On 13 June 2013, at about 1924 Eastern Standard Time, a Bell 412 helicopter, registered VH‑EMZ (EMZ), departed Horn Island, Queensland on a training flight to Prince of Wales Island, Torres Strait. On board the helicopter was the pilot flying (PF) who was under instruction, a training pilot, and a crewman. The PF conducted a practice approach using the ‘Nightsun’ to illuminate the ground below the helicopter.
At 400 ft AGL, and about 0.6 NM from the targeted landing area, the training pilot noted that the indicated airspeed (IAS) was 60 kt with a 500 ft/min rate of descent and the GPS showed a ground speed of 45 kt. The PF looked outside to confirm that the profile and sight picture were correct for a visual approach and adjusted the Nightsun beam onto the landing site. The training pilot also looked out of the cockpit and when he looked back inside the cockpit he observed a high rate of descent of about 800 ft/min and he called ‘rate of descent’. The training pilot also observed that the IAS was below 35 kt and called ‘go around’. The training pilot reported that there was no immediate response so he repeated the call to ‘go around’.
The PF reported commencing a go-around and responded ‘going around’. The training pilot reported that he felt the collective move. The helicopter entered an incipient vortex ring state and impacted the ground heavily. The crew were uninjured, and the helicopter was substantially damaged.
On 4 June 2013, a Boeing 737-800 (737) aircraft, registered VH-YIR and operated by Virgin Australia (Virgin), was on a scheduled passenger service from Melbourne, Victoria to Sydney, New South Wales. During descent into Sydney, the crew was advised by air traffic control (ATC) to expect an independent visual approach (IVA) to runway 16 Right (16R).
As the aircraft approached the extended centre-line of runway 16R, the aircraft’s traffic collision avoidance system (TCAS) provided a traffic alert followed by a resolution advisory (RA) in relation to an Airbus A320 aircraft on approach to parallel runway 16 Left (16L). As the crew commenced descent in response to the RA, the aircraft continued through the extended centreline of runway 16R by about 300 m. When the TCAS alert ceased, the pilot flying captured the extended runway centre-line from the other side. The flight crew continued the approach and landed, whilst the A320 executed a go-around procedure.
As both aircraft were cleared and utilising IVA procedures, the occurrence did not constitute an ATC loss of separation assurance.
What the ATSB found
The ATSB found that the 737 passed through the centreline as a result of the aircraft’s automatic flight control system not being set to the correct flight mode for an intercept and turn onto the runway 16R localiser. This most likely occurred due to insufficient force being applied to the approach mode push-button and, as the flight crew did not perform an effective check of either the mode control panel or the flight mode annunciator to verify a mode change, they were unaware that the aircraft’s flight mode was not set as intended.
The ATSB also found that the risk of an undetected mode selection error was increased as the Virgin procedures did not mandate that flight crew announce flight mode changes.
There were no technical failures of the aircraft, aircraft tracking systems or ground equipment in relation to this occurrence.
What's been done as a result
In response to this occurrence, Virgin introduced a flight policy requiring crews to verbally announce flight mode changes when operating above 500 ft.
Safety message
During an IVA, accurate interception and tracking of the runway centreline is essential to maintain separation with aircraft using the parallel runway. This occurrence highlights the importance of pilots remaining vigilant during this type of approach, including to the consideration of and response to all RAs. The importance of crews conducting comprehensive checks of the mode control panel and flight mode annunciator to ensure that the flight mode selected is consistent with the crew’s intention is also reinforced.
The occurrence
On 4 June 2013, a Boeing 737-800 aircraft (737), registered VH-YIR and operated by Virgin Australia, was on a scheduled passenger service from Melbourne, Victoria to Sydney, New South Wales. The captain was the pilot flying (PF) and the first officer (FO) was the pilot monitoring (PM).[1]
On first contact with an approach controller, flight crews are required to notify receipt of the latest automatic terminal information service[2] broadcast for their destination. At 1845 Eastern Standard Time,[3] the flight crew contacted the Sydney Approach controller for approach clearance and were cleared to descend to 8,000 ft for runway 16 Right (16R).[4] The crew were also advised to expect an independent visual approach (IVA).[5] At that time, the aircraft was 46 NM (85 km) to the southwest of Sydney.
At 1848, the flight crew of an Airbus A320 (A320), registered VH-VFL and operated by Jetstar Airways, contacted the Sydney Approach controller from a location 53 NM (98 km) to the north of Sydney. The approach controller cleared the A320 flight crew to descend to 8,000 ft for runway 16L and advised the crew to expect an IVA. At that time the 737 was 25 NM (46 km) to the southwest of Sydney on descent to a re-cleared 6,000 ft.
The Sydney Director[6] assigned the 737 flight crew a right turn heading 070° and descent to 3,000 ft at 1854:51. Shortly after at 1855:07, the Director instructed the crew to continue on their current heading and advised that there was a chance they may be taken off the approach as there was medical traffic heading north-east from Bankstown. Figure 1 illustrates the relative position and flight paths of the 737 and A320 to Bankstown Airport.
Figure 1: 737 and A320 approach paths for arrival at Sydney Airport
Source: Google earth (modified by the ATSB)
At 1856:10 the Director instructed the 737 flight crew to turn right onto a heading of 125° to join final for runway 16R, and cleared the crew for a visual approach with a requirement to not descend below 2,000 ft until established on the precision approach path indicator (PAPI).[7] The 737 flight crew reported that at that stage of an approach they would normally select the approach mode, which involves the PF pressing the approach push-button on the aircraft’s mode control panel. With this flight mode armed, and given the reliance on the instrument landing system (ILS)[8] at Sydney to conduct IVA approaches (see the section titled Independent visual approaches), the aircraft’s automatic flight control system (AFCS) flies the heading selected by the crew until the aircraft captures the ILS localiser signal. Once captured, the AFCS follows the localiser course, which coincides with the runway extended centre-line in this case. The captain recalled selecting the approach mode using the push-button but stated that the action may not have been sufficiently forceful to enable the selection. The FO also recalled the captain announcing a mode change by stating ‘arming approach’, and considered that the approach pushbutton was pressed. However, recorded data indicated that the captain’s action to select the push-button was not effective. Neither pilot could clearly recall verifying a change on the flight mode annunciator (FMA)[9] to confirm that the approach mode was armed.
Radar data indicated that at 1857:27, the 737 was at 3,100 ft, with a ground speed of 210 kt, on a heading of 124° and 1.7 NM (3.2 km) from the extended centre-line of runway 16R. The 737 was 1.6 NM (3.0 km) from the A320 with 700 ft vertical distance between the aircraft.
At 1857:42, the A320’s traffic collision avoidance system (TCAS)[10] provided a traffic advisory (TA)[11] in relation to the proximity of the 737. The 737 crew received a TCAS TA at about the same time. The FO on the 737 recalled hearing the TA and initially thought it related to the medical traffic referred to earlier by air traffic control (ATC). The FO looked at the aircraft’s navigation display and noted an aircraft behind and to the left of the 737.
At 1857:53, the 737’s TCAS provided a resolution advisory (RA)[12] alert to descend. The captain reported becoming aware of a flight path deviation at about the same time as the RA, and noticed that the aircraft was about to pass through the centre-line for runway 16R. The captain disconnected the autopilot at 1857:57 and manually commenced a descent to comply with the RA.
At the same time, the 737 passed through the extended centre-line for runway 16R and, at 1858:03, the RA changed from ‘descend’ to ‘don’t climb’. At 1858:12 the aircraft reached its maximum deviation (320 m) to the left of the runway 16R extended centre-line.
The A320’s flight crew received a ‘climb’ RA at 1857:54 before advising ATC at 1858:05 of that alert. The 737 crew advised ATC at 1858:10 that they too had received an RA.
At 1858:17 the 737’s TCAS provided a ‘clear of conflict’ annunciation, which the crew conveyed to ATC. The Director asked the 737 crew if they were able to continue the approach, and the crew replied that they could. At 1858:30, the 737 captain armed the approach mode. At 18:58:32 the AFCS captured the localiser and, at 18:59:44, the AFCS captured the glideslope before the aircraft landed at about 1900.
At 1858:26 the A320 crew advised they were clear of the conflict, and were provided instructions to re-sequence for another approach. This entailed a left turn to re-intercept the extended centreline of runway 16L (Figure 1).
As both aircraft were cleared for and using IVA procedures, the occurrence did not constitute an ATC loss of separation assurance.[13]
The aircraft flight paths during final approach and the sequence of recorded TCAS events for the 737 and A320 are shown in Figure 2.
Figure 2: Aircraft flight paths (737 shown in blue (runway 16R) and the A320 in red (runway 16L)) and TCAS events
Source: Google earth (modified by the ATSB)
Flight data for selected parameters associated with approach mode selection and the TCAS advisories from the 737 are illustrated at Figure 3.
The captain held an Air Transport Pilot (Aeroplane) Licence ATP(A)L. The captain’s total aviation experience was about 16,000 hours of which 7,000 hours were on the 737. They had flown the 737 with Virgin Australia (Virgin) for 11 years, with 6 years as pilot in command.
The first officer (FO) held an ATP(A)L and had previously obtained their Commercial Pilot Licence in 2004. The FO had flown with Virgin for about 5 years and had about 2 years’ experience in 737 aircraft.
Crew duty
The captain and FO each commenced duty in Sydney at approximately 0755 and operated a sector from Sydney to Cairns and then from Cairns to Melbourne before the occurrence flight from Melbourne to Sydney. Total duty time[14] for each pilot was 11 hours 43 minutes and the flights were conducted within the same time zone.
Prior to the occurrence flight the captain had 3 days off duty and a 9-day break from flying.
The captain stated that they commuted from Brisbane to Sydney on the previous evening to preposition for the first leg out of Sydney the next morning. The Captain reported experiencing a typically restful sleep the night before the occurrence flight.
The FO was returning to work from a period of simulator training in the days preceding the occurrence. They reported receiving about 7 hours of quality sleep during the night before the occurrence.
Aircraft and operational information
Mode control panel
The aircraft’s automatic flight control system includes an autopilot flight director system (AFDS) and auto-throttle, which are controlled using the AFDS mode control panel (MCP) and the flight management computer (FMC). Normally, the AFDS and auto-throttle are controlled automatically by the FMC to fly an optimised lateral and vertical flight path through climb, cruise and descent.
The MCP facilitates more direct control of the aircraft flight modes and can be used to intervene between the FMC and either of the AFDS and auto-throttle. Pilots may use the MCP to make a range of flight mode selections including autopilot engagement, course and heading, target speeds and altitudes.
The mode selector switches are pushed to select desired command modes for the AFDS and auto-throttle. The switch illuminates to indicate mode selection and that the mode can be deselected by pushing the switch again. While a mode is active, deselection can be automatically inhibited and is indicated by the switch being extinguished.
When engagement of a mode would conflict with current AFDS operation, pushing the mode selector switch has no effect. All AFDS modes can be disengaged either by selecting another command mode or by disengaging the autopilot and turning the flight directors off (Figure 4).
Figure 4: Mode control panel with the location of the approach mode push-button highlighted in red
Source: ATSB
Once the approach mode push-button on the MCP is pressed, the system approach mode is armed. The approach mode becomes active when the localiser and glide slope signals are captured, causing the green light on the MCP approach push-button to extinguish. Once this occurs, the only means to deactivate approach mode include:
selecting take-off/go around (TOGA) thrust
disengaging the autopilots and switching the flight directors off
retuning the very high frequency navigation (VHF NAV) receivers.
Illumination of the approach push-button light indicates that the armed or active mode can be deselected. It is not an accurate indicator of the status of a selected mode. As discussed in the following description, only the flight mode annunciator (FMA) provides a true indication of the mode status.
Flight mode annunciator
As previously stated, auto-flight system mode selections are made using the MCP. They are displayed from left to right on the FMA at the top of each pilots primary flight display (PFD) and include:
auto-throttle
roll (or lateral) mode
pitch (or vertical) mode.
Engaged flight modes are displayed on the top line of the FMA in green letters and are the primary indication to the pilot that an AFDS mode is active. Armed modes are displayed in smaller, white letters beneath the engaged modes. A highlighting rectangle appears around the relevant mode annunciation for a period of 10 seconds following mode engagement. Confirmation of the status of the auto-flight system modes is made by reference to the FMA.
Figure 5 is an example of a B737 PFD for an aircraft on approach to Sydney runway 16R. The representation shows the aircraft close to touchdown with both Localiser and Approach correctly captured by dual autopilots.
Figure 5: B737 PFD with the FMA highlighted in red. The top line of the FMA shows in green text that Mode Control Panel Speed is engaged, VOR or Localiser Tracking is engaged, and Glideslope is engaged and captured. The second, lower line shows in white that Flare is armed
Source: Aerosoft Australia (modified by the ATSB)
Flight mode monitoring
The Virgin Operations Manual provided guidance regarding crew monitoring of indications, limitations and on systematic cooperation and exchange of information between flight crew. The manual also included a requirement to check FMA annunciations that may indicate downgraded capability, and for flight crew to monitor the FMA during low visibility approach and auto-land operations. However, the document did not mandate that flight crews verbally announce FMA changes.
Aircraft and ground-based systems
Under certain conditions, including as a result of ground-based technical factors, the aircraft’s automation may allow the aircraft to fly through the localiser briefly, before intercepting the localiser from the other side. However, the known onboard or ground-based technical factors with the potential to cause these brief incursions did not contribute to this occurrence.
Independent visual approaches
Sydney Airport operates two parallel runways (16L/34R and 16R/34L) that are set 1,037 m apart (Figure 6). Use of the independent visual approach (IVA) procedure, which is only possible in visual meteorological conditions,[15] allows two aircraft to be established on final approach to the parallel runways at the same time.
IVA procedures were introduced with the commissioning of runway 16L/34R in 1994. At the time of writing, Sydney Airport was the only location in Australia with parallel runways and where IVAs were in operation. All runways at Sydney Airport are instrument landing system (ILS) equipped.
Figure 6: Sydney Airport runway configuration
Source: Airservices Australia (modified by the ATSB)
Procedures
At the time of the occurrence, Airservices Australia (Airservices) published the requirements and procedures for IVAs in the Australian Aeronautical Information Publication (AIP). The AIP stated that IVAs could be conducted to parallel runways with centrelines separated by at least 760m, provided that aircraft were making straight-in approaches that commenced at the ILS outer marker or 4 NM (7.4 km) from the runway threshold, and air traffic control (ATC) maintained a minimum 1,000 ft vertical or 3 NM (5.6 km) radar separation between the affected aircraft. These preconditions were stipulated until a number of additional conditions were met and an aircraft was cleared for an IVA. When an aircraft was vectored by ATC to intercept the final course, the final vector had to permit the aircraft to intercept that course at an angle not greater than 30°.
The AIP also contained a section on pilot responsibilities during IVAs. These included flying accurate headings, ensuring that the runway centre-line was not crossed during intercept, accurately tracking the runway centreline, maintaining a visual lookout for aircraft approaching the parallel runway and advising ATC immediately when avoiding action was initiated or contact with the runway was lost.
During an IVA, accurate tracking of the runway centreline is essential. This can be achieved either electronically (localiser, RNAV[16] or ground-based augmentation system (GBAS) landing system[17]) or visually.
The difference between reporting ‘visual’[18] and reporting the runway in sight is significant. A visual call by the flight crew to ATC is all that is required if established on the centreline using electronic indications. In contrast, a specific runway in sight call is required from flight crews if they are not electronically established on centreline (Figure 7).
Any aircraft being processed for an IVA needs to inform the Sydney ATC Director[19] when they have the runway in sight as soon as possible.
A local ATC restriction was in place that did not allow foreign-based airlines (excluding New Zealand-based operators) to participate in IVAs until their aircraft was established on the runway centreline. In addition, Qantas Airways Limited Boeing 747 aircraft were treated by ATC as a foreign international operator, due to a limitation imposed by the airline.
At Sydney Airport, the ILS critical areas[20] were not protected during IVAs. Airservices reported that aircraft operating on IVAs to runways 16L and 16R were spaced 4 NM (7.4 km) apart in trail for each runway in order to optimise the efficiency of landing rate. If the ILS critical areas were to be protected, this spacing would need to be increased to 7 NM (13 km). Together with the requirement for ground traffic to be positioned outside the ILS critical areas, leading to congestion and towing and taxiing restrictions in the remaining aircraft ground movement areas, this would result in a 39 per cent reduction in the arrival rate into Sydney per hour.
Design and implementation
The design and implementation of IVAs at Sydney was a collaborative undertaking involving industry, the Civil Aviation Safety Authority (CASA) and Airservices.
Due to the different runway lengths and airport layout in Sydney, extensive analysis was conducted by Airservices on the operation of a parallel runway (16L/34R) prior to its construction and commissioning in 1994. A design group was formed to investigate different methods of achieving an efficient mixed-mode operation, where both runways are used for arrivals and departures. The group focused on airports with a distance between their parallel runways similar to that of Sydney.
Minneapolis St Paul and Raleigh Durham Airports, which are both located in the United States (US) and operated under Federal Aviation Administration (FAA) regulations, were selected as possible models given the distance between their parallel runways was similar to Sydney. As those airports had accumulated years of experience with mixed-mode parallel runway operations, it was determined the most efficient approach would be to adopt a similar mixed-mode operation at Sydney Airport.
Conservative restrictions were applied to the baseline FAA model to develop the IVA procedures for application at Sydney Airport. These restrictions reflected the local conditions at Sydney Airport and the then inexperience in high-capacity parallel runway operations in Australia.
Implementation of IVAs at Sydney was managed by Airservices. With support from industry and CASA, this included an information/education program involving industry meetings and presentations, the development of a training video outlining pilot and ATC responsibilities and the production of operational documentation.
Reviews of IVA operations at Sydney Airport prior to the occurrence
Civil Aviation Safety Authority
Civil Aviation Safety Regulation 1998 (CASR) Part 172 Manual of Standards - Air Traffic Services specifies the regulatory framework for the approval of air traffic service providers in Australia, and includes standards for air traffic facilities, safety management and the provision of air traffic services. In January 2013, CASA published the results of a review of its CASR Part 172 approval of Airservices. The review sought to identify areas for improvement of Airservices safety and service performance and to determine if conditions needed to be applied to the Airservices approval.
CASA’s review determined that there was a high instance of traffic alert and collision avoidance system (TCAS) alerts at Sydney involving IVAs, and that the IVA operation was implicated in TCAS incident reports more frequently than any other event. These included traffic advisory (TA) and resolution advisory (RA) occurrences. In addition, CASA noted a number of changes in operating dynamics since the introduction of IVA procedures, such as aircraft utilising the TCAS RA mode instead of the original TCAS TA mode once used for arrivals into Sydney.
CASA also found that there were frequent occasions during the conduct of IVAs where controllers issued visual approach clearances from the downwind or base legs that resulted in aircraft intercepting adjacent, parallel final approach paths. Aircraft would then fly either ‘side-by-side’ or one overtake the other during final approach. It was also noted that ATC standards did not require controllers to provide positive separation between aircraft on adjacent final approaches, nor was ATC required to provide traffic information to flight crews of aircraft vectored into close proximity.
One of the recommendations from the CASA review was that, in respect of IVA operations at Sydney Airport:
Airservices should review Independent Visual Approach (IVA) procedures at Sydney in order to determine if sufficient risk mitigators are in place to ensure the continued safe operation of aircraft.
Airservices Australia
Airservices conducted a review of parallel runway operations at Sydney Airport in 2003. The review included an evaluation of the IVA procedures at Sydney and recommended a number of improvements.
Subsequently, Airservices became aware of a potential recommendation from CASA’s CASR Part 172 review in respect of IVA procedures at Sydney Airport. Airservices advised CASA that they would review the procedures in conjunction with industry and implement any agreed actions.
On 14 November 2012 Airservices revised the wording on the Sydney Automatic Terminal Information Service (ATIS)[21] in respect of IVA approaches to advise pilots to ‘not pass through the assigned runway centreline’. This revision of the Sydney ATIS was in place at the time of the occurrence on 4 June 2013.
On 20 November 2012, following a discussion with airlines, Airservices published a safety bulletin on IVAs[22] noting that their reporting system had identified the inconsistent application by some pilots of their responsibilities during IVAs. The following five pilot responsibilities were reiterated in the safety bulletin:
Pilots ensuring that the runway centreline is not crossed during intercept (Figure 8)
Pilots maintaining a visual lookout for aircraft approaching the adjacent parallel runway centreline
TCAS RAs during IVAs
Importance of “Visual” reports and the specific requirements for “specific Runway XX Right or left in sight”
Pilots understanding of radio failure procedure as described in ERSA (En Route Supplement Australia)
In addition to the safety bulletin, Airservices reported that they planned to review the IVA procedures, with airline input, in the first quarter of 2013.
The Airservices review of IVA procedures was attended by CASA and airline and Airservices representatives. It identified a number of potential risk mitigation factors and considered the release of an order by the US FAA for ATC to review approach speeds to reduce potential overshoot situations during parallel runway operations.
On 24 November 2013, standard circuit speeds for application by Sydney ATC during IVA procedures were introduced. On 12 December 2013, a change was introduced to the phraseology used by ATC to clear an aircraft for an IVA to increase pilot awareness of the IVA procedure, given pilot responsibilities ‘change slightly between a visual approach and a visual approach as part of the IVA procedure’. This required controllers to include the word ‘independent’ when clearing an aircraft for a visual approach during IVA procedures.
Related occurrences
A review of the ATSB and Airservices occurrence databases was undertaken to examine the number of TCAS RA occurrences during IVA operations at Sydney Airport in the period 20082013 (Figure 8). Based on the available information, in that period there were 277 reported RAs during IVA procedures.
Figure 9: TCAS RA occurrences during IVA operations at Sydney airport (2008–2013)
Source: ATSB
Airservices advised that over the period 2008–2013, there were about 154,000 approaches per year at Sydney Airport, and that about 60 to 85 per cent involved IVAs. Given this broad range, it was not possible to determine whether there were any trends in the rate of RA occurrences (per approach) in recent years. The ATSB was unable to compare this alert rate against the US airports using similar procedures because there were more limited requirements for reporting TCAS RA occurrences in the US.
A review of the ATSB database indicated that about 30 per cent of all reported TCAS RA occurrences in Australia, or involving Australian aircraft overseas, occurred at Sydney Airport. Data provided by Airservices indicated that about 62 per cent of all RAs in the Sydney terminal area occurred during IVA procedures. The other 38 per cent included other approaches, departures and overflying aircraft.
In certain situations, two aircraft can be flown on IVAs at Sydney in a manner consistent with the required procedures, yet both aircraft’s TCAS can generate an RA. For example, of the 277 reported RA occurrences during IVAs at Sydney during 2008–2013, 30 involved one of the aircraft passing through the extended centre-line. For many of the other occurrences there was insufficient information to determine whether a deviation occurred.
The extent to which centre-line deviations occurred during IVAs when no RA was reported could not be reliably determined as there were no specific reporting requirements for such events.
While conducting an independent visual approach (IVA) to runway 16R at Sydney Airport, the flight crew of a Boeing 737 (737), registered VH-YIR and operated by Virgin Australia (Virgin), passed through the runway centreline as an Airbus A320 aircraft, registered VH-VFL, was conducting an IVA to parallel runway 16L. Both flight crews received a traffic advisory (TA) followed by a resolution advisory (RA) through their aircraft’s traffic collision avoidance systems (TCAS) and acted in accordance with their respective RA instructions. At that time, the 737 captain realised they were passing through the extended runway 16R centreline and disengaged the autopilot before manually flying the aircraft back towards the centre-line. The flight crew then armed the approach mode, which captured the localiser course. The flight crew of the A320 performed a go-around, and both aircraft landed safely. Although the occurrence was far from developing into a collision, and both aircraft were in visual meteorological conditions, this analysis examines the crew actions and IVA procedures at Sydney Airport and their influence on the development of the occurrence.
Arming the approach and mode awareness
The mode control panel (MCP) is the interface between the flight crew and the aircraft’s automatic flight control system (AFCS). Pressing the approach mode pushbutton on the MCP illuminates the integral pushbutton light, indicating that the crew have ‘issued an instruction’ to the AFCS. The flight mode annunciator (FMA) then displays a message as a means of feedback from the AFCS to the crew confirming either that a mode is armed or active. Checking both the MCP and the FMA is therefore critical to the flight crew’s continued awareness of the aircraft’s mode status.
Checking the FMA
Although it seems logical that a pilot would check the FMA, research has shown that they do not always do so, even when flight crews are required to call out auto-flight mode changes (Björklund and others 2006). Other research has suggested that 32 per cent of pilots do not observe the FMA within the first 20 seconds of a manually-selected mode change (Mumaw and others 2001).
Björklund and others (2006) state:
Two out of five mode transitions on the FMA were never “seen” by the flight crews. In contrast to instrument monitoring in non-glass-cockpit aircraft, monitoring for mode transitions is likely to be based more on a pilot’s mental model of the automation that drives expectations of where and when to look. Such models are often incomplete and buggy (e.g., Sarter, 1995). Therefore, it may not be surprising that many mode transitions in this study are neither visually nor verbally verified by flight crews, and that the FMA triggered only 4% of call-outs in this study, of which one out of four was not the official call-out. The FMA did not get consulted for 40% of all mode transitions.
Goteman and Dekker (2006) also found that FMA call-outs were shed when pilots were under a higher task load. It was also found that compliance with the required call-outs was higher when the operator only required a selected set of mode changes to be called.
Mode awareness
In an effort to increase pilots’ mode awareness, aircraft manufacturers have recommended that pilots call out mode (or FMA) changes. Many operators have incorporated this into their operating procedures.
Ineffective auto-flight system mode awareness has been identified as a contributing factor in many occurrences since the introduction of complex auto-flight systems (Federal Aviation Administration 1996). Attempts to address the problem have only been partially successful, as a 2013 report into operations of flight path management systems (Flight Deck Automation Working Group 2013) stated:[23]
The 1996 FAA report[[24]] identified insufficient autoflight mode awareness as an important vulnerability area. Since that [the 1996] report was published, some changes to flight deck equipment design have been made in new aircraft to address this vulnerability area (e.g., only showing selected target values or modes on the PFD [Primary Flight Display], to foster the pilots reviewing the information on the mode annunciator display rather than on the mode selection panel).
In addition, the issue has been addressed in training through increased emphasis on mode awareness and in some operators’ flight crew procedures by having the pilots call out all mode changes. However, other operators find this use of callouts to be too burdensome and a potential distraction.
These mitigations are only partially successful. The data analysis reveals that autoflight mode selection, awareness and understanding continue to be common vulnerabilities.
In this occurrence, it was found that, contrary to their intentions, the 737 flight crew did not arm the approach mode, probably as a result of applying insufficient force to the mode push-button. Subsequently, the crew’s check/confirmation, if any, of the mode change using either the MCP or FMA was ineffective. This meant the crew’s belief of their selection of the approach flight mode led to an incorrect expectation that the aircraft would automatically capture the localiser.
Virgin did not mandate that its crew announce mode changes at the time of the occurrence. Despite the absence of this procedural requirement, the 737 captain made a verbal call announcing their selection of the approach mode. However, the flight crew did not effectively verify via the FMA display that approach mode was armed.
Flight path monitoring
As the 737 turned on final to intercept the extended runway 16R centre-line, the flight crew expected that the aircraft’s autopilot system would capture the localiser via the automated approach mode. As the flight continued, the flight crew did not anticipate that the aircraft was about to fly through the centre-line.
About 16 seconds prior to passing through the centre-line, the TA provided by the aircraft’s TCAS was the first indication of a potential problem with the aircraft’s flight path. The first officer (FO), who was the pilot monitoring, reported hearing the TCAS TA and that they considered it may have been due to the previously advised medical traffic departing Bankstown. Although the captain, who was the pilot flying, described the medical traffic as a minor distraction, the FO remained concerned about its location. This distraction probably reduced the crew’s ability to fully appreciate the location of their own aircraft and anticipate its future position relative to the runway centreline.
Flight crew response to the TCAS advisories
Automated conflict detection and alerting systems such as TCAS form a fundamental layer of defence against collision through their traffic and resolution advisory functionality. In this occurrence, the TCAS system performed as expected and provided the necessary stimulus for the crew of the 737 to disengage the autopilot and manually fly the aircraft back toward the runway16R centre-line.
The FO recalled the TCAS TA and, although immediately thinking it must have been a consequence of the medical traffic, decided to refer to the TCAS display. This display indicated the traffic as an A320 to the left-rear of the 737. The captain was previously aware of the A320 and acknowledged the value of TCAS as a situation awareness tool.
Twelve seconds passed between the TCAS TA and RA, with recorded data verifying that the 737 flight crew responded by disengaging the autopilot within 4 seconds of the RA. Recorded data also showed that, during the initial 12-second period, the captain described observing a number of unexpected external visual cues, until such time as the aircraft continued through the runway 16R centreline. There was no obvious action taken to rectify the aircraft’s flight path during the period between the TA and the RA.
Although Virgin’s Operations Manual stated that all TCAS warnings were to be treated as genuine and action taken in response, there have been anecdotal reports among a number of operators of TCAS TAs being considered of nuisance value during IVAs at Sydney. This may be due to the technical limitations of TCAS, leading to a ‘cry wolf’ effect (Wickens and others 2009), or reduced trust and use of the system during IVAs due to false alarms associated with the inherent parallel runway proximity to other aircraft.
There was no irrefutable evidence to suggest that pilots ignore TCAS TAs or RAs during IVA procedures at Sydney.
Fatigue
Thomas and Ferguson (2010) examined the effects of different amounts of sleep on the performance of Australian airline flight crews. The study found that the average amount of sleep in the previous 24 hours for captains was 7.0 hours and 7.1 hours for FOs. The occurrence of crew errors was higher, and threat management poorer, during flights when the crew included a captain with less than 6 hours sleep or a FO with less than 5 hours sleep in the previous 24 hours.
It is generally agreed that most people need at least 7 to 8 hours of sleep each day to achieve maximum levels of alertness and performance. Both pilots reported receiving an adequate amount of rest which was conducive to the vigilance, reaction times and cognitive requirements of the flight.
Although both of the crew reported obtaining an adequate amount of quality sleep the previous night, the occurrence took place during the last leg of an 11 hour 43 minute duty period. Although consistent with the captain’s report of experiencing some degree of tiredness during the final leg, it could not be concluded that this equated to a level of fatigue likely to affect the crew’s performance of the task; in particular, their ability to make and verify flight mode selections.
Independent visual approaches
As the crew approached the extended centreline of runway 16R, the captain was aware of an A320 on approach to runway 16L. The captain reported that it was only from using the TCAS that they gained an accurate understanding of the A320’s position and altitude. This may be due, in part, because ATC did not, and was not required to provide, traffic information to aircraft using adjacent runways and abeam each other during independent visual approach procedures at Sydney. Traffic information regarding the A320 may have assisted the crew of the 737 form an accurate traffic picture at a critical point during the final approach phase.
Relative to other approaches and flight operations, there is a relatively high rate of RAs during IVA procedures at Sydney. This reinforces the need for compliance with the IVA procedures into Sydney and for flight crew to assure themselves that the correct flight mode has been selected. Confirmation of the correct flight mode will allow crews to focus on additional aspects of the approach, including surrounding traffic and flight path monitoring during the critical stages of flight.
From the evidence available, the following findings are made with respect to the flight path management occurrence involving Boeing 737, registered VH-YIR and operated by Virgin Australia, which occurred at Sydney Airport, New South Wales on 9 June 2013. These findings should not be read as apportioning blame or liability to any particular organisation or individual.
Safety issues, or system problems, are highlighted in bold to emphasise their importance. A safety issue is an event or condition that increases safety risk and (a) can reasonably be regarded as having the potential to adversely affect the safety of future operations, and (b) is a characteristic of an organisation or a system, rather than a characteristic of a specific individual, or characteristic of an operating environment at a specific point in time.
Contributing factors
The flight crew applied insufficient force to the approach mode push-button on the mode control panel to arm the mode, which was not identified during their subsequent check of the flight mode annunciator.
Due to their expectation that the aircraft approach mode was armed, and a subsequent degradation in monitoring the aircraft’s systems, the flight crew did not anticipate that the aircraft’s automated systems would not capture the runway 16R localiser and did not immediately detect the flight path deviation.
The flight crew did not recognise that the traffic advisory alert provided by the aircraft’s traffic collision avoidance system 12 seconds prior to the aircraft passing through the runway centreline was a potential indication of a problem with the aircraft’s flight path.
Other factors that increased risk
The Virgin Australia procedures did not require its flight crew to, whenever practicable, announce flight mode changes. [Safety issue]
Air traffic control did not, and was not required to provide traffic information to aircraft using adjacent runways and abeam each other during independent visual approach procedures at Sydney. [Safety issue]
Other findings
Relative to other approaches or flight operations, there is a relatively high rate of traffic collision avoidance system resolution advisory alerts during independent visual approaches at Sydney Airport.
Safety issues and actions
The safety issues identified during this investigation are listed in the Findings and Safety issues and actions sections of this report. The Australian Transport Safety Bureau (ATSB) expects that all safety issues identified by the investigation should be addressed by the relevant organisation(s). In addressing those issues, the ATSB prefers to encourage relevant organisation(s) to proactively initiate safety action, rather than to issue formal safety recommendations or safety advisory notices.
All of the directly involved parties were provided with a draft report and invited to provide submissions. As part of that process, each organisation was asked to communicate what safety actions, if any, they had carried out or were planning to carry out in relation to each safety issue relevant to their organisation.
Where relevant, these safety issues and actions will be updated on the ATSB website as information comes to hand. The initial public version of these safety issues and actions are in PDF on the ATSB website.
Annunciation of mode changes
The Virgin Australia procedures did not require its flight crews to, whenever practicable, announce flight mode changes.
Air traffic control did not, and was not required to provide traffic information to aircraft using adjacent runways and abeam each other during independent visual approach procedures at Sydney.
The sources of information during the investigation included:
Virgin Australia (Virgin) and Jetstar Airways
the 737 flight crew
Airservices Australia (Airservices)
the Manual of Air Traffic Services
References
Björklund, CM Alfredson, J & Dekker, SWA 2006, ‘Mode monitoring and call-outs: An eyetracking study of two-crew automated flight deck operations’, The International Journal of AviationPsychology, vol. 16, pp. 257–269.
Civil Aviation Safety Authority 2013, Review of CASR Part 172Air Traffic Service approval of Airservices Australia.
Federal Aviation Administration 1996, Human Factors Team Report on The Interfaces Between Flightcrews and Modern Flight Deck Systems.
Goteman, O & Dekker, S 2006, Flight crew callouts and aircraft automation modes: An observational study of task shedding, International Journal of Applied Aviation Studies, vol. 6, pp. 235–248.
Mumaw, RJ Sarter, NB and Wickens, CD 2001, ‘Analysis of pilots’ monitoring and performance on an automated flight deck’.Proceedings of the Eleventh International Symposium in Aviation Psychology,p. 6.
Performance-based operations Aviation Rulemaking Committee/Commercial Aviation Safety Team Flight Deck Automation Working Group 2013, Operational use of flight path management systems. Available from www.faa.gov.
Thomas, MJW & Ferguson, SA 2010, ‘Prior sleep, prior wake, and crew performance during normal flight operations’, Aviation, Space, and Environmental Medicine, vol. 81, pp. 665–670.
Wickens, CD Rice, S Keller, D, Hutchins, Hughes, J & Clayton, K 2009, ‘False alerts in air traffic control alerting system: Is there a “cry wolf” effect’, Human Factors, vol. 51, pp. 446–462.
Submissions
Under Part 4, Division 2 (Investigation Reports), Section 26 of the Transport Safety Investigation Act 2003 (the Act), the ATSB may provide a draft report, on a confidential basis, to any person whom the ATSB considers appropriate. Section 26 (1) (a) of the Act allows a person receiving a draft report to make submissions to the ATSB about the draft report.
A draft of this report was provided to the flight crew, Virgin, Airservices, Jetstar and the Civil Aviation Safety Authority.
Submissions were received from Virgin and Airservices. The submissions were reviewed and where considered appropriate, the text of the report was amended accordingly.
Purpose of safety investigations
The objective of a safety investigation is to enhance transport safety. This is done through:
identifying safety issues and facilitating safety action to address those issues
providing information about occurrences and their associated safety factors to facilitate learning within the transport industry.
It is not a function of the ATSB to apportion blame or provide a means for determining liability. At the same time, an investigation report must include factual material of sufficient weight to support the analysis and findings. At all times the ATSB endeavours to balance the use of material that could imply adverse comment with the need to properly explain what happened, and why, in a fair and unbiased manner. The ATSB does not investigate for the purpose of taking administrative, regulatory or criminal action.
Terminology
An explanation of terminology used in ATSB investigation reports is available here. This includes terms such as occurrence, contributing factor, other factor that increased risk, and safety issue.
Publishing information
Released in accordance with section 25 of the Transport Safety Investigation Act 2003
Ownership of intellectual property rights in this publication
Unless otherwise noted, copyright (and any other intellectual property rights, if any) in this report publication is owned by the Commonwealth of Australia.
Creative Commons licence
With the exception of the Coat of Arms, ATSB logo, and photos and graphics in which a third party holds copyright, this publication is licensed under a Creative Commons Attribution 3.0 Australia licence.
Creative Commons Attribution 3.0 Australia Licence is a standard form licence agreement that allows you to copy, distribute, transmit and adapt this publication provided that you attribute the work.
The ATSB’s preference is that you attribute this publication (and any material sourced from it) using the following wording: Source: Australian Transport Safety Bureau
Copyright in material obtained from other agencies, private individuals or organisations, belongs to those agencies, individuals or organisations. Where you wish to use their material, you will need to contact them directly.
On 4 June 2013, track workers were preparing a road-rail vehicle to travel to a worksite near Rinadeena Station on the West Coast Wilderness Railway, Tasmania, when the vehicle unexpectedly started to roll backwards down a 1:20 grade. The driver was unable to slow the vehicle, so he and the passenger jumped clear, sustaining minor injuries.
The now unmanned out-of-control vehicle continued to accelerate down the steep grade, heading towards a second road-rail vehicle containing four track workers. Two passengers of the second vehicle jumped clear, sustaining minor injuries, but a third passenger and the driver were still inside when the unmanned road-rail vehicle collided with theirs.
The passenger sustained minor injuries but the driver was trapped and seriously injured in the collision. He was subsequently removed from the vehicle and air lifted to hospital. Both road-rail vehicles were extensively damaged.
What the ATSB found
The ATSB found that the vehicle’s rear road tyres were lifted from the track to examine a suspected problem with the rear rail guidance wheels. As a result, the braking force provided by the rear road wheels was lost and the vehicle began to roll down the incline. The rail guidance wheel electric brake controller had not yet been set correctly and, as a result, little braking effort was applied through the rail guidance wheels.
The investigation also found that the West Coast Wilderness Railway had not considered all of the risks associated with the operation of road-rail vehicles on the steep railway. As a result, documented operational procedures had not been developed and locations where vehicles could be safely on and off railed had not been defined.
Other findings related to the effectiveness of the training provided to the road-rail vehicle operators, road-rail vehicle procurement standards, acceptance testing and commissioning of road-rail vehicles, reliability of the very high frequency (VHF) radio network in the Rinadeena area and radio communication protocols.
What's been done as a result
West Coast Wilderness Railway has reviewed its risk register and implemented operational procedures covering the safe operation of road-rail vehicles on the network. This has led to the development and implementation of an updated training package, procurement specifications and documented on/off tracking points. The company has also taken action to improve radio reception in the Rinadeena area and to ensure reliable communications at the station.
Safety message
All organisations operating road-rail vehicles should consider the risks associated with operating the vehicles on their networks. Information on the risk of operating road-rail vehicles can be found at the Office of National Rail Safety Regulator website.
Preliminary report
Report release date: 15/08/2013
This preliminary report details factual information established in the investigation’s early evidence collection phase and has been prepared to provide timely information to the industry and public. Preliminary reports contain no analysis or findings, which will be detailed in the investigation’s final report. The information contained in this preliminary report is released in accordance with section 25 of the Transport Safety Investigation Act 2003.
Safety summary
What happened
On 4 June 2013, track workers were preparing a road rail vehicle to travel to a worksite near Rinadeena Station on the West Coast Wilderness Railway, Tasmania, when the vehicle unexpectedly started to roll backwards down a 1:20 grade. The driver was unable to slow the vehicle, so he and the passenger jumped clear, sustaining minor injuries.
The now unmanned out-of-control vehicle continued to accelerate down the steep grade, heading towards a second road rail vehicle containing four track workers. Two passengers of the second vehicle jumped clear, sustaining minor injuries, but a third passenger and the driver were still inside when the unmanned road rail vehicle collided with theirs. The passenger sustained minor injuries but the driver was trapped and seriously injured in the collision. He was subsequently removed from the vehicle and air lifted to hospital. Both road rail vehicles were extensively damaged.
What the ATSB has found so far
The preliminary ATSB investigation has found that the vehicle’s rear road-going tyres were lifted from the track during an inspection of the rear rail guidance wheels. As a result, the braking force provided by them was lost and the vehicle began to roll down the grade.
The investigation has also found that West Coast Wilderness Railway had not considered all of the risks associated with operating road rail vehicles on the steep railway.
What's been done as a result
Immediately after the incident, West Coast Wilderness Railway suspended all road rail operations and initiated a full review of its management of the safety of these vehicles. This has led to the development of updated procedures and training and a review of suitable on/off tracking points.
Safety message
All rail organisations operating road rail equipment should consider the advice in safety alert RISN Number 7/2012 Risk associated with Hirail Operations (issued by the Department of Infrastructure, Energy & Resources, Tasmania) and review their management of the risks associated with these operations.
Section 21 (2) of the Transport Safety Investigation Act 2003 (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.
On 12 June 2013, the ATSB commenced an investigation into an occurrence involving a Traffic Alert and Collision Avoidance System (TCAS) Resolution Advisory received by the flight crew of a Boeing 737-800, registered VH-VUX, during an independent visual approach to runway 34R at Sydney Airport, New South Wales on 10 June 2013. The second aircraft involved in the resolution advisory, another Boeing 737, registered ZK-PBF, was on approach to runway 34L.
Examination of the information collected during the initial stages of the investigation identified that the flight crews in both aircraft complied with all air traffic control requirements and there was no loss of separation.
The ATSB is investigating an earlier occurrence involving independent visual approaches at Sydney Airport that occurred on 4 June 2013 (investigation AO-2013-095). This will include a review of other occurrences associated with these types of approaches, including that at Sydney on 12 June 2013. Any safety issues associated with independent visual approaches at Sydney can be expected to be identified as a result of investigation AO-2013-095.
Accordingly, the ATSB has decided to discontinue investigation AO-2013-098.
On 8 June 2013, the pilot of a Bell LongRanger helicopter, registered VH-RHF, was conducting an aerial survey flight with four passengers in the Buccaneer Archipelago area north of Derby, Western Australia. The helicopter was being flown at about 1,000 ft to a planned fuel stop on an island in Cone Bay and was over water when the engine flamed out. The pilot entered autorotation to glide towards land but was unable to reach it. During the glide the pilot deployed the helicopter’s pop-out floats in preparation for an emergency ditching. Shortly after touchdown the helicopter rolled inverted. The pilot and the four passengers exited without injury. A boat crew observed the emergency landing and rescued the occupants from on top of the upturned floating helicopter.
What the ATSB found
The ATSB found that, without the pilot realising, the fuel on board was probably sufficiently low to allow momentary un-porting of the fuel boost pumps, which interrupted the flow of fuel to the engine, resulting in an engine flame-out and ditching. Contributing to the pilot’s lack of awareness of the fuel state was a likely malfunction of the helicopter’s fuel quantity indicating system and a faulty low fuel caution system. In addition, the operator’s fuel management system was almost totally reliant on the fuel quantity indicating system and as a consequence, lacked a high level of assurance.
The ATSB also found that the guidance provided by the Civil Aviation Safety Authority in relation to pre-flight crosschecking of fuel on board allowed for a reliance on aircraft fuel quantity indicating systems without reference to independent sources of fuel quantity information.
What has been done as a result
The helicopter operator advised that as a result of this occurrence they have redesigned their fuel tracking form to improve usability. In addition, the operator is considering the fitment of a fuel totaliser to their LongRanger helicopter types.
Safety message
As shown by this and other occurrences, there is a need for operators to ensure that their fuel management policy and procedures provide for at least two independent and reliable means of establishing fuel on board. These should be supplemented by criteria for identifying, recording, and resolving any discrepancy between the amounts generated by the different methods. In situations where visual or other direct means of establishing fuel quantity are not possible, equipment that measures and totalises fuel flow can provide a valid basis for derivation of fuel on board.
Low fuel level caution systems are valuable elements in a safe fuel management framework but can fail without detection and should not be relied upon as a substitute for an independent crosscheck of fuel quantity indicating systems. Operators should consider the criticality of on-board fuel quantity measuring equipment in the context of their particular operations and manage the risk of malfunction accordingly.
On 6 June 2013, the pilot in command and student pilot of a Cessna 172 aircraft, registered VH‑WYG (WYG), were conducting a training navigational flight from Bankstown to Goulburn, New South Wales. At 1325 Eastern Standard Time, WYG was cleared to take off on runway 29 Right (29R). Immediately after this, a Cessna 185 aircraft, registered VH‑OZX (OZX), was cleared to line up on runway 29 Centre (29C). There was also active circuit traffic on runway 29 Left (29L).
The pilot of OZX was conducting a ferry flight from Bankstown to Moruya and was cleared for take-off on 29C at 1326, just as WYG became airborne on runway 29R. Thirty seconds after this, OZX was given WYG as traffic.
As OZX climbed, the pilot lost sight of WYG. When the tower controller asked whether he still had the Cessna 172 in sight, the pilot of OZX replied in the negative and commenced looking for it to his right. In looking right, the pilot believed he may have rolled the aircraft to the right. OZX crossed over and above WYG, which was maintaining a track slightly to the north of the extended centreline of runway 29R.
The pilot of OZX reported that he had already reached the departure altitude of 1,000 ft crossing the upwind threshold prior to losing sight of WYG. WYG appeared below him and to his left.
WYG was climbing through about 700 to 800 ft above ground level when the PIC sighted OZX above the right wing. He estimated that OZX was then about 30 ft above WYG and 15 m to his right.
In Class D airspace, the onus is on the pilots of visual flight rules (VFR) aircraft to maintain their own separation. It is important to keep other aircraft in sight at all times, irrespective of the aircraft performance.
Section 21 (2) of the Transport Safety Investigation Act 2003 (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.
On 7 June 2013, the ATSB commenced an investigation into a ground proximity warning system (GPWS) event involving a Hawker Beechcraft B200C, registered VH-LTQ, near Strahan aerodrome, Tasmania on 3 June 2013.
Examination of the information collected during the investigation determined that the GPWS warnings received by the pilot were false. The operator advised the ATSB that they had identified that the aircraft’s terrain awareness and warning system (TAWS) database did not include Strahan aerodrome resulting in the false warnings and have put in place measures to prevent its reoccurrence.
Accordingly, the ATSB decided that there was limited potential to enhance transport safety by continuing this investigation, and has elected to discontinue it.
The pilot declared a "Mayday", approximately 15 NM from runway 15 at Cairns. He advised "I think I've run out of fuel, going in". The aircraft ditched a short time later, 2 NM off the coast of Wangetti Beach.
All 5 persons on board were rescued safely.
FACTUAL INFORMATION
History of the flight
The pilot was to fly a normally aspirated PA23-250 (Aztec) aircraft from Cairns to Horn Island and to return in VH-CAR, a turbo-charged Aztec. Three non-paying passengers were to be carried on each flight. The flight from Cairns to Horn Island was conducted on 2 November and the return flight was planned for the afternoon of 3 November. The senior pilot at Horn Island tasked the pilot to divert to Bamaga to pick up a fare-paying passenger and transport him to Cairns. At that time the senior pilot explained to the pilot the changes to aircraft equipment requirements and other aspects related to the change of category from private to charter.
Prior to departure from Horn Island, the pilot obtained a briefing on some aspects of the operation of the aircraft, as this was the first time he had been tasked to fly the turbo-charged variant of the Aztec. The engine turbo system involved differences in engine handling, fuel flow and aircraft performance. The pilot refuelled the aircraft to full tanks. The aircraft departed from Horn Island at 1508 and after boarding the passenger at Bamaga, departed there at 1543. Fuel was drawn initially from the outboard tank in each wing. Both fuel gauge readings initially reduced evenly. The aircraft was flown at 9,000 ft with a ground speed of 135 kts until about Princess Charlotte Bay, when the ground speed increased to 140 kts. During the flight the pilot monitored all tank caps for evidence of leaking, as was his normal practice. No leaks were observed.
The aircraft was equipped with four fuel tanks, one located inboard and one outboard in each wing. Fuel can be drawn from any of the four tanks to provide fuel to either engine by the use of two three-position selectors (inboard, outboard and off) and a cross-feed selector in the cockpit of the aircraft. It is normal for fuel to be drawn from either tank in each wing to feed the adjacent engine. There were two fuel gauges, one for each wing of the aircraft. The gauges indicate the fuel remaining in the tank which is selected on the cockpit fuel tank selector.
About 10 NM before the aircraft crossed the coast at the southern shore of Princess Charlotte Bay, the left engine began running roughly. This position was earlier than the pilot's intended fuel tank change point; however, he regained smooth engine operation by selecting the left inboard tank to feed fuel to the left engine. Prior to the engine running roughly, the left gauge had been reading between one-half and three-quarters full. The gauge indicated full when the inboard tank was selected. The right fuel selector was changed from right outboard to the right inboard tank about 10 minutes later, the time at which the pilot had expected to make that selection. Its gauge reading then changed from below one-quarter full to a full indication.
The flight continued uneventfully until about 50 NM from Cairns when, at 1825, the left engine again began to run roughly. The pilot reported that on this occasion he could only regain smooth operation by cross-feeding fuel from the right inboard tank. Both engines were then drawing fuel from the right inboard tank. He did not notice the left fuel gauge at that time but saw that the right gauge was indicating half-full. Shortly afterwards, the pilot initiated a descent from 9,000 ft.
At 1829, the pilot changed his radio frequency from the Cairns Centre frequency to Cairns Approach. During the initial contact with Cairns Approach he reported maintaining 7,000 ft and was told to expect an ILS approach for runway 15. He then requested minimum delay and was cleared to use maximum speed to the field. The recorded radar data indicated a speed reduction at 1838, when the aircraft was at 3,600 ft. Normal operational transmissions between the controller and pilot continued until 1839, when the pilot made a Mayday transmission during which he said "I think I'm out of fuel here. I'm just going over towards the coast". Radar contact was lost by the Cairns Approach controller at 1842, when the aircraft was descending through 1,000 ft with a ground speed of 84 kts. The controller subsequently calculated the last known position of the aircraft, a datum for search purposes, as a point on the 317 VOR radial at 13 NM from the Cairns DME site.
The pilot told the passengers to ensure that they were strapped in and to prepare for a ditching. He directed the front right seat passenger to unlock the aircraft's entry door and hold it ajar for the ditching. The pilot aligned the aircraft to ditch along the sea swell and did not extend the landing gear or flaps. Because he was concerned that he might hinder the evacuation of the passengers by the need to undo his seat belt, the pilot undid his seat belt prior to the ditching.
The ditching occurred about 1.5 km from shore, after which all occupants promptly vacated the aircraft. The pilot was dazed after his head struck the instrument panel, and was the last to leave the aircraft, assisted by a passenger. The survivors stayed together and commenced moving towards the shore. No life jackets were carried in the aircraft, nor were they required by legislation. After about 45 minutes in the water, two people decided to swim for shore ahead of the others. Soon after this a rescue helicopter located all the survivors about 500-600 m from shore and began rescue winching operations. The survivors determined the order in which each would be rescued, depending on their well-being at the time. During winching of the fourth person, an inflatable rescue boat operated by the Cairns Airport Rescue and Fire Fighting Service arrived at the scene and was able to rescue the remaining survivor.
The pilot and one passenger received minor injuries during the ditching. One passenger was injured during the winching operation. The other two passengers were not injured. The aircraft sank within minutes of the ditching.
Pilot in command
The pilot had been working as a flying instructor and held a Grade 2 instructor rating prior to commencing employment with the company as a Cairns-based pilot. He had recently gained a command instrument rating for multi-engine aircraft. The flying for the rating had included his Aztec aircraft type endorsement training. Before the trip to Horn Island, he had refamiliarised himself with aircraft handling procedures and had conducted simulated instrument approaches for Horn Island and Cairns, in the company's simulator. He believed that the indications provided by fuel gauges fitted to this category of aircraft were generally unreliable. The pilot was appropriately qualified to undertake the flight.
Operational aspects
The operator employed a chief flying instructor and a chief pilot at Cairns. The chief flying instructor was responsible for type conversion training as well as pilot check-and-training duties. A senior base pilot and a senior pilot were employed at Horn Island in addition to three other pilots. Flying statistics were forwarded to Cairns regularly; however, the operations at the two bases were conducted independently. New pilots employed at Horn Island usually commenced by flying the Aztec aircraft, progressing to other types as they gained additional experience. Each day's flying in the Torres Strait area was arranged by Horn Island staff on the previous day. The arrangements for the diversion to pick up the Bamaga passenger were also made on the day prior to the accident flight.
An air operators certificate (AOC) had been renewed to the operator on 26 June 1996 and was valid to 30 June 1997. An amendment to the AOC, to introduce a new aircraft type, had been promulgated on 11 September 1996. This revised AOC was also valid to 30 June 1997.
Operational surveillance of the operator by the Civil Aviation Safety Authority (CASA) had last been conducted on 16 January 1996, in accordance with planned schedules. Ramp checks of the operator had been conducted in accordance with CASA schedules, mainly in the Torres Strait area. No significant shortcomings had been detected during the various checks. The latest ramp check on the operator was conducted on 18 April 1996. The latest ramp check involving VH-CAR was conducted on 20 June 1995. Dangerous Goods inspections of the operator were conducted on 4 December 1995 and again on 23 July 1996.
VH-CAR was maintained by its owner, who was a licensed aircraft maintenance engineer based in north Queensland and not an employee of the operator. Surveillance of the maintenance organisation had been conducted early in 1996 because the owner had established the business under its own certificate of approval. No deficiencies likely to have contributed to the problems encountered during the accident flight were found.
Fuel system
The aircraft was fitted with four flexible rubber fuel cells, two in each wing outboard of the engines. Each cell was fitted with button-type fasteners on the bottom, and bayonet-type clips on the top side, to allow the cell to be attached to the wing structure. These fasteners were intended to retain the cell in its correct shape. As a result of tank material deterioration or problems with the tank venting system, these fuel cells can detach from their retaining points and become distorted. Such distortion can reduce the tank capacity, prevent fuel flow to the engine, and/or cause erroneous fuel gauge indications. Instructions for operation of the fuel system were contained in the aircraft flight manual and in the manufacturer's owner's handbook. The VH-CAR flight manual section on fuel management stated, in part, that "Crossfeed shall be used only to extend range after failure of one engine". Flight manuals for other Aztecs were checked and found to contain no such instruction. The pilot indicated that he had not read the flight manual for this aircraft. The investigation found nothing in the manufacturer's owner's handbook for this aircraft to indicate that crossfeed should not be used when both engines were operating. In fact, the owner's handbook gave instructions for the use of crossfeed to run both engines from one tank. A number of experienced pilots provided information to the investigation, indicating that the engines ran quite successfully when both were fed from one fuel tank, provided the manufacturer's procedures were followed. Tests conducted in 1979 by an experienced pilot, following an incident involving fuel crossfeeding in an Aztec, showed that it could take up to 17 seconds to restart an engine after fuel starvation.
The investigation team conducted an analysis of estimated fuel consumption. At the time of the first rough-running event, the left outboard tank should have been almost empty. When the right fuel tank selection was changed, its outboard tank should have been empty. On the second rough-running occasion, the left inboard tank should have contained about 78 L of fuel. At that time, when both engines began operating from the right inboard tank, the tank should have contained about 84 L of fuel. When both engines stopped some 13 minutes later, there should still have been 57 L of fuel in the right inboard tank.
Two passengers reported seeing smoke, mist or vapour coming from each engine during the flight. The pilot's attention was drawn to the matter, but he indicated that it was not of concern. The pilot could not recall this event during interview. The investigation was unable to discover any evidence of possible defects involving either engine.
The pilot said that no unserviceabilities which would have affected the aircraft's operational capability for the flight had been recorded in the maintenance release. He had been instructed by the senior base pilot on Horn Island to enter on the maintenance release any problems he discovered so that they could be corrected. Because of the apparent anomalous indications from the left fuel gauge experienced during the flight, the pilot intended recording the gauge as faulty upon his arrival in Cairns. He considered that the left gauge reading was too high on a number of occasions during the flight.
Examination of the aircraft maintenance records did not disclose any ongoing problem with the fuel system of this aircraft. All fuel system maintenance undertaken had been as a result of normal wear or airworthiness directive requirements. There was no record of recent maintenance work performed on the aircraft which could have had an influence on the problems encountered during the accident flight. The operator provided flying times and refuelling quantities for some months prior to the accident. Comparison of the hours flown and the fuel used did not reveal excessive fuel consumption rates, nor was there a trend towards high fuel consumption.
Weather information
The general weather was such that most of the flight was conducted above cloud. Scattered cloud was encountered during the descent. Throughout the flight, a headwind component of about 20 to 25 kts was experienced. In the area of the ditching the surface wind was south-easterly at about 15 kts.
Communications
Communication and radar facilities operated normally throughout the period of the flight. Contact with the aircraft was lost at about 1,000 ft above sea level due to terrain shielding. The same problem hindered communications during the search and rescue operations.
The approach controller handling the aircraft provided prompt assistance when the pilot reported his problem. Telephone system congestion between Cairns and the Rescue Co-ordination Centre in Brisbane hindered communications between the parties. Steps have since been taken to provide additional capacity through another service provider should a similar problem occur. Air traffic controllers also experienced difficulty contacting the Queensland Emergency Services helicopter crew because they were using an outdated list of telephone numbers. The list has since been corrected.
The ditching
No ditching instructions were contained in the various publications relating to the operation of the aircraft. The pilot ditched the aircraft using information derived from prior discussions of the circumstances likely to be encountered in a ditching. He did not extend the flaps because he wanted to keep the aircraft attitude flat and was concerned about a possible "pitch-down and nose-dive" if the flaps contacted the water first. The propellers were not feathered after engine power was lost. During the latter stages of the flight, the pilot noticed an indicated airspeed of 80 kts. He was not aware of the airspeed at touchdown, nor was he trying to maintain a particular speed.
Search and rescue
The Queensland Emergency Services helicopter crew was available at its Cairns Airport base. In accordance with their procedures, a second pilot was required because the flight was likely to involve night winching over water. The departure for the accident location was delayed by about 5 minutes for the second pilot to arrive. The crew used the ditching position of the aircraft calculated by the approach controller as a datum for the search. The helicopter was equipped to conduct a visual search using the night sun searchlight. The crew commenced the search 1 NM before the datum and continued until 1 NM after the datum. Their intention was to search a small area either side of the likely position of the aircraft. Although the sun had set there was still some light from the west, and they elected to fly a reciprocal track further to the east in the hope that survivors might be silhouetted against the light. This was unsuccessful, so the next search leg was flown to reposition the helicopter at the datum. A reciprocal track towards the west was then flown as the start of a series of search patterns working from the search datum towards the coast. The survivors were found about half-way along that first search leg. A crewman was lowered to assist each survivor from the water and to supervise each winching aboard the helicopter. The first person to be winched aboard the helicopter was suffering from asthma and was given oxygen. The helicopter pilots, aware that the survivors might have difficulty countering the downwash from the main rotors, moved the helicopter away from the survivors during each winching operation.
Two inflatable boats crewed by Rescue and Fire Fighting Service (RFFS) staff from Cairns Airport arrived near the end of the helicopter winching activities. These boats carried only flotation devices (similar to inflatable rafts) for survivors of an aircraft ditching because the boats were intended to be operated within 2 km of the airport. On this occasion the boats were transported by road to a beach close to the search area before being launched. The last survivor was passed a life jacket from the helicopter so that he would have one to wear in the boat.
ANALYSIS
Fuel system handling
Although the pilot considered that the left fuel gauge read higher than he had expected on a number of occasions, he did not perceive any trends in the fuel gauge indications, probably due to his mistrust of the accuracy of fuel gauges. However, had he consistently monitored the gauges, he may have recognised the need for earlier alternative action. Fuel system management by the pilot was in accordance with the operator's procedures, which were compatible with the manufacturer's owner's handbook. The first rough-running engine event took place to the north of Cooktown. Since the pilot needed to change tanks to correct the problem, and this was before his planned tank-change time, a landing at Cooktown to visually check the actual fuel state would have been prudent. Such a decision should have been reinforced by the left gauge indicating between one-half and three-quarters full. This was markedly different from the right gauge and much higher than the reading should have been for the elapsed flight time. The pilot continued towards Cairns, still satisfied that sufficient fuel remained. This decision was probably influenced by his opinion that fuel gauges in this category of aircraft were generally unreliable. When the second rough-running engine event occurred, the pilot should have become more concerned. He regained engine power by cross-feeding both engines from the right inboard tank. This situation was not planned, nor should it have been considered normal. Options to increase range were available. A reduction to long-range cruise power would have significantly reduced the fuel flow, and by maintaining 9,000 ft or 7,000 ft until much closer to Cairns, the aircraft's range could have been extended.
These options were not considered by the pilot. Examination of the aircraft logbooks and maintenance worksheets found no history of fuel system problems. The aircraft should have contained sufficient fuel for the flight plus company-required reserves. No trends or possible explanations for the fuel problem could be determined by the analysis of refuelling records for a number of months earlier, nor from an examination of the timings and circumstances described by the pilot. Consideration was given to a number of scenarios which could have led to the problems encountered on this flight:
1. One or more fuel tanks could have been deformed prior to the refuelling at Horn Island. This would have reduced tank capacity and may not have been noticeable during a visual inspection.
2. Tank venting problems during flight could have deformed one or more tanks, resulting in reduced fuel flow to the engines.
3. Fuel leakage could have developed during the flight. As the aircraft wreckage was not recovered, the reason for the loss of engine power could not be determined.
Survival aspects
The pilot's decision to release his seat belt prior to the ditching was unsound. He had not considered that he could be incapacitated by impact forces. Had he been injured more severely he may have been a considerable handicap to the egress of the other occupants from the aircraft.
SIGNIFICANT FACTORS
1. Fuel-flow interruptions involving the left engine were experienced on two occasions during the flight.
2. The pilot did not divert or take actions to minimise the effects of possible fuel starvation or exhaustion.
3. It is likely that both engines failed due to fuel starvation or exhaustion.
4. The pilot's head injuries resulted from his decision to not wear a seat belt during the ditching.
SAFETY ACTION
As a result of the investigation into this occurrence, the Bureau of Air Safety Investigation issued Safety Advisory Notice SAN970048 to Airservices Australia. The SAN highlighted the following deficiencies in the role and use of inflatable rescue craft in the rescue operations:
1. The rescue boats were designed and equipped to respond to an aircraft ditching within a distance of 1,000 m from the boundary of the airport. In this case they were operating some 12 NM from the airport. The rescue crews experienced radio communication problems with the control tower due to the distance from the aerodrome and the terrain between the ditching area and the aerodrome. Also the crews of the boats were not able to communicate with the police service. The police service is responsible for the coordination of all rescue services at the scene.
2. The boats were dispatched by the local RFFS district officer. There was no definitive policy in place at the time to cover the operation of the boats at this distance from the airport. No coordination was conducted between the crews of the inflatable boats and the Rescue Coordination Centre in Brisbane, which was responsible for the conduct of the search-and-rescue operation.
3. The survivor who was rescued by boat had been in the water for over one hour, and had removed most of his clothing in an attempt to remain afloat. The boat crew had no clothing or other material available for the survivor to prevent any further heat loss.
On 4 June 2013, a Cessna 210 aircraft, registered VH-ZMT, departed Elcho Island for Ramingining aerodrome, Northern Territory. The purpose of the flight was to pick up a passenger and return to Elcho Island about 15 minutes prior to last light.
The aircraft arrived at Ramingining and joined the circuit. The pilot completed his pre-landing checks, but inadvertently omitted to lower the landing gear and confirm that it had been extended. When on final, the pilot could not recall conducting his final checks.
During the landing, the pilot reported hearing a scraping sound and noticed that the propeller had come into contact with the ground. He then realised that the aircraft had landed with the landing gear retracted. The pilot stated that the aircraft’s landing gear warning horn did not sound at any stage during the landing.
The pilot reported that he had a difficult passenger on board one of his previous flights and was 20 minutes behind schedule. As a result, the pilot reported that he was keen to get home, which had distracted him somewhat during the flight.
As a result of this occurrence, the aircraft operator has advised the ATSB that they have issued a Flight Operation Service Notification to all pilots highlighting the requirement to follow checklists, including the pre-landing checklist.
This incident highlights the impact distractions can have on aircraft operations and the need for pilots to remain vigilant when completing checklists.
On 2 June 2013, the pilot of a Mooney M20J aircraft, registered VH‑NFP (NFP), conducted his pre-flight checks at Canberra Airport, Australian Capital Territory for a private flight to Albury, New South Wales under visual flight rules.
As it had rained at Canberra Airport earlier in the morning, the pilot paid particular attention to conducting pre-flight fuel drains and checking for water, with none found.
During the take-off run, at about 1038 Eastern Standard Time, the pilot reported that all cockpit indications were normal, the aircraft obtained full power and achieved the expected rotate speed followed by a positive rate of climb.
The pilot retracted the landing gear at about 100 ft above ground level. Within seconds of retracting the gear, the engine stopped. The aircraft descended and landed on the runway heavily on the left wing and landing gear, with the propeller striking the ground. The aircraft was substantially damaged, and the pilot suffered minor injuries.
An inspection of the engine after the accident revealed water in the left tank, fuel system and fuel injector lines. The pilot reported that he contacted the Mooney Service Centre and was advised that incorrect re-sealing of the M20 series aircraft fuel tanks could allow 1 to 2 litres of water to be retained in the wing, which could not be drained.
This accident highlights the importance of currency in emergency procedures training. A pre-take-off safety briefing can remind the pilot of the procedure in event of an engine failure at low altitude. Controlling the aircraft at low altitude and maintaining airspeed can reduce the severity of these incidents.
