and did not maintain adequate airspeed during the turn, resulting in an aerodynamic stall, a loss of control and collision with terrain.

Other factors that increased risk

• The aircraft manufacturer could not determine the reason for the fluctuations. Propeller speed fluctuations have occurred in other aircraft with the E4 engine and MTV-6-R propeller, and either resolved without pilot input or by moving the power lever.
• Cylinder heads for the aircraft type were cracking prior to reaching their service life.

Safety issues and actions

Additional safety actions

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

Australian International Aviation College

The aircraft operator, Australian International Aviation College, reported to the ATSB that after the accident, they:

• added the following to their required take-off safety briefing:

In the event of a partial power loss, set partial power safety speed (DA40 NG – 78 kt take-off flap or 88 kt flap up), assess performance. If unable to maintain altitude find suitable place to conduct a forced landing.

• conducted partial engine failure after take-off training for instructors and students, comprising pre-flight planning and self-briefing, ground training, and flight training
• performed flight simulator tests for partial engine failure after take-off conditions in each single engine aircraft model operated by the flying school to assess the power required to maintain altitude.

Pilot details

Pilot details – Instructor

Pilot details – Student

Sources and submissions

Sources of information

The sources of information during the investigation included the:

• instructor
• student
• flight training school/aircraft operator
• aircraft and engine manufacturer
• Civil Aviation Safety Authority
• Airservices Australia
• Bureau of Meteorology.

References

Orasanu, J., 2010, Flight Crew Decision-Making, in Kanki, B., Helmreich, R. and Anca, J., Crew Resource Management, Elsevier, San Diego, USA

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 aircraft operator, manufacturer and maintainer, the engine and propeller manufacturers, instructor and student pilot, the Civil Aviation Safety Authority, Austrian Federal Safety Investigation Authority and the German Federal Bureau of Aircraft Accident Investigation.

Submissions were received from the aircraft manufacturer, operator and maintainer, the Civil Aviation Safety Authority, Austrian Federal Safety Investigation Authority and the instructor. 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 Published by: Australian Transport Safety Bureau © Commonwealth of Australia 2020 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 22 August 2017, the ATSB commenced a marine investigation into the grounding occurrence of a domestic commercial vessel (The Big Duck) in South Australia.

The initial notification indicated that the vessel broke its moorings while moored at Granite Island and drifted onto rocks in Victor Harbour. 

During the course of the investigation it was identified that, during the early hours of 15 August 2017, the unoccupied vessel broke its moorings as a weather front passed through the area. The ATSB has concluded that further investigation of this occurrence would yield minimal benefits to the enhancement of transport safety.

Therefore, the ATSB has discontinued the investigation in line with the provisions of the Transport Safety Investigation Act 2003 (TSI Act).

Section 21 (2) of the TSI Act empowers the Australian Transport Safety Bureau (ATSB) to discontinue an investigation into a transport safety matter at any time. Section 21 (3) of the TSI Act requires the ATSB to publish a statement setting out the reasons for discontinuing an investigation.

The ATSB commenced an investigation into a flight control system event that occurred on 25 May 2017 involving a Boeing Company 737-800, registered VH-YIJ and operated by Virgin Australia International, on a flight from Brisbane, Queensland to Wellington, New Zealand.

The captain was the pilot flying, and he was conducting the night arrival into Wellington. The weather conditions were fine, and the descent (below flight level 250) and approach were briefed to be flown with the autopilot and autothrottle disengaged for practice.

During the approach to runway 34, the flight crew progressively selected flaps 1 then flaps 15. The landing gear was selected down and then flaps 25. The right flap moved to flap 25 but the left flap initially remained at flap 15, before moving very slowly to flap 25, which was not initially detected by the flight crew. Flap 40 was then selected, however, the flaps remained in the 25 and 15+ positions.

While carrying out the landing checklist, the flight crew detected the flap asymmetry. The flight crew attempted to rectify the problem with various flap lever selections, which were unsuccessful. Approaching 1,000 ft, the aircraft did not meet the operator’s stable approach criteria, so the flight crew initiated a missed approach.

During the missed approach, the left flap slowly extended to flap 25, correcting the initial ‘Trailing Edge Flap Asymmetry’ to a ‘Trailing Edge Flap Disagree’ condition. The aircraft, still being manually flown, subsequently climbed above the cleared altitude of 5,000 ft (reaching 5,340 ft) and the flap limit speed was marginally exceeded on two occasions. The flight crew positioned the aircraft into a holding pattern, completed the ‘After Takeoff’ checklist and ‘Trailing Edge Flap Disagree’ non-normal checklist and briefed for a second approach to runway 34 with flaps 25.

Prior to leaving the holding pattern, the captain briefed the cabin supervisor about the situation. However, the format of that briefing was the same as what the cabin supervisor would expect for an emergency. As a result, the cabin supervisor perceived that the cabin needed to be prepared for an emergency landing, which was not the captain’s intention.

A second approach was conducted with the autopilot and autothrottle engaged to 136 ft. The aircraft landed without further incident.

Engineers later performed the required aircraft inspections. They could not reproduce the flap fault, however, replaced the left-hand trailing edge flap position transmitter as a precaution.

The ATSB obtained the operator’s investigation report, and interviewed the flight crew and the cabin supervisor. The ATSB also obtained data from the aircraft’s flight data recorder, aircraft maintenance records and relevant sections of the operator’s operations manual. Based on its review of this information, the ATSB concluded that the operator had conducted a detailed investigation and it was unlikely that further ATSB investigation would identify any systemic safety issues.

The ATSB noted that, although there were flight crew errors made during the approach and the subsequent missed approach, the missed approach was conducted at an appropriate time. In addition, while the cabin crew conducted the cabin preparation drill for a non-normal landing when it was not required, doing so was an example of the operation ‘failing safe’ rather than increasing risk.

The operator has subsequently used this incident as a basis for some recurrent training for its flight crew and cabin crew. Consequently, the ATSB has discontinued this investigation.

Safety summary

What happened

On the 5 September 2017, Queensland Rail suburban passenger train 1W33 was being transferred from Mayne Depot to Roma Street Station, Brisbane, Queensland, as a non-revenue operation. Nearing the Roma Street Station at 1628, train 1W33 passed signal RS57 at danger. Signal RS57 was displaying a stop indication (red aspect). Train 1W33 ran through points 226 and travelled east into platform 8. At the same time, another train (15X2) was approaching Roma Street Station on a converging route from the east.

As the train passed signal RS57 at danger, a ‘signal passed at danger’ (SPAD) alarm was generated in the Queensland Rail Management Centre. The Network Control Officer broadcast an emergency call to 1W33 and converging train 15X2. Both trains came to a stop approximately 550 m apart.

The rail infrastructure (points 226) was damaged by train 1W33. There were no reported injuries or damage to the rolling stock.

What the ATSB found

The ATSB found that the driver of 1W33 was distracted from his primary task of driving the train, including observing and reacting to signals, by personal emotional thoughts, when approaching signal RS57 displaying a stop indication.

What's been done as a result

Queensland Rail has initiated a number of strategies to manage the risk of SPADs, including administrative processes, human factors analysis, plus reviews and improvements to existing SPAD controls.

Longer term, Queensland Rail have sought expressions of interest from market leaders to partner with Queensland Rail to implement European Train Control System (ETCS) on sections of the Queensland Rail network. ETCS incorporates automatic train protection that provides for monitoring of train speed and limits of authority to ensure trains stay within designated speed limits and authorised safeworking limits.

Safety message

This incident highlights how distraction increases the risk of a SPAD event. Distractions can be external, but internal thoughts related to significant personal events or circumstances can take away a driver’s attention from their primary task of safely managing the train.

Queensland Rail suburban passenger train 1W33

Source: Queensland Rail

The occurrence

At about 1623 Eastern Standard Time on 5 September 2017, the Queensland Rail (QR) suburban non-revenue passenger train 1W33, departed Mayne Depot and travelled on the Exhibition loop towards Roma Street Station, Brisbane, Queensland (Figure 1).

Figure 1: Train 1W33 route from Mayne depot to Roma Street Station, Brisbane, Queensland

Image shows signal and station locations. Note that the colour of the signal markers denotes the aspect of the signal passed by 1W33.
Source: Google Earth annotated by ATSB.

The penultimate signal prior to Roma Street Station (RS49) displayed a restricted indication (flashing yellow aspect^[1]) for the driver. As train 1W33 approached signal RS49, the driver slowed down in preparation for a 25 km/h speed board placed adjacent the signal. The driver recalled that the Automatic Warning System (AWS)^[2] generated an in-cab alert of the restricted signal RS49 indication ahead and that he acknowledged the AWS alert. At about 1628, train 1W33 continued past signal RS49 towards signal RS57, just west of Roma Street Station. Signal RS57 was displaying a stop indication (red aspect) to allow other train movements to clear Roma Street Station before train 1W33 would be authorised to continue.

Meanwhile, passenger train 1K56 was preparing to depart Roma Street Station from platform 8 (Figure 2). At about 1628, train 1K56 departed Roma Street Station travelling west towards the Springfield line via points 226 set normal.^[3] About 11 seconds after train 1K56 cleared 226 points, train 1W33 proceeded past signal RS57 at danger. Train 1W33 ran through^[4] points 226 about 26 seconds after passing RS57 at danger and travelled east into platform 8.

Figure 2: Western end Roma Street Station layout, Brisbane, Queensland

Image shows track infrastructure layout west of Roma Street Station with the directions of travel and path for trains 1W33 and 1K56.
Source: Queensland Rail annotated by ATSB.

At approximately the same time, another train (15X2) had just departed Central Station and was travelling towards Roma Street Station on a converging route with train 1W33 from the east.

As train 1W33 passed signal RS57, an alarm activated at the QR Rail Management Centre. The Network Control Officer, broadcast an emergency radio message, calling for the driver of 1W33 and the driver of 15X2 to stop. At about 1630 the two drivers confirmed they had stopped.

Train 15X2 came to a stop approximately 25 m prior to signal RS102. Signal RS102 is the signal before the eastern entry signal (RS100) into platform 8 at Roma Street Station (Figure 3). Signal RS102 was displaying a single yellow caution^[5] indication at this time due to signal RS100 displaying a stop indication. The signal interlocking system had restored Signal RS100 to stop when train 1W33 entered platform 8 at Roma Street Station. The two converging trains stopped approximately 550 m from each other.

Figure 3: Train 15X2 route from Central Station towards eastern entry of Roma Street Station, Platform 8

Image shows signal locations and stopping locations of train 1W33 and 15X2. Note that colour of the signal markers denote the aspects displayed after 1W33 entered the platform track circuit at Roma Street Station.
Source: Google Earth annotated by ATSB.

__________

1. Special Caution indication advising the driver to proceed to the next stop signal at a speed not exceeding 40 km/h. Source: Queensland Rail Observance of Signals Manual.
2. An advisory system that provides audible and visual warnings to the driver on the approach to signals.
3. The normal position of points generally refers to the position set to give optimum protection to other routes. The opposite position is referred to as points set reverse.
4. A movement through trailing points set in the wrong position. Damage to the point mechanism and switch rail usually results.
5. A Caution indication, advises drivers that they must expect the next signal to be at Stop. Drivers can proceed towards the next signal being prepared to stop prior to the next signal. Source: Queensland Rail Observance of Signals Manual.

Safety analysis

The ATSB explored the following aspects which have been known as contributing factors in other signal passed at danger incidents.

Environmental conditions

A review of the CCTV^[6] images from the time of the incident and driver comments indicated that the weather conditions were fine and clear. Further to this, no rainfall had been recorded for Brisbane on the day of the incident by the Bureau of Meteorology. The ATSB found that environmental conditions did not contribute to this incident.

Human performance

The ATSB reviewed the health assessment records, drugs and alcohol test results, rosters (with respect to fatigue) and competency records related to the driver of train 1W33. In conjunction with the driver comments, the ATSB found that these human performance factors did not contribute to this incident.

Rolling stock

Train 1W33 consisted of two coupled Suburban Multiple Units (SMU’s), with SMU 261 leading and SMU 279 trailing. Train 1W33 weighed approximately 256 tonnes and had a length of approximately 150 m.

As permitted by the Queensland Rail standard for operational integrity of trains, event recorders^[7] for suburban multiple units SMU 261 and SMU 279 had both been temporarily removed for repairs and therefore did not record rolling stock inputs in relation to this incident (refer to ATSB comment below). However, driver comments and examination of available CCTV footage did not suggest any degradation of braking performance of train 1W33.

Visibility of signal RS57

Signal RS57 was a 4-aspect^[8] main line signal fitted with a junction indicator and low speed shunt signal (Figure 4). Signal RS57 authorised train movements towards signals RS83 or RS85 located at the eastern end of Roma Street Station (platforms 8 and 7).

Figure 4: Signal configuration of RS57, at Roma Street Station, Brisbane

Source: Queensland Rail

Signal RS57 was located at the end of a left hand curve adjacent to Roma Street Station on the right hand side of the track in the direction of travel. Due to the curvature of the track and the placement of the overhead catenary masts, RS57 signal sighting was limited to approximately 103 m.

Queensland Rail managed the limited signal sighting with the application of a 25 km/h speed restriction from signal RS49 leading up to signal RS57 and LED signal lamps had recently been installed. In addition, other controls have been considered and implemented as part of Queensland Rail’s ongoing signal sighting reviews.

The ATSB found that the configuration and sighting of signal RS57 did not contribute to this incident.

Signal SPAD history

The ATSB found that signal RS57 had been passed at danger on six occasions since 2007. Queensland Rail records for these SPAD events noted varying sources of driver distraction as the causal factor for each event.

Figure 5: Signal RS57 10 year SPAD history

Image shows that in 2007 three reports for signals passed at danger for RS57 were recorded, plus three more each in 2009, 2012, and this event in 2017.Source: Queensland Rail records graphed by ATSB.

Following the three SPAD’s in 2007, Queensland Rail identified RS57 as a multi-SPAD signal and had undertaken SPAD risk studies and regular reviews of the signal sighting. These risk studies and reviews led to the identification and, where accepted, the implementation of additional controls. This incident is the first time RS57 had been passed at danger in five years.

Driver distraction

Driver distraction can be understood as a type of inattention and has been defined as ‘the diversion of attention away from activities critical for safe driving toward a competing activity (occurring) voluntarily or involuntarily’.^[9] A physical source of distraction common in today’s society is the use of mobile phones. In this case, the driver reported he was not using his phone and there was evidence showing that his phone was not used at the time of the incident. Similarly, the driver reported no distractions from any radio chatter or noticing anything external to the train.

While the source of distraction can often be physical, it can also be non-physical. For example, a situation where task-irrelevant thoughts interfere with task-relevant thoughts can also contribute to distraction and decrease task performance.

A contributor to irrelevant thoughts is emotional mood states which have been described as providing a third processing layer on top of cognitive and physiological levels. ‘Emotions play an important role in motivating people to initiate and maintain a task in the first place, but they may also interfere with cognitive processing. In particular, under time pressure or threatening conditions, the regulation of our emotions is critical for efficient task performance’.^[10]

A research study into the effects of emotional mood states on irrelevant thoughts and cognitive task performance concluded that ‘people produce more irrelevant thoughts during emotional mood states, and when the proportion of irrelevant thoughts to relevant thoughts increases, performance suffers.’^[11]

In relation to this incident, the driver advised that prior to commencing work he recalled a significant past personal event that resulted in an emotional response. It is likely this contributed to an increase in task-irrelevant thoughts that led to the driver being distracted from observing and reacting appropriately when approaching signal RS57 displaying a stop indication.

ATSB comment

The function of a safety investigation is to identify and reduce safety-related risk, by gathering information and evidence to form conclusions based on the facts.

The ATSB noted that the Queensland Rail standard for operational integrity of trains permitted trains into service on their network without an operational event recorder. In this case, the ATSB relied on CCTV footage and the driver’s recollection for identifying the sequence of events. The absence of recorded data prevented further verification of the driver’s performance against operational practice.

__________

6. Abbreviation for Closed Circuit Television.
7. A device / system designed to resist tampering, with crash-protected non-volatile memory, that records event data to support accident or incident analysis. Source: Queensland Rail Event recorders for rolling stock specification.
8. Although RS57 was fitted with a 4 aspect signal head, only 3 aspects are in use, with the signal not configured to display a green aspect.
9. Regan, M.A., Hallett, C. & Gordon, C.P. (2011). Driver distraction and driver inattention: Definition, relationship and taxonomy. Accident Analysis and Prevention, 43, 1771-1781.
10. Gaillard, A.W.K. (2001). Stress, workload and fatigue as three bio-behavioural states: A general overview. In P.A. Hancock, & P.A. Desmond (Eds.), Stress, workload, and fatigue. Mahwah, NJ: L. Erlbaum.
11. Pennie S. Seibert and Henry C. Ellis, (1991), Memory & Cognition Irrelevant thoughts, emotional mood states, and cognitive task performance, 507-513.

Findings

From the evidence available, the following findings are made with respect to train 1W33 passing signal RS57 at danger near Roma St Station, Brisbane, Queensland, on 5 September 2017. These findings should not be read as apportioning blame or liability to any particular organisation or individual.

Contributing factors

• The driver of 1W33 was distracted by personal emotional thoughts from his task of observing and reacting appropriately when approaching signal RS57 displaying a stop indication.

Other findings

• A SPAD alarm activated at the QR Rail Management Centre prompting the network control officer to broadcast an emergency radio message stopping all affected trains.

Safety action

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

As a result of this occurrence, Queensland Rail has advised the ATSB they are taking the following safety actions:

• A qualitative analysis of the SPAD scenarios at RS57 using a bowtie model.
• Present a health and wellbeing strategic plan to the Executive Leadership Team which includes an organisational resilience and psychological wellbeing education and awareness program.
• Human factors review of driver response to AWS audible indications and reaction times for green and restricted signals.
• Facilitate the decrease in volume of the AWS audible indication at a proceed signal (green) and increase the volume of an AWS audible indication at a restricted signal (double yellow, yellow, and red).
• Queensland Rail have sought expressions of interest from market leaders to partner with Queensland Rail to implement ETCS (European Train Control System) on sections of the Queensland Rail network. ETCS incorporates automatic train protection that provides for monitoring of train speed and limits of authority to ensure trains stay within designated speed limits and authorised safeworking limits.

Sources and submissions

Sources of information

The sources of information during the investigation included the:

• Queensland Rail (QR)
• Train driver involved in incident
• Recorded data
• Rail Industry Safety and Standards Board (RISSB)
• Bureau of Meteorology.

References

Queensland Rail Observance of Signals Manual (MD-10-109).

Queensland Rail Event recorders for rolling stock specification (MD-10-217).

Queensland Rail Operational Integrity Of Trains standard (MD-10-106).

Rail Industry and Safety Standards Board, 2010, National Guideline Glossary of Railway Terminology.

Pennie S. Seibert and Henry C. Ellis, (1991), Memory & Cognition Irrelevant thoughts, emotional mood states, and cognitive task performance, 507-513.

Regan, M.A., Hallett, C. & Gordon, C.P. (2011). Driver distraction and driver inattention: Definition, relationship and taxonomy. Accident Analysis and Prevention, 43, 1771-1781.

Gaillard, A.W.K. (2001). Stress, workload and fatigue as three bio-behavioural states: A general overview. In P.A. Hancock, & P.A. Desmond (Eds.), Stress, workload, and fatigue. Mahwah, NJ: L. Erlbaum.

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 Queensland Rail, the driver of train 1W33, and the Office of the National Rail Safety Regulator.

Any submissions from those parties will be reviewed and where considered appropriate, the text of the draft report will be amended accordingly.

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

What happened

On 3 September 2017, a Fairchild SA227-AC aircraft, registered VH-SEZ, was operating Sharp Airlines flight SH843 from Portland to Essendon, Victoria. The first officer was pilot flying (PF) and the captain was pilot monitoring (PM).^[1]

About 46 NM east of Portland, overhead Warrnambool, and in the cruise at flight level (FL) 170,^[2] the left engine fire warning lights on the annunciator and fire warning panels started to momentarily illuminate. At 1608 Eastern Standard Time,^[3] the flight crew contacted air traffic control (ATC) advising that they required a clearance to descend to 9,000 ft and that they would require a direct track to Essendon.

The fire warning lights continued to flicker and then both remained illuminated. As a result, the flight crew assessed that there was an engine fire. They conducted the memory checklist for in‑flight engine fire, which included:

• shutting down the left engine
• feathering^[4] the left propeller
• shutting off the left fuel supply and hydraulic system
• discharging the fire retardant.

The fire warning lights then extinguished.

The captain looked over his left shoulder at the left engine and could not see any smoke, flames or scorch marks. The flight crew then declared a MAYDAY^[5] to ATC, and reported that they had experienced an engine fire and that the fire was extinguished.

The captain made a public address to the passengers, explaining that they had shut down the left engine due to suspected fire and asking any passengers who had seen smoke or flames to come forward and let him know. No one reported any visual indication of fire.

The flight crew assessed their options for landing, considering both Warrnambool and Avalon Airports. Although they were nearer to Warrnambool Airport, the runway there was shorter and narrower and the wind was gusty, so the crew elected to divert to Avalon Airport. The flight crew reported that a tailwind en route to Avalon and the presence of emergency services also influenced their decision.

At 1613, the flight crew advised ATC that they required direct tracking to Avalon Airport. They received a clearance to do so.

As the aircraft tracked to Avalon, and about 5 minutes after the fire warning lights had extinguished, the lights started to flicker and then came back on to a steady warning. By that time, the crew had tested the integrity of the fire warning loop, discharged the bottle of fire retardant and shut down and secured the left engine.

ATC advised the crew that Runway 36 at Avalon had an occasional 5 kt tailwind and Runway 18 had an occasional tailwind of 9 kt. The flight crew responded that they would require Runway 36. When the aircraft was about 15 NM from Avalon, the captain took over the pilot flying role. ATC provided heading guidance to the flight crew, and the aircraft landed at Avalon at about 1626.

After landing, firefighters advised that there was no sign of fire in the engine. Consequently, there was no need to conduct an emergency evacuation and the passengers and crew disembarked normally.

Engineering inspection

Engineering examination did not identify any signs of fire, smoke or heat damage in the left engine bay. Engineers inspected the four left engine fire sensor probes and all appeared to be functioning normally, but they replaced the lower rear sensor due to cracking of the ceramic insulator.

On the fire extinguisher control panel, each engine has a button and a tri-coloured light – red for fire, green to indicate system integrity of the fire loop, and yellow to indicate a bottle of fire retardant had been discharged and was empty. After the crew pressed the fire button, the empty (yellow) light did not illuminate. This was despite the fact that the fire bottle pressure gauge was indicating zero, after being full pre-flight, and there was indication of discharge into the engine. The yellow light did illuminate when the bulb was tested however, which indicated an issue with the circuit.

Engineers found that the logic control unit in the fire extinguisher control assembly had failed, resulting in erroneous illumination of the engine fire warning. The logic module was replaced, along with the fire extinguisher control panel.

Approach and landing

The captain commented that two out of their six cyclic biannual simulator checks included fire warnings, engine shutdowns and single engine actions. In this occurrence, the engine shutdown, diversion and single-engine approach and landing were consistent with the simulator training. There were no issues with controlling the aircraft or with its performance.

Findings

These findings should not be read as apportioning blame or liability to any particular organisation or individual.

• The fire extinguisher control logic module failed, resulting in an erroneous engine fire warning.

Safety message

This incident highlights the importance of well-designed simulator training and robust threat and error management procedures. The captain commented that it was important to treat fire warnings as legitimate indications of fire. Additionally, the declaration of an emergency alerts air traffic control and enables the provision of appropriate assistance.

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

__________

1. Pilot Flying (PF) and Pilot Monitoring (PM): procedurally assigned roles with specifically assigned duties at specific stages of a flight. The PF does most of the flying, except in defined circumstances such as planning for descent, approach and landing. The PM carries out support duties and monitors the PF’s actions and the aircraft’s flight path.
2. Flight level: at altitudes above 10,000 ft in Australia, an aircraft’s height above mean sea level is referred to as a flight level (FL). FL 170 equates to 17,000 ft.
3. Eastern Standard Time (EST): Coordinated Universal Time (UTC) + 10 hours
4. Feathering: the rotation of propeller blades to an edge-on angle to the airflow to minimise aircraft drag following an inflight engine failure or shutdown.
5. MAYDAY: an internationally recognised radio call announcing a distress condition where an aircraft or its occupants are being threatened by serious and/or imminent danger and the flight crew require immediate assistance.

What happened

At about 0219 Western Standard Time^[1] on 7 August 2017, a Pilatus PC‑12/47E aircraft, registered VH-OWS (OWS), taxied at Meekatharra Airport, Western Australia. The aircraft was operating as a Royal Flying Doctor Service (RFDS) air ambulance flight to Jandakot Airport, and had a pilot, two medical staff, and a patient on board.

At about 0227, OWS lined up and departed from runway 09. The night was clear with almost a full moon. At about 1,000 ft above ground level, when the pilot turned the aircraft to depart from overhead the airport, he noticed small beads forming on the outside of the windscreen. As there was no rain in the area, he shone a torch on the windscreen. The beads had formed into a steady stream of fluid moving up the windscreen, and noting its thickness, the pilot surmised that it was oil.

The pilot also knew that the aircraft’s propeller shaft seal had been replaced during maintenance on the previous day. Before this flight, he had followed his normal routine of checking the engine oil quantity and found that it was indicating ‘full’.

In assessing the situation, the pilot noted that all the engine parameters were in the normal range and he elected to return to Meekatharra Airport. He levelled the aircraft at an altitude of about 5,500 ft, and informed air traffic control of his intention to return for a landing on runway 09.

The pilot then conducted a wide descending circuit, slightly higher than normal to maintain glide capability in case the leak got worse and affected the engine’s operation. The visibility through the windscreen had reduced and by the time OWS turned onto the base leg of the circuit, oil was streaming down the side windows. On the final leg of the circuit, the pilot saw the runway lights blurring ‘like in heavy rain’. He conducted a normal landing, taxied to the parking bay, and shut down the engine. The aircraft was not damaged and no one was injured.

After landing, the pilot checked the aircraft and found oil on the engine cowling (Figure 1) and dripping/pooling on the tarmac. When the engine cowling was opened, a small amount of smoke could be seen coming from where oil had contacted hot engine parts.

Figure 1: Oil on the outside of the engine cowling

Source: RFDS

Propeller shaft seal replacement

The aircraft operator (RFDS) investigated the incident and identified that the propeller shaft seal had been incorrectly assembled (Figure 2). The seal had been replaced at Jandakot on 6 August 2017 (the day before the incident) after which the aircraft had flown to Meekatharra, a flight of about 1 hour 40 minutes.

Figure 2: Propeller shaft seal assembly at the time of the incident

Source: Pratt & Whitney Canada, modified by the ATSB

The day before the occurrence, the on-call licensed aircraft maintenance engineer (engineer) at the Jandakot maintenance base was asked to check an engine oil leak on OWS. When he examined the engine, he found oil on the engine cowling and the forward part of the engine. Having removed the oil, he investigated further by running the engine three times without any further evidence of an oil leak.

The engineer then contacted the engineering manager to discuss the defect. Based on their experience and the signs of the oil leak, they decided that the engineer would replace the propeller’s shaft seal.

When replacing the seal, the engineer assembled the parts in the order that he recalled from disassembly, and the diagram in the engine manufacturer’s, Pratt & Whitney Canada (P&WC), illustrated parts catalogue (IPC). He recalled that the half flat spacer was located forward of the seal—the same as the IPC (Figure 3). After replacing the seal, he ran the engine and no oil leaks were evident.

Figure 3: Propeller shaft seal assembly parts diagram in the IPC

Source: Pratt & Whitney Canada, annotated by the ATSB

The RFDS investigation found that the IPC was not intended to be used for seal replacement nor did it show the correct order of parts in the seal assembly. Only the P&WC engine maintenance manual (EMM) was intended to be used for maintenance, including propeller shaft seal replacement. The diagram in the EMM showed the half flat spacer correctly located aft of the seal (Figure 4). The EMM assembly procedure permitted the positions of the seal and the seal spacer shown in Figure 4 to be interchanged. In both configurations however, the half flat spacer was to be installed aft of the seal. The engineer had referred to the EMM but had not identified the discrepancy between the seal assembly diagrams in the IPC and the EMM.

Figure 4: Propeller shaft seal assembly diagram in the EMM

Source: Pratt & Whitney Canada, annotated by the ATSB

The RFDS investigation also identified that in the time leading up to the incident the engineer had worked extended work hours, and had had 4 days off in the previous 27 days. The engineer also indicated to RFDS that he had been feeling tired when replacing the seal as he had not slept well the previous night. He also indicated receiving a work-related phone call while replacing the seal which he believed was a distraction.

Safety analysis

The oil leaked from the propeller shaft seal assembly because parts of the seal had not been assembled in the correct order. As the half flat spacer had been placed forward (instead of aft) of the seal, the seal was compressed and was pushed against the seal runner (Figure 2). In that configuration, the friction from the rotating seal runner resulted in abnormal wear of the seal allowing engine oil to leak.

When assembling the propeller shaft seal, the engineer had used the indicative diagram in the IPC instead of the diagram in the EMM, which showed the correct order of parts. He did refer to the EMM but did not notice that the assembly procedure and diagram in it were different to the diagram in the IPC. He also relied on his memory to recall the order in which the parts were removed, which may have been affected by fatigue associated with longer working hours and disrupted sleep.

Findings

These findings should not be read as apportioning blame or liability to any particular organisation or individual.

• Shortly after VH-OWS took off from Meekatharra Airport the pilot saw oil leaking from the engine on to the windscreen. In response, he turned the aircraft back and safely landed at the airport.
• Engine oil leaked from the aircraft’s propeller shaft seal as it had not been correctly assembled when replaced the previous day.
• When assembling the propeller shaft seal, the engineer used an indicative (but technically incorrect) diagram in the engine manufacturer’s illustrated parts catalogue instead of the correct sequence detailed in the manufacturer’s maintenance manual. He also relied on his memory of disassembling the seal, which may have been affected by fatigue associated with disrupted sleep and recent longer working hours.

Safety action

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

Royal Flying Doctor Service

The aircraft operator, RFDS, advised the ATSB that the following safety action had been taken:

• Maintenance engineers have been reminded that the use of memory, IPCs and engineering by comparison, are not suitable methods for conducting aircraft maintenance.
• The engineers have also been reminded to use the EMM and associated diagrams for the maintenance, including the propeller shaft seal assembly.
• Engineer’s overtime is being monitored while a more permanent solution for workplace fatigue management is being considered.

Transport Canada

Transport Canada, Canada’s transport safety regulator, advised the ATSB of the following safety action, which it believes will mitigate future risk of this type of occurrence.

• Pratt & Whitney has initiated a manual revision for the IPC to add an illustration of the most current (post service bulletin) configuration of the seal assembly, and amend the EMM to add a caution in the reduction gearbox maintenance practices section.

Safety message

Maintenance engineers should ensure that the appropriate technical documents are used for any maintenance task (Civil Aviation Safety Authority (CASA) maintenance poster Check the data also refers). In this occurrence, use of the inappropriate technical document resulted in the incorrect assembly of the propeller shaft seal, which then leaked.

The CASA publication Safety Behaviours: Human Factors Resource Guide for Engineers, notes that many maintenance engineers use personal sources of unapproved technical documents. A common problem faced by engineers is the requirement to follow procedures and time pressures to complete maintenance. A better understanding of these demands to complete the maintenance by operators and maintenance organisations could help them identify informal work practices and areas for improvement, including the use of appropriate technical documents.

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

__________

1. Western Standard Time (WST): Coordinated Universal Time (UTC) + 8 hours.

Safety summary

What happened

On the morning of 29 August 2017, freight train 8279 (one locomotive and 23 wagons) was travelling north from Mackay, Queensland. At about 0945, after passing over the Wintons Road level crossing, about 64 km north of Mackay, the train driver observed a truck approaching the next level crossing. The unsealed crossing (ID 5318) provided vehicle access between Wagoora-Yalboroo Road and the Bruce Highway. It was equipped with passive traffic control devices (stop signs).

The train driver sounded the train horn while observing the truck slow and stop at the crossing. The train driver noticed that the truck appeared to be facing slightly away from the approaching train when stopped, before it then proceeded to cross the track in front of the train. The train driver gave a second sustained sounding of the horn, placed the brake handle into the emergency position, before moving to the foot-well for protection.

Soon after, train 8279 collided with the driver’s cab of the truck. The truck driver sustained fatal injuries. The train driver was shaken but otherwise unhurt.

What the ATSB found

The truck driver’s ability to sight the approaching train was probably restricted due to the truck’s cab-design and the likelihood that the vehicle stopped at an angle to the railway track. Once the truck proceeded into the path of the train, there was insufficient time for the train driver to stop the train before colliding with the truck.

The investigation also noted that road vehicles had progressively cut the corner when entering/exiting the private road, resulting in the gradual widening of junction at Wagoora-Yalboroo Road. This could influence the position of road vehicles when stopped at the crossing and in turn, affect the ability for drivers of some larger vehicles to sight an approaching train.

In considering the Queensland government’s long-term level crossing safety strategy, the ATSB found that private (occupation) crossings had been specifically excluded. However, the ATSB notes that providing unrestricted public access to a private (occupation) crossings could present a level of risk similar to that of a public road crossing. While the program has achieved a wide range of positive safety initiatives and improvements, the ATSB found that exclusion of private (occupation) crossings from the strategy potentially removes an opportunity to further improve safety outcomes for all level crossing users across Queensland.

What's been done as a result

While the Queensland Level Crossing Safety Group (QLCSG) focus remains on the higher risk public level crossings, they acknowledge that private (occupational) level crossings also require management under the Rail Safety National Law. Stakeholders of the QLCSG agree to review the status and treatment at private level crossings used by the public (such as in this case). Where appropriate, private crossings will be included as a subset of public level crossings for the purposes of their safety strategy.

Queensland Rail (QR) conducted an audit of level crossing ID 5318. The installation was found to be consistent with the QR standard. However, some enhancements were identified and programmed for installation under routine maintenance activities.

With respect to all private (occupation) crossings in Queensland, QR is liaising with private property owners to enter into licence or interface agreements to ensure the safe operation and use of the private level crossings. Where appropriate, private crossings would be closed or, if used by the general public, upgrade to public crossing status.

Safety message

Truck drivers are reminded that cabin designs can limit their view to the left out of the passenger window. As a result, when stopped at a rail level crossing, this can influence the distance away a train approaching from their left can be seen. Taking care to maximise this sighting distance is important to ensure a safe crossing.

Rail operators and governments should continue to implement programs for improved safety at railway level crossings. This should include considering safety strategies for private (occupation) crossings, especially where the level of risk may be similar to that of a public road crossing.

The occurrence

On 27 August 2017, train 8279 departed Acacia Ridge, Brisbane, Queensland, for Stuart Yard in Townsville. Train 8279 consisted of one locomotive (2813) and 23 wagons. The train was 480 m in length, with a total weight of 1370 t.

On the morning of 29 August 2017, train 8279 arrived at Mackay, about 964 km^[1] north of Brisbane. A crew change was undertaken, before train 8279 departed Mackay at about 0845 Eastern Standard Time (EST), and continued traveling north.

At about 0937, train 8279 passed through a crossing loop at Calen (1018.590 km). The track largely runs parallel to the Bruce Highway (to the west) with sugar cane fields on both sides. About 8 km after the crossing loop, the track tops a rise before turning in a more westerly direction and descending towards Wintons Road level crossing. Wintons Road is a sealed road with active traffic control devices (flashing lights) at the level crossing.

At about 0945, train 8279 passed over the level crossing at Wintons Road (Figure 1). Soon after, while travelling at about 84 km/h,^[2] the train driver observed a dust cloud ahead and assumed it was a vehicle travelling on the parallel unsealed road to the north—Wagoora-Yalboroo Road. The train driver then sighted a truck approach the next level crossing (1028.693 km). The unsealed level crossing was equipped with passive traffic control devices (Stop signs).

The train driver sounded the train horn while observing the truck slow and stop at the crossing. The train was about 230 m from the crossing. The train driver described the truck as having stopped at an angle facing away from the approaching train. The truck then proceeded to cross the track in front of the train. The train driver immediately gave a second sustained sounding of the horn (about 3 seconds), placed the brake handle into the emergency position, before moving to the foot-well for protection.

Figure 1: Aerial view of the level crossing approach and road alignment

The truck was travelling west along Wagoora-Yalboroo Road before turning left towards the railway crossing and the Bruce Highway. The train was also travelling west, parallel to the road.
Source: Google Earth, annotated by ATSB

A few seconds later, train 8279 collided with the driver’s cab of the truck. The train travelled a further 640 m before stopping. Shortly after, the driver of train 8279 contacted the Queensland Rail Network Control Centre to report the collision.

The truck driver sustained fatal injuries. The train driver was shaken but otherwise unhurt.

__________

1. Rail distance measured from the track kilometre zero point located at Roma Street Station, Brisbane.
2. The posted speed limit was 100 km/h.

Context

Level crossings provide for traffic management where a railway line and road intersect. Crossings are usually categorised as either public crossings or private crossings. Public crossings provide unrestricted access to all road users (the general public) for everyday use. Private crossings (or ‘occupation’ crossings) are on private land and are often exclusively used by the landowner.^[3]

The collision on 29 August 2017 occurred at a crossing defined by Queensland Rail (QR) as an Occupation/Private rail crossing, QR ID number 5318. The road provided vehicle access between Wagoora-Yalboroo Road and the Bruce Highway, and was about 120 m in length.

Standards for traffic control devices at railway crossings

Australian Standard 1742.7:2016 Manual of uniform traffic control devices, Part 7: Railway crossings, prescribes the requirements for road markings, roadside signs and configuration of traffic controls at railway crossings throughtout Australia. The standard states that the requirements are:

…not applicable to railway crossings provided for the exclusive use of the occupier of private land or by other people with the knowledge and agreement of the occupier (sometimes known as ‘occupation’ crossings).

While crossing ID 5318 was classified as an occupation crossing, the road provided unrestricted access between Wagoora-Yalboroo Road and the Bruce Highway. Consequently, the crossing was almost certainly used as a public thoroughfare, in which case the requirements of Australian Standard 1742.7:2016 may apply. The minimum treatment specified in the standard was a RX-1 assembly (Give-way sign), subject to documented conditions.^[4]

In 1997, QR engaged a consultant to develop an organisational standard for all occupation crossings on their rail network. The consultant report recommended a RX-2 assembly (Stop sign) as the standard for all occupation crossings. This recommendation exceeded the minimum treatment requirement of Australian Standard 1742.7:2016^[5] and was consistent with the installation at crossing ID 5318 (Figure 2).

The consultant report also recommended minimum sighting distances at occupation crossings. For train speeds of 100 km/h on single track, the minimum sighting distance for a vehicle stopped at the stop sign was recommended as 450 m. QR records indicate that sighting distances at occupation crossing ID 5318 was in excess of 700 m.

In general, the traffic control devices installed at occupation crossing ID 5318 were consistent with the requirements documented in the QR organisational standard.

Figure 2: Traffic control devices at occupation crossing ID 5318

RX-2 assembly (Stop sign) to Australian Standard 1742.7:2016. Source: AS1742.7 and Queensland Rail

Environmental conditions

The closest Bureau of Meteorology (BOM) weather station was at Proserpine, about 56 km north of occupation crossing ID 5318. At 0900 (about 45 minutes before the collision), the weather was clear with a light south-easterly breeze and the temperature was about 22 °C.

The train driver reported seeing dust from a vehicle travelling on Wagoora-Yalboroo Road. Considering the road was north of the track and the wind was from the south-east, any dust would have been blown away from the railway line. It is very unlikely that the dust would have obscured the truck driver’s vision of the approaching train.

The weather conditions were unlikely to have been a factor in the collision.

The sun’s azimuth^[6] and altitude^[7] was 52° 09’ 51” and 43° 42’ 47” respectively. Therefore, the sun was relatively high in the sky, and to the north-east.

In this instance, train 8279 was approaching from the east. For a vehicle stopped at the crossing, the sun would have been almost directly behind the truck driver at an altitude of about 43° above the horizon (Figure 3).

It is unlikely that sun-glare or any other sun related effects would have been a factor in the collision.

Figure 3: Sun position and sighting direction of truck driver

The sighting direction for the truck driver was towards the east. The sun was almost directly behind the truck driver at an altitude of about 43° above the horizon. Source: Google Earth, annotated by ATSB

Road vehicle

The vehicle was a 1983 Scania model P112 series 2, twin-steer truck and had been fitted with a flat tray body in May 2017 (Figure 4). The vehicle had a tare mass of 12.2 t and a gross vehicle mass of 26.5 t. Inspection, registration, insurance and modification certificates were all current.

The ATSB noted that the truck had a relatively prominent B-Pillar (Figure 4). The B-pillar of heavy vehicles is a feature known to obscure a driver's vision. It is likely that, under some conditions, a driver of this truck would experience a restricted view to the left due to the vehicle’s B-pillar.

Figure 4: Road vehicle

The vehicle was a 1983 Scania model P112 series 2, twin-steer truck fitted with a flat tray body. The vehicle was loaded with a storage container, fuel tank and air compressor (inset). Note that under some conditions, it is very likely the truck driver’s vision to the left may be restricted due to the vehicle’s B-pillar. Source: Qld Police, Annotated by ATSB

At the time of the accident, the vehicle was loaded with a storage container, fuel tank and air compressor (Figure 4, inset). All were securely mounted on the flat tray body.

The storage container was relatively lightly loaded; an oil drum secured to the container frame, with a number of smaller fluid containers and hand tools stored on racks. The fuel tank^[8] was found to be full (about 1650 lt).

There was no indication that the vehicle was overloaded. Similarly, and considering the fuel tank was full, it is unlikely that any load shift would have affected vehicle dynamics.

The damaged vehicle was examined by the Queensland Police Vehicle Inspection Unit. The inspection found an amount of water retained in the air (brake) system tanks. However, the report noted the train driver’s observations that the truck had stopped at the crossing before proceeding, suggesting the brakes appeared to have been operating immediately prior to the incident. The report concluded that there were no obvious or apparent mechanical defects that could have contributed to the cause of the incident.

Road vehicle driver

The truck driver was appropriately qualified and had extensive experience driving heavy vehicles.

Fatigue

In the context of human performance, fatigue is a physical and psychological condition primarily caused by prolonged wakefulness and/or insufficient or disturbed sleep.^[9] Fatigue can have a range of influences on performance, such as decreased short-term memory, slowed reaction time, decreased work efficiency, reduced motivational drive, increased variability in work performance, and increased errors of omission.^[10] Fatigue impairment has been identified as causal in many transport related accidents.

The truck driver’s work-hours were generally between 0630 and mid-afternoon. The driver had worked these times in the three days prior to the incident, having previously come off of a three-day break. It was evident that work was generally undertaken during daylight hours. There were also sufficient off-duty hours at night, providing ample opportunity for restorative sleep prior to commencing work the next day. In addition, full rest-days were provided, giving the opportunity for recovery away from the work environment.

On the day of the incident, the truck driver started work at 0630. He had worked in an on-site shed until about 0800 before travelling out to the cane fields in the truck. The accident occurred at about 0945, less than half way through the driver’s normal workday. It was considered unlikely that the effects of fatigue were present either leading up to or at the time of the collision.

Medical and toxicology

Post-mortem examination found no evidence of any significant natural disease. Similarly, toxicology testing revealed no evidence of alcohol or drugs in blood and urine.

Looked but did not see

Research has shown that, in road accidents, critical/important information may have been detectible but the motorist did not attend to or notice it because their mental resources were elsewhere.^[11] Furthermore, research into a phenomenon known as ‘inattentional blindness’^[12] has shown how a person may fail to detect an object even though they were looking directly at it.

The human mind has limited resources for perceptual and memory processing. To cope with this limitation, a mechanism called ‘attention’ acts as a filter to focus this resource on specific tasks. Research suggests that inattentional blindness can occur when attention is mistakenly filtered away from important information^[13] and can be affected by mental workload, expectation, conspicuity and capacity.

In this case, the vehicle driver approached the railway crossing and was observed to have stopped at the stop sign, albeit at an angle facing away from the approaching train. It is possible that the driver looked towards the approaching train, but did not see or perceive its approach. Soon after stopping, the truck driver proceeded into the path of the train.

Expectancy

Expectancy can be understood as the extent to which an event or condition is expected to occur or be present at a particular time and place. An individual’s expectation can influence their attention to (and preparation for) that event or condition.^[14]

Studies undertaken into motorist behaviour have found that drivers who are familiar with a railway crossing are more likely to be involved in a crossing incident than drivers unfamiliar with the crossing.^[15] An influencing factor is where motorists familiar with a crossing with relatively low train frequencies, don’t expect a train since they are rarely seen.

The frequency of train movements across the railway crossing was relatively low—about 15 train movements per day. Therefore, while the truck driver was aware of the railway crossing due to regularly working in the area, the probability of him encountering trains was relatively low. It is therefore possible that the truck driver’s familiarity with the railway crossing and a low expectation of encountering a train contributed to him not noticing the approaching train.

Driver distraction

Distraction can be understood as a type of inattention, where a person’s attention is diverted by a particular event or object. Driver distraction has been more specifically defined as ‘the diversion of attention away from activities critical for safe driving toward a competing activity (occurring) voluntarily or involuntarily.’^[16]

Driver distraction can involve a range of factors either inside or outside a vehicle that draw on limited human physical, visual and cognitive resources, and can result in a degradation of the driver’s performance. For example, eating, drinking, operating devices integral to (or brought into) the vehicle (such as a mobile telephone), and smoking, are all activities that may distract from the driving task.^[17]

While the source of distraction may often be physical, it can also be non-physical. For example, a situation where task-irrelevant thoughts interfere with task-relevant thoughts, may contribute to distraction and decrease task performance. A research study into the effects of irrelevant thoughts and cognitive task performance concluded that ‘when the proportion of irrelevant thoughts to relevant thoughts increases, performance suffers.’^[18]

The truck driver was the sole occupant, so distraction by another person did not occur. The truck driver had a mobile phone, but there was no evidence of it being used at the time of the incident.

Based on the available evidence, there was nothing to indicate that the truck driver’s attention had been diverted by a distracting object or event.

It is possible, however, that the truck driver’s thoughts were directed towards other tasks. That is, rather than thinking about the task of look for an approaching train, the driver’s attention may have been directed towards some other task-irrelevant thoughts.

Summary of driver behaviour

Other than witness observations, there was very little objective evidence for verifying driver behaviour. The truck was observed stopping at the crossing before the truck driver proceeded into the path of the train. There was no evidence to suggest that the effects of fatigue, drugs, alcohol, or any medical condition contributed to the collision.

In the absence of any evidence to the contrary, it is possible that the truck driver looked towards the approaching train but did not see or perceive its approach. It is also possible that distraction and a low expectation of encountering a train contributed to him not noticing the approaching train. However, the combination of the truck’s cab-design (B-pillar) and that it stopped at an angle facing slightly away from the approaching train likely restricted the truck driver’s ability to sight the train.

Rail vehicle

Freight train 8279 was operated by Aurizon and consisted of one locomotive (2813) and 23 wagons. The train was 480 m in length with a total weight of 1370 t.

The train was crewed by a single driver, who had about nine years’ train driving experience. At the time of the collision, the train driver was appropriately qualified, assessed as competent and medically fit for duty.

The train driver was rostered to start work on 29 August 2017 at 0700, having had the previous two days off. The driver took control of train 8279 at 0830, about 75 minutes before the accident. It was considered very unlikely that train driver fatigue contributed in any way to the collision.

Following the accident, the train driver undertook routine drug and alcohol testing which returned a negative result.

Conspicuity

Conspicuity refers to an object’s ability to capture attention. Physical factors that affect the conspicuity of an object include size, contrast and movement. The approaching train was large and travelling relatively quickly. However, from the truck driver’s perspective, the perception of size and speed are likely to have been low due to its distance away (more than 200 m) and angle of approach (almost directly towards the truck driver).

The Rail Industry Safety and Standards Board (RISSB) develop and publish Australian Standards for the rail industry. Australian Standard AS 7531:2015 Rolling stock standard – Lighting and Visibility documents the requirements for lighting and rolling stock visibility and is applicable to new and existing locomotives.

The Australian standards states that the front of locomotives shall have areas of high visibility colour, either yellow, orange, orange-red or red. Locomotives shall also have white marker lights and at least one white headlight, mounted at least 2.3 m above the rail. The standard also makes reference to a requirement for visibility lights. The RISSB standard stated:

The primary purpose of locomotive visibility lights is to enhance the visibility of the front-end locomotive of a train from the perspective of a driver of a motor vehicle approaching a level crossing.

The two white visibility lights are mounted on either side of the locomotive front, between 600 mm and 1200 mm above the top of the rail. The standard stated that lights must alternately flash on and off when the horn is sounded and continue flashing for at least fifteen seconds after the horn has been sounded.

QR also publish their Interface Standards (MD-10-194, dated 23 July 2014) which prescribes the minimum requirements for rolling stock operating on QR infrastructure. The QR standard was largely consistent with the RISSB standards regarding colour, headlights and marker lights. While the QR standard also included the requirement for visibility lights, it did not make mention of any requirement for visibility lights to flash when the horn is sounded.

In this case, the front of locomotive 2813 was painted yellow as per the RISSB and Queensland Rail visibility standards.

The operational condition of the 2800 class locomotives headlights is a parameter recorded on the locomotive data log. In this case, the data log for locomotive 2813 indicates the headlights were illuminated on hi-beam while approaching the level crossing. Figure 5 also shows that both the headlights and the marker lights were still illuminated after collision.

Figure 5: Rail vehicle – Locomotive 2813 (post-collision)

To improve conspicuity, the front of locomotive 2813 was painted yellow and its headlights were illuminated. Source: Qld Police

The 2800 class locomotives also have visibility lights, which sit to the side of the coupler behind a perforated protection plate (Figure 5). The lights are manually controlled by the train driver and operate continuously when switched on, but do not flash when the horn is sounded.

When photographed following the collision, one of the visibility light was destroyed as a consequence of the collision and the other was not illuminated (Figure 5). The ATSB noted that a photograph of the driver’s overhead console indicated the switch controlling the visibility lights was in the off position.^[19] The train driver advised that it is normal to travel with the visibility lights on, and recalled that the lights were on in this case.

Discussions with other drivers of 2800 class locomotives noted that it is routine to operate the train with the headlights and visibility lights on. When stopping at a station or crossing loop, a driver might turn the visibility lights off and switch headlights to low beam to avoid light glare affecting an opposing train/driver. However, when departing, drivers usually turn on both the locomotive generator field and the visibility lights at the same time, since these switches are adjacent each other. In this case, it was not unexpected that the headlights remained on after the train stopped following the collision and the train driver had secured the locomotive, as this would be normal practice when stopping. However, it could not be confirmed at what time the visibility lights were switched on or off, since the operational condition of the visibility lights is not a parameter recorded on the locomotive data log.

Locomotive 2813 incorporated all lighting and visibility features documented in the QR interface standard. However, an enhancement documented in the Australian standard (flashing lights when the horn is sounded) was not included on locomotive 2813.

In this instance, the headlights were switched on, but the condition of the visibility lights could not be verified. While visibility lights, and their flashing feature if available, may capture a road vehicle driver’s attention under some conditions, it is unknown if it would have resulted in a different outcome in this case.

Audible devices

Given the size and weight of most trains it is not possible to brake at anywhere near the rate of a road vehicle. Heavy freight and locomotive hauled passenger trains normally take in excess of 1 km to stop from high track speeds, even if the train driver initiates an emergency application of the brakes.

In an event such as this, by the time the train driver observes a truck begin to move into the path of the train, there is insufficient time to take any avoiding action to prevent the collision other than sounding the whistle and applying the brakes. In this specific case, the brakes had only barely begun to take effect (speed reduced from 84 km/h to 83 km/h) as train 8279 traversed the crossing.

Historically, audible devices have been considered an important component in the systems used to warn motorists of an approaching train. However, soundproofing, air conditioning and entertainment systems in modern vehicles have generally reduced the effectiveness of audible warnings.^[20] Similarly, surrounding sounds such as vehicle engine noise (potentially significant for heavy vehicles) and other road noises (nearby traffic) may reduce the ability for a vehicle driver to hear the sounding of a train horn.

In this case, the driver of train 8279 sounded the locomotive horn twice while approaching the level crossing. The first sounding was when the truck was observed having stopped at the crossing. The second sustained sounding was when the truck was observed starting to move into the path of the train, at which point the train driver placed the brake handle into the emergency position, before moving to the foot-well for protection.

Once the truck moved forward onto the crossing, a collision was inevitable considering the proximity of train.

__________

3. Office of National Rail Safety Regulator (ONRSR) – Policy: Railway crossings
4. Conditions related to (road vehicle) approach speed, sighting and vehicle densities. If the conditions cannot be met, then full treatments would be required.
5. AS 1742.7:2016 specifies limitations on the use of minimum treatment crossings such as road traffic speed and visibility distance. Full passive control treatments would be required at crossings not meeting these limitations.
6. Azimuth is the clockwise horizontal angle (in degrees, minutes and seconds) from true north to the sun/moon.
7. Altitude is the vertical angle (in degrees minutes and seconds) from an ideal horizon, to the sun/moon.
8. External measurements: 900 mm x 920 mm x 2000 mm
9. National Transport Commission (2008). National Rail Safety Guideline. Management of Fatigue in Rail Safety Workers.
10. Battelle Memorial Institute (1998). An Overview of the scientific literature concerning fatigue, sleep, and the circadian cycle, Report prepared for the Office of the Chief Scientific and Technical Advisor for Human Factors, US Federal Aviation Administration.
11. Green, M. & Senders, J. (2004). Human error in road accidents. Retrieved 13 March 2007 from www.visualexpert.com
12. Mack, A. & Rock, I. (1998), Inattentional Blindness. MIT Press: Cambridge, USA
13. Green, M (2004), Inattentional Blindness & Conspicuity. Retrieved 13 March 2007 from www.visualexpert.com
14. Wickens C.D. & McCarley, J.S. (2008). Applied Attention Theory. CRC Press: Boca Raton. pp 55-57.
15. Yeh, M. & Multzer, J. (2008). Driver Behaviour at Highway-Railroad Grade Crossings: A Literature Review from 1990-2006. Human Factors in Railroad Operations. United States Department of Transportation, Federal Railroad Administration: Washington DC.
16. Regan, M.A., Hallett, C. & Gordon, C.P. (2011). Driver distraction and driver inattention: Definition, relationship and taxonomy. Accident Analysis and Prevention, 43, 1771-1781.
17. Young, K.L., Regan, M.A., & Hammer, M. (2003). Driver Distraction: A review of the literature. Monash University Accident Research Centre. Available from: www.monash.edu.au
18. Pennie S. Seibert and Henry C. Ellis, Memory & Cognition (1991), Irrelevant thoughts, emotional mood states, and cognitive task performance, 507-513.
19. The photographs were provided by Queensland Police. Metadata indicates that the photographs of the locomotive were taken about 2.5 to 3 hours after the collision.
20. It should be noted that the use of portable entertainment systems (iPods etc.) has also significantly reduce the effectiveness of audible warning devices for bicycle riders and pedestrians.

Safety analysis

For this accident, there was no evidence to suggest that factors such as sun-glare, mechanical defect, fatigue, drugs, alcohol, or medical condition contributed to the collision. Similarly, the traffic control devices installed at this location were consistent with Queensland Rail’s design standards for private (occupation) crossings.

In this case, the train driver described the truck as having stopped at an angle facing away from the approaching train. Consequently, the truck driver’s ability to sight a train approaching from the left was examined further.

The ATSB also examined the long-term programs and strategies for improving safety outcomes for all level crossing users across Queensland.

Vision obstructions due to vehicle

While most vehicles exhibit features that may obscure a driver's vision, it is particularly true for heavy vehicles. Considering the cab-design of the vehicle involved in this incident (Figure 4), it is very likely that under some conditions, the driver’s vision to the left would have been restricted due to the vehicle’s B-pillar.

The Queensland Police conducted visibility testing using a vehicle with very similar side window and B-pillar configurations.^[21] Testing showed that clear visibility was available to 90° for a normally seated driver simply turning the head to look out the passenger side window. If the driver was to lean forward over the steering wheel, the angle of vision increased to about 100°. However, vision beyond 100° to the left was not possible through the passenger side window (Figure 6).

Figure 6: Road vehicle visibility testing

Road vehicle visibility testing indicated a complete restriction to driver visibility beyond 100°, when the driver was looking out the left side passenger window. Source: Qld Police, annotated by ATSB

Table 1 shows the results of the Queensland Police road vehicle visibility testing and is presented in terms of road vehicle angle relative to the rail line. As this angle decreases, a road vehicle driver is required to look further to the rear of 90° to sight along the rail line. It is evident from testing, that sighting objects through the passenger side window is significantly reduced if the road vehicle stops at an angle to the rail line. If the vehicle angle relative to the rail line was less than 80°, the driver would be unable to sight the full distance along the rail line, even when leaning forward over the steering wheel.

Table 1: Road vehicle visibility – test results

Results from road vehicle visibility testing on a vehicle similar in configuration to the incident vehicle, based on a stopping distance of 2.5 m before the rail line. Note, the measured results have been rounded to the nearest metre. Source: Qld Police

The Austroads Guide to Road Design provides guidance to road designers on the geometric design of all types of road intersections and crossings. The Queensland Department of Transport and Main Roads (TMR) publish their Road Planning and Design Manual that largely accepts the requirements documented in the Austroads publication.

The Austroads guide notes that there are no design rules dealing with visibility from vehicles, but makes reference to published information regarding restrictions due to vehicle design^[22] (Figure 7). Considering the published restrictions due to vehicle design, TMR policy was that road centre lines should be designed to intersect at between 70° and 110° in both urban and rural situations.

Figure 7: Sight restrictions due to vehicle design

The Austroads guide notes that vision form the cab of a vehicle becomes increasingly difficult to the side and rear of the vehicle. The rectangle represents a vehicle; the circle represents the vehicle driver. Note the area highlighted indicating where vision from a truck may be restricted due to the vehicle’s B-pillar. Source: Austroads Guide to Road Design Part 4A: Unsignalised and Signalised Intersections, Annotated by ATSB

The tests undertaken by Queensland Police showed similar results to the limitations documented in the road design guidelines. That is, vision out the left side passenger window became increasingly difficult as the angle increased to 20° behind 90°. At an angle variation of 20°, a driver in the test vehicle would only be able to sight another vehicle approaching from the left when it was within about 26 m.

Road vehicles are capable of braking relatively quickly. In the context of a road intersection, the driver of an approaching road vehicle is likely to sight another vehicle starting to enter the intersection and take action to avoid an imminent collision.

However, given the size and weight of most trains it is not possible to brake at anywhere near the rate of a road vehicle, nor is it possible to rapidly accelerate or decelerate a train. Therefore, at railway level crossings, it is much more important for the driver of the road vehicle to take action to avoid a collision.

At level crossings protected by Stop signs, motorists are required to stop in order to look for any approaching trains. Consequently, sufficient sighting distance must be available for a motorist to see an approaching train in time to make an informed decision whether it is safe to proceed over the crossing. Australian Standard 1742.7:2016 includes design criteria for calculating sighting distance requirements at level crossings protected by Stop signs. For the configuration at occupation crossing ID 5318, at least 430 m sighting distance is required to provide adequate time for a large road vehicle to accelerate from stop and safely clear the crossing before the arrival of an approaching train. When considering the results of the visibility testing, this could only be achieved if the road vehicle has stopped at an angle greater than 80° to the rail line.

At the accident location, the private road intersected the rail line at an angle of about 90°. However, the junction between the private road and Wagoora-Yalboroo Road (approximately 19 m from the railway crossing signage) showed evidence of widening at its eastern corner. Both roads were unsealed, so widening probably occurred gradually as road vehicles progressively cut the corner when entering/exiting the private road. Consequently, the effective centre line of the road varied from the designed 90° (Figure 8). The ability for a road vehicle to cut the corner and enter the private road at an angle, combined with a relatively short distance to the crossing, may influence the positioning of larger vehicles when stopped at the crossing.

Figure 8: Road angle

The design angle for the road centre line at the level crossing was approximately 90° to the rail line. However, due to widening of the road junction eastern corner, the effective centre line of the road varied from the designed 90°. Source: Qld Police, annotated by ATSB

In this case, the train driver described observing the road vehicle having stopped at the crossing, but it appeared to be angled facing slightly away from the approaching train. Considering both the train driver observation and the geometry of the private road, the angle between the vehicle direction and the railway was very likely less than 90°. Visibility testing suggested that at almost any angle less than 90°, the truck driver’s ability to sight sufficient distance along the rail line would have significantly reduced. However, there is insufficient evidence to determine the exact angle the vehicle stopped, nor what action the truck driver may have taken to check for an approaching train before proceeding over the crossing.

Railway level crossing safety strategy

A 2009 study found there were about 21,800 road/railway level crossings in Australia. Of these, about 8,800 were public road crossings and 13,000 were classified as private/occupation crossings.^[23] While the number of private crossings is relatively high, they are normally provided for the exclusive use of a landowner to access private land. Private crossings are not intended as a thoroughfare for access by the general public, consequently the lower usage of the road lessens the level of risk that may arise from the crossing.

The Office of the National Rail Safety Regulator (ONRSR) publishes a document titled Policy - Railway crossings. The policy focus is on addressing public crossings which are generally considered a greater safety risk to road users. However, it notes that the application of the policy with respect to private crossings should also be considered in relation to risk. That is, regardless of a level crossing being defined as public or private, a rail infrastructure manager is still required to eliminate or minimise the risk that may arise from railway crossings, so far as is reasonably practicable. The policy states that ONRSR expects to see continuous improvement in the safety of railway crossings and ultimately seeks a reduction in the number of railway crossings.

The ONRSR policy refers to a Rail Industry Safety and Standards Board (RISSB) guideline for the consolidation of level crossings.^[24] The guideline acknowledges that managing risk so far as is reasonably practicable may include permanently closing the level crossing to road/pedestrian traffic. The guideline suggests that multiple level crossings in close proximity (within a five kilometre distance or less) present prime opportunities for consolidation.

In July 2012, the Queensland government (Department of Transport and Main Roads) published their Level Crossing Safety Strategy 2012-2021 as a ‘…commitment to further improve safety outcomes for all level crossing users’. The document detailed 12 key strategies, one of which is to ‘…eliminate level crossings where appropriate’.^[25]

There are multiple stakeholders (state/local government and rail/road organisations) involved in managing safety risk at railway level crossings in Queensland. The Queensland Level Crossing Safety Group (QLCSG) provides a forum to bring these stakeholders together to work collaboratively to address the 12 key safety strategies. The QLCSG is chaired by the Department of Transport and Main Roads. Queensland Rail and the infrastructure manager for level crossing ID5318, is a stakeholder and active member of the group.

The QLCSG has documented^[26] a number of achievements since the launch of the safety strategy. While increased education and enforcement has been significant, there has also been progress made with respect to reducing the number of level crossings in Queensland. This has been achieved by closing level crossings where possible, grade separations (overpasses), the closure of branch lines and by not introducing new level crossings onto the network.

It is evident that the strategy and actions to date are generally consistent with the ONRSR policy in that the removal of unnecessary or rarely used public level crossings has been considered.

However, the level crossing safety strategy and the QLCSG report both state that private (occupation) crossings are excluded and that private crossings are considered a workplace health and safety matter. Notwithstanding the exclusion, the QLCSG advised that issues in managing private (occupation) crossings has occurred informally in the past and will continue.

With respect to managing risk, the objective of work health and safety legislation^[27] is largely consistent with that of rail safety legislation.^[28] That is, the duty imposed is to eliminate or minimise the risk to health and safety, so far as is reasonably practicable.

As part of the management of risk at level crossings, Queensland Rail undertake data collection and assessment of all crossings (public and private) on their network. The data and assessments help guide Queensland Rail’s programs for ensuring level crossings (public and private) are maintained in accordance with their documented standards. While crossing ID 5318 provided vehicle access similar to that of a public crossing, the data recorded relatively low risk scores for most elements of the assessment. The ATSB noted that the annual average daily traffic recorded at the crossing was 100 vehicles per day, seasonal. The seasonal notation suggested that the traffic was very likely associated with cane production and less likely associated with general public. However, it was also likely that much of the traffic used the crossing as a thoroughfare between the two public roads, rather than for the sole purpose of accessing a small strip of cane field adjacent the railway track.

While crossing ID 5318 was classified as a private (occupation) crossing, its unrestricted access as a thoroughfare between two public roads likely presented a level of risk similar to that of a public road crossing. Consequently, the strategies for addressing the risk should be consistent for those used as a public thoroughfare whether the crossing is a public or a private (occupation) crossing.

It was reported that the truck driver in this case was travelling between properties located on the Wagoora-Yalboroo Road and another located on the opposite side of the Bruce Highway. It is evident that in this case, crossing ID 5318 was being accessed as a thoroughfare between Wagoora-Yalboroo Road and the Bruce Highway.

There were a number of level crossings in the relative vicinity of crossing ID 5318 (both public and private). Of specific relevance was the Wintons Road level crossing, located about 900 m east. Wintons Road is a sealed road with active traffic control devices (flashing lights) at the level crossing. For the truck in this case, traversing the railway track at the Wintons Road level crossing was a viable and safer alternative for crossing the railway track (Figure 9).

However, road related issues should also be taken into account when considering alternative routes. For example, information provided the truck operator (Fox Hall Harvesting Group) suggested that road signage near the Wintons Road – Bruce Highway road intersection may pose sighting issues for drivers of larger vehicles such as tractors and trucks.

Some cane harvesting vehicles are not permitted to operate on the highway. Consequently, in some cases, the sole access to property may be via a private (occupation) crossing. However, the ATSB notes that this scenario is consistent with the intention of a private (occupation) crossing and its provision should not encourage unrestricted public access as a thoroughfare.

Figure 9: Chosen route (red) and alternative route (orange)

The truck was travelling between properties located on the Wagoora-Yalboroo Road and another located on the opposite side of the Bruce Highway via crossing ID 5318. There was a viable and safer alternative for travelling to the intended destination via Wintons Road with active traffic control devices (flashing lights) at the level crossing. Source: Google Earth, annotated by ATSB

In the absence of measures preventing unrestricted public access and considering the proximity of an alternative crossing with active traffic control devices (flashing lights), it is possible that crossing ID 5318 might present an opportunity for Queensland Rail to consider consolidation as encouraged by documented level crossing safety policies and safety strategies.

The exclusion of private (occupation) crossings from the Queensland level crossing safety strategy potentially removes an opportunity to further improve safety outcomes for all level crossing users across Queensland. This is especially the case for crossing ID 5318 and other private (occupation) crossings with a similar configuration, where unrestricted public access may present a level of risk similar to that of a public road crossing.

__________

21. An undamaged 1983 Scania model P112 series 2, twin-steer truck was not available for examination.
22. Ackerman C. (1989): Vehicle Characteristics (Course notes titled Traffic Engineering Practice edited by Ogden and Bennett for Monash University, Melbourne)
23. RISSB - National Level Crossing Stocktake (2009)
24. RISSB Guideline - Consolidation of public level crossings
25. The strategy compliments the Queensland government’s Road Safety Strategy 2015 – 2021.
26. Queensland Level Crossing Safety Strategy 2012-2021 - 2013-2014 Annual Report. The strategy was also reviewed in 2017 with the intention to update and acknowledge further improvements.
27. Work Health and Safety Act 2011
28. Rail Safety National Law (Queensland)

Findings

From the evidence available, the following findings are made with respect to the collision between freight train 8279 and truck, at level crossing ID 5318 near Yalboroo, Queensland on 29 August 2017. These findings should not be read as apportioning blame or liability to any particular organisation or individual.

Contributing factors

• After stopping at the stop sign, the road vehicle driver then proceeded onto the level crossing and into the path of the train.
• There was insufficient time for the train driver to stop the train before colliding with the truck.
• The combination of the truck’s cab-design and the likelihood that the vehicle stopped at an angle less than 90° to the railway track likely restricted the truck drivers sighting of the approaching train.

Other factors that increased risk

• Some private (occupation) crossings in Queensland, such as crossing ID 5318, provide unrestricted public access between two public roads. These crossings present a level of risk similar to that of a public road crossing where either a higher level of treatment may be applicable with respect to traffic control devices, or closure may be a consideration.
• The exclusion of private (occupation) crossings from the Queensland level crossing safety strategy potentially removes an opportunity to further improve safety outcomes for all level crossing users across Queensland. This is especially the case for crossing configurations where unrestricted public access may present a level of risk similar to that of a public road crossing.
• The combination of the widening of junction at Wagoora-Yalboroo Road with the relatively short distance to the railway crossing could influence the position of road vehicles when stopped at the crossing and in turn, could affect the ability for drivers of some larger vehicles to sight an approaching train.

Other findings

• While there is insufficient evidence to determine conclusively, there are a number of other factors that may have influenced why the truck driver proceeded onto the level crossing:
  • The truck driver looked towards the approaching train, but did not see or perceive its approach (inattentional blindness).
  • The truck driver’s familiarity with the railway crossing and a low expectation of encountering a train contributed to him not noticing the approaching train.
  • The truck driver’s attention may have been directed towards some other task-irrelevant thoughts.
• The traffic control devices installed at occupation crossing ID 5318 were consistent with the requirements documented in the Queensland Rail organisational standard.
• It is unlikely that sun-glare or any other environmental conditions contributed to the collision.
• There were no obvious or apparent mechanical defects to the truck that could have contributed to the collision.
• The locomotive was relatively conspicuous in colour (yellow), its headlights were illuminated and the train driver had sounded the train whistle twice while approaching the crossing.
• For both the truck driver and the train driver, it is unlikely that the effects of fatigue, drugs, alcohol, or any medical condition contributed to the collision.
• There was a viable and safer alternative for the truck to travel to the intended destination. This alternative involved traversing the railway track via a sealed road with active traffic control devices (flashing lights) at the level crossing.

Safety actions

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

Proactive safety action taken by Aurizon

• External communications – Aurizon has ongoing engagement with local authorities (and in this case Mackay Regional Council) prior to sugar seasons, for the purpose of raising awareness of above rail operations during the sugar crushing.
• Internal communications – In addition to the existing ‘pre-start’ and ‘tool box’ talks provided for rail traffic crew, specific information has been shared with rail traffic crew subsequent to this occurrence to reinforce procedures and re-remind rail traffic crew of increased presence of vehicles due to the cane crushing season and to continue to report all occurrences as required and to continue to report near misses as required.

Proactive safety action taken by Queensland Rail

• On the 6 November 2017, Queensland Rail (QR) conducted an audit (site survey) to confirm site distances and the signage has been installed to the QR standard. The result of this audit found that the level crossing had appropriate sighting distance and signage for the controls in place in accordance with QR’s level crossing standard for private crossings. It is assumed that during cane season the crossing is used by the land owner’s trucks and harvesting equipment and it appears the crossing is well maintained.
• Whistle boards are installed at the 428 m from the crossing to the south and 375 m from the north. The audit identified that additional whistle boards should be installed at 250 m either side of the crossing due to the rail line speed of 100 km/h and that the R26 sign has started to fade and should be replaced. These will be addressed by the local asset manager under routine maintenance activities.
• Independently to this incident occurring, QR had already put in place an audit programme to assess all private crossings, upgrade them to the QR standard as necessary, and seek to enter into interface agreements. This programme was progressing at the time of the accident. Although the programme had not yet covered the specific location of this crossing at the time of the accident, QR is actively working towards improving or eliminating private crossings.
• QR is requesting the relevant party (local government or landholder) enter into an appropriate licence or interface agreement with Queensland Rail for the safe operation and use of the private level crossings. Audits have so far been completed within the Far North, North West and Mackay regions of Queensland. To date, the audits have identified eight crossings for closure (as they are no longer required by any party) and eight crossings that will be changed from a private status to a public status crossing as they appear to be used by the general public. One crossing has been identified as requiring relocation. As at May 2018, licence/interface agreements for 85 crossings have been sent to relevant parties in the shire areas noted above and 15 licence/interface agreements have been formalised.
• QRl reviews its level crossing standard on a 2 yearly basis and the current review is due to be completed in 2019. It has been identified in this review that consideration needs to be given to the frequency of audit of private level crossings.

Proactive safety action taken by Fox Hall Harvesting Group

Fox Hall Harvesting Group advised that they are prepared to erect no access signs as well as private property signs on both sides of the access road. They are also prepared to consider lockable gates and a fence on the boundary between the property and the railway easement to stop public use.

ATSB comment:

The ATSB acknowledges the willingness of the property owner to take action preventing public use of the private crossing. However, it is noted that any action should be done in consultation with the rail operator to ensure unintended hazards are not introduced.

Proactive safety action taken by the Queensland Level Crossing Safety Group (QLCSG)

At the QLCSG meeting held on 30 May 2018, the responsibilities under the Rail Safety National Law in relation to private (occupational) level crossings was acknowledged as being the primary means by which to manage safety at these locations. It was agreed that the QLCSG should remain focused on the higher risk public level crossings. However, in recognition of ATSB’s draft finding, it was also agreed that rail infrastructure managers would undertake a one-off review of private level crossings that are being used by the public (similarly to Yalboroo). The intent would be to review their status and treatment, and where assessed, include them as a sub-set of public level crossings for the purposes of the Strategy and QLCSQ functions.

Sources and submissions

Sources of information

The sources of information during the investigation included the:

• Aurizon
• Queensland Rail
• Queensland Police Service
• Office of the national rail safety regulator
• Bureau of Meteorology
• Crew of freight train 8279

References

Australian Government, Geoscience Australia

Australian Level Crossing Assessment Model ALCAM

Australian Standard 1742.7:2016 Manual of uniform traffic control devices, Part 7: Railway crossings

Australian Standard AS 7531:2015 Rolling stock standard – Lighting and Visibility

Austroads Guide to Road Design Part 4A: Unsignalised and Signalised Intersections

Ackerman C. (1989): Vehicle Characteristics (Course notes titled Traffic Engineering Practice edited by Ogden and Bennett for Monash University, Melbourne)

Battelle Memorial Institute (1998). An Overview of the scientific literature concerning fatigue, sleep, and the circadian cycle, Report prepared for the Office of the Chief Scientific and Technical Advisor for Human Factors, US Federal Aviation Administration.

Green, M (2004), Inattentional Blindness & Conspicuity. Retrieved 13 March 2007 from www.visualexpert.com

Green, M. & Senders, J. (2004). Human error in road accidents. Retrieved 13 March 2007 from www.visualexpert.com

Mack, A. & Rock, I. (1998), Inattentional Blindness. MIT Press: Cambridge, USA

National Transport Commission (2008). National Rail Safety Guideline. Management of Fatigue in Rail Safety Workers.

Office of the National Rail Safety Regulator – Policy: Railway crossings

Pennie S. Seibert and Henry C. Ellis, Memory & Cognition (1991), Irrelevant thoughts, emotional mood states, and cognitive task performance, 507-513.

Queensland Department of Transport and Main Roads Road Planning and Design Manual

Queensland Department of Transport and Main Roads Level Crossing Safety Strategy 2012-2021

Queensland Level Crossing Safety Strategy 2012-2021 - 2013-2014 Annual Report

Queensland Rail Interface Standards (MD-10-194, dated 23 July 2014)

Rail Industry Safety and Standards Board - National Level Crossing Stocktake (2009)

Rail Industry Safety and Standards Board Guideline - Consolidation of public level crossings

Rail Safety National Law (Queensland)

Regan, M.A., Hallett, C. & Gordon, C.P. (2011). Driver distraction and driver inattention: Definition, relationship and taxonomy. Accident Analysis and Prevention, 43, 1771-1781.

Wickens C.D. & McCarley, J.S. (2008). Applied Attention Theory. CRC Press: Boca Raton. pp 55-57.

Work Health and Safety Act 2011.

Yeh, M. & Multzer, J. (2008). Driver Behaviour at Highway-Railroad Grade Crossings: A Literature Review from 1990-2006. Human Factors in Railroad Operations. United States Department of Transportation, Federal Railroad Administration: Washington DC.

Young, K.L., Regan, M.A., & Hammer, M. (2003). Driver Distraction: A review of the literature. Monash University Accident Research Centre. Available from: www.monash.edu.au

Submissions

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

A draft of this report was provided to the next of kin, the crew of train 8279, Aurizon, Queensland Rail, Queensland Police Service, the Queensland Department of Transport and Main Roads, the trucking company and the Office of the National Rail Safety Regulator.

Submissions were received from the crew of train 8279, Aurizon, Queensland Rail, Queensland Police Service, the Queensland Department of Transport and Main Roads, the trucking company and the Office of the National Rail Safety Regulator. 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 Published by: Australian Transport Safety Bureau © Commonwealth of Australia 2018 Ownership of intellectual property rights in this publication Unless otherwise noted, copyright (and any other intellectual property rights, if any) in this report publication is owned by the Commonwealth of Australia. Creative Commons licence With the exception of the Coat of Arms, ATSB logo, and photos and graphics in which a third party holds copyright, this publication is licensed under a Creative Commons Attribution 3.0 Australia licence. Creative Commons Attribution 3.0 Australia Licence is a standard form licence agreement that allows you to copy, distribute, transmit and adapt this publication provided that you attribute the work. The ATSB’s preference is that you attribute this publication (and any material sourced from it) using the following wording: Source: Australian Transport Safety Bureau Copyright in material obtained from other agencies, private individuals or organisations, belongs to those agencies, individuals or organisations. Where you wish to use their material, you will need to contact them directly.