Collision with water, Piper PA31-350 Chieftain, VH-MZK, Spencer Gulf, South Australia, on 31 May 2000

Executive summary

Report release date: 19/12/2001

Since the release of this report on 19 December 2001, further events took place, and additional information become available, regarding issues that were examined in this report. A Supplementary Aviation Safety Investigation Report was released in October 2003 after the ATSB formally re-opened its investigation in November 2002 to address the most significant among those issues.

On the evening of 31 May 2000, Piper Chieftain, VH-MZK, was being operated by Whyalla Airlines as Flight WW904 on a regular public transport service from Adelaide to Whyalla, South Australia. One pilot and seven passengers were on board. The aircraft departed at 1823 central Standard Time (CST) and, after being radar vectored a short distance to the west of Adelaide for traffic separation purposes, the pilot was cleared to track direct to Whyalla at 6,000 ft. A significant proportion of the track from Adelaide to Whyalla passed over the waters of Gulf St Vincent and Spencer Gulf. The entire flight was conducted in darkness.

The aircraft reached 6,000 ft and proceeded apparently normally at that altitude on the direct track to Whyalla. At 1856 CST, the pilot reported to Adelaide Flight Information Service (FIS) that the aircraft was 35 NM south-south-east of Whyalla, commencing descent from 6,000 ft. Five minutes later the pilot transmitted a MAYDAY report to FIS. He indicated that both engines of the aircraft had failed, that there were eight persons on board and that he was going to have to ditch the aircraft, but was trying to reach Whyalla. He requested that assistance be arranged and that his company be advised of the situation. About three minutes later, the pilot reported his position as about 15 NM off the coast from Whyalla. FIS advised the pilot to communicate through another aircraft that was in the area if he lost contact with FIS. The pilot's acknowledgment was the last transmission heard from the aircraft. A few minutes later, the crew of another aircraft heard an emergency locater transmitter (ELT) signal for 10-20 seconds.

Early the following morning, a search and rescue operation located two deceased persons and a small amount of wreckage in Spencer Gulf, near the last reported position of the aircraft. The aircraft, together with five deceased occupants, was located several days later on the sea-bed. One passenger remained missing.

On 9 June 2000, the wreckage of the aircraft was recovered for examination. Aside from the engines, no fault was found in the aircraft that might have contributed to the accident. Both engines had malfunctioned due to the failure of components of the engines.

The crankshaft of the left engine fractured at the Number 6 connecting rod journal. Fatigue cracking was initiated by the presence of a planar discontinuity in the journal surface. It was evident that the discontinuity had been caused by localised thermal expansion of the nitrided journal surface following contact with the edge of the Number 6 connecting rod big end bearing insert. The crankshaft failed approximately 50 flights after fatigue crack initiation.

The Number 6 bearing insert was damaged during engine operation through the combined effect of:

  • high bearing loads created by lead oxybromide deposit induced preignition, and
  • lowered bearing insert retention forces associated with the inclusion of an anti-galling compound between the bearing inserts and the housings.

Fatigue cracking in the Number 6 connecting rod big end housing had developed following the gradual destruction of the bearing insert. The left engine probably continued to operate for 8-10 minutes after the final fracture of the Number 6 connecting rod housing before the final disconnection of the Number 6 journal of the crankshaft. It is likely that the engine would have displayed signs of rough running and some power loss during this time. The final disconnection of the crankshaft resulted in a loss of drive to the magnetos, fuel pump, camshaft and, consequently, the sudden stoppage of the engine. The left propeller was in the feathered position when the aircraft struck the water, confirming that the engine was not operating at that time.

The physical damage sustained by the right engine was restricted to the localised melting of the Number 6 cylinder head and piston. The piston damage had allowed combustion gases to bypass the piston rings. The overheating of the right engine combustion chamber components was a result of changes in heat transfer to cylinder head and piston surfaces created by combustion end-gas detonation. The carbonaceous nature of the residual deposits on the piston crowns indicated that detonation had occurred under a rich fuel-air mixture setting. Rich mixture settings are used with high engine power settings.

The damaged piston would have caused a loss of engine oil and erratic engine operation, particularly at higher power settings. Engine lubrication was still effective at impact, indicating that oil loss was incomplete and that the piston holing occurred at a late stage of the flight.

Examination of the right propeller indicated that the blades were in a normal operating pitch range (i.e. not feathered) when the aircraft struck the water. It could not be confirmed that the right engine was operating when the aircraft struck the water, although it most probably was operating when radar contact was lost as the aircraft descended through 4,260 ft when 25.8 NM from Whyalla.

The aircraft was not fitted with a Flight Data Recorder (FDR) or a Cockpit Voice Recorder (CVR), nor was it required to be. Analysis of recorded radar data confirmed that the aircraft performed normally during the flight until the latter stages of the cruise segment when the speed gradually decreased. Speed variations, accompanied by track irregularities, then became more pronounced. Analysis of recorded voice transmissions revealed that propeller (and engine) RPM during the climb from Adelaide was 2,400. The RPM was 2,200 after the aircraft levelled at 6,000 ft. These were normal climb and cruise engine settings used by the company and the performance achieved by the aircraft during these segments was consistent with normal engine performance. Just prior to the commencement of descent, an RPM of 2,400 was identified. That was not a normal engine power setting for that stage of the flight.

The aircraft speed and propeller RPM information, coupled with the engine failure analysis, was consistent with the following likely sequence of events:

  • The power output from the left engine deteriorated during the first third of the cruise segment of the flight after the Number 6 connecting rod big end housing had fractured. The engine ceased operating completely 8-10 minutes later.
  • In response to the failure of the left engine, the pilot increased the power setting of the right engine.
  • Increased combustion chamber component temperatures via detonation within the right engine led to the Number 6 piston being holed. That resulted in the erratic operation of the right engine with reduced power and controllability and left the pilot with little alternative but to ditch the aircraft.
  • The double engine failure was a dependent failure.

Examination of eight failures of Textron Lycoming engines from a number of operators that had occurred over the period January 2000 to November 2001 revealed that deposits of lead oxybromide on combustion chamber surfaces were not restricted to the engines from MZK; seven other engines had such deposits. The inclusion of a copper-based anti-galling compound between the bearing insert and big end housing was noted in three of the engines examined. The quantity of anti-galling compound present varied between those engines.

Lead oxybromide deposits and anti-galling compounds act in different ways to weaken the defences for reliable engine operation. The relative contribution to engine failure of the factors cannot be predicted easily because of variations in the extent of each effect and the complexity inherent in engine assembly and operation. It is likely that the formation of lead oxybromides that cause deposit induced preignition is linked to the temperature of the fuel-air charge temperature in the combustion chamber just prior to the passing of the flame front. Leaning the mixture during climb, and using near "best economy" cruise power settings appeared to favour the formation of lead oxybromide deposits that resulted in deposit induced preignition. Mixture settings of "full rich" mixture during climb and "best power" cruise settings appeared to favour reactions that resulted in less extensive and different deposits being formed. The Whyalla Airlines procedure was to lean the mixture during climb, and to use a cruise power setting close to "best economy". Those procedures were in accordance with the US Federal Aviation Administration (FAA) approved Pilot's Operating Handbook for the Piper Chieftain aircraft.

The combination of the use of leaded aviation gasoline, mixture leaning during climb, and leaning for best economy during cruise was not restricted to Lycoming engines. The ATSB also found evidence of high combustion loads and lead oxybromide deposits during the examination of components from two Teledyne Continental TIO-520 engines that were defective.

Anecdotal reports indicated that there were fewer engine problems (including component failures) in engines that were operated full rich during climb, and "best power" during cruise, compared with those where the mixture was leaned during climb and "best economy" cruise power was used. A comparison of the engine operating procedures of twelve other operators of Piper Chieftain aircraft revealed considerable disparity in procedures, particularly for climb and cruise. In fact, no two operators used the same procedure.

The incidence of lead oxybromide deposits in engines that had experienced defects, coupled with the range of fuel leaning techniques used, indicated a deficiency in the operation and maintenance of those engines, at least among some of the operators of high-powered piston engine aircraft in Australia.

On 30 October 2000, the ATSB issued a recommendation that the Civil Aviation Safety Authority alert operators regarding the risks of detonation, and encourage the adoption of conservative fuel leaning practices. This report includes further recommendations addressing the following:

  • the engine operating conditions under which combustion chamber deposits that may cause preignition are formed (addressed to the US Federal Aviation Administration);
  • the effect on engine reliability of the use of anti-galling compounds between connecting rod bearing inserts and housings (addressed to the US Federal Aviation Administration and the engine manufacturer); and
  • the reliability of high-powered aircraft piston engines operated in Australia (addressed to CASA).

This accident was the first recorded ditching involving a Piper Chieftain aircraft in Australia. Available records world-wide of previous Piper Chieftain engine failure/ditching events illustrate that, in most instances, successful night ditchings occurred in better visibility and weather conditions than those confronting the pilot of MZK. The relatively minor injuries suffered by the occupants of the aircraft indicated that the pilot demonstrated a high level of skill in ditching the aircraft. The report includes a recommendation to CASA regarding guidance material for pilots on ditching.

It is likely that the survival prospects of the occupants would have been enhanced had the passenger seats been fitted with upper body restraints, and life jackets or equivalent flotation devices had been available to the occupants. As a result of a separate investigation, the Bureau issued a recommendation concerning upper body restraints on 31 March 1999. On 30 October 2000, arising from the Whyalla investigation, the ATSB issued recommendations to the Civil Aviation Safety Authority concerning the provision of adequate emergency and life saving equipment for the protection of fare-paying passengers in smaller aircraft during over-water flights.

Full details of safety action including the CASA response to recommendations made on 31 March 1999 and 30 October 2000 are in Section 4 of this report.

The investigation included a detailed examination of the regulatory history of Whyalla Airlines from June 1997 to June 2000. In common with the published findings of other reports on CASA surveillance activities, there was a significant under-achievement of surveillance of the company against CASA's planned levels during that period. However, there was insufficient information to conclude that the level of surveillance achieved was of significance with respect to the accident.

With regard to Whyalla Airlines itself, issues were identified in the company that had the potential to adversely influence safety. There was insufficient information to conclude that any of these issues were of significance with respect to the accident.

As a result of the accident and ATSB's investigation, improved refuelling procedures were introduced nationally by the refuelling organisation to reduce the chance of error.

  • Supplementary report was released on 28 October 2003

Occurrence summary

Investigation number 200002157
Occurrence date 31/05/2000
Location 28 km SE Whyalla Airport
State South Australia
Report release date 19/12/2001
Report status Final
Investigation type Occurrence Investigation
Investigation status Completed
Mode of transport Aviation
Occurrence class Accident
Highest injury level Fatal

Aircraft details

Manufacturer Piper Aircraft Corp
Model PA-31
Registration VH-MZK
Serial number 31-8152180
Sector Piston
Operation type Air Transport Low Capacity
Departure point Adelaide, SA
Destination Whyalla, SA
Damage Destroyed

Boeing 747SP-38, VH-EAB

Summary

While at approximately 8,000 ft after departure from Sydney, the crew heard a loud bang from the left side of the aircraft. The number 1 engine pressure ratio (EPR) had dropped slightly while the compressor speed (N1), exhaust gas temperature and fuel flow indications increased. The engine vibration level remained within limits. A flight attendant reported seeing something entering into or flaying around the engine intake. After discussions with ground maintenance, the aircraft was returned to Sydney, where an uneventful landing was made after about 50,000 kg of fuel had been jettisoned off the coast.

Examination of the engine by the operator revealed evidence of a bird strike. Blood was found on the engine fan exit nozzle, and the glass-reinforced plastic of the ice impact panel was damaged. About 90 per cent of the circumference of the ice impact panel that covered the acoustic liner aft of the fan was found to be missing

The fan was inspected for damage and the remains of the ice impact panel were removed. (The engine maintenance manual allows the ice impact panel to remain damaged beyond limits or missing for up to 125 hours, or until the next aircraft maintenance inspection.) The EPR probes and their lines were flushed clean and appropriate line connectors were replaced.

A subsequent engine ground run found EPR indications had returned to normal and the engine was cleared back to line operations.

Occurrence summary

Investigation number 200002130
Occurrence date 13/05/2000
Location Richmond
State New South Wales
Report release date 27/09/2000
Report status Final
Investigation type Occurrence Investigation
Investigation status Completed
Mode of transport Aviation
Aviation occurrence category Birdstrike
Occurrence class Incident
Highest injury level None

Aircraft details

Manufacturer The Boeing Company
Model 747
Registration VH-EAB
Serial number 22672
Sector Jet
Operation type Air Transport High Capacity
Departure point Sydney, NSW
Destination Manila, Philippines
Damage Minor

Boeing 737-377, VH-CZA

Safety Action

As a result of the investigation, the Australian Transport Safety Bureau issues the following recommendations:

R20000301

That Airservices Australia review the documentation in relation to air traffic controller recency requirements, in particular, the methodology of how individuals can meet the requirements.

R20000302

That the Civil Aviation Safety Authority review the requirements for air traffic controller recency, such as the number of hours required, shifts that do or do not qualify for such recency, and documentation for recording such recency.

Related safety action

Related issues have been found in several Australian Transport Safety Bureau investigation reports and the Bureau is developing recommendations in respect of the affect of personal stress on controllers.

Significant Factors

  1. The Inverell controller was experiencing personal stress factors that may have affected his work performance.
  2. The Inverell controller had minimal recency time on the sector, which may have affected his performance.
  3. The crew of CZA requested and were approved to climb to non-standard level FL370.
  4. The Inverell controller did not implement a separation assurance plan when he became aware of the likelihood of a conflict.
  5. The Inverell controller allowed secondary considerations to override separation actions.
  6. The short-term conflict alert activated in sufficient time to allow remedial action by the controller.
  7. The TCAS on each of the aircraft provided each crew with an alert indicating a required positive separation action.

Factual Information

Sequence of events

A Boeing 737, VH-CZA, departed Maroochydore for Sydney and was tracking southbound on air route H62. The crew planned to cruise at flight level (FL) 350 but experienced turbulence at that level. Several other aircraft had encountered turbulence at various levels up to FL350 but those at FL370 were in smooth flying conditions. Consequently, the crew had obtained approval from air traffic control to climb to non-standard level FL370 well before the aircraft was handed off to the Inverell sector controller.

A Boeing 737, VH-CZX, departed Melbourne for Brisbane, tracking via air route H29 and on climb to the flight planned level, FL370.

The air routes crossed near the Gibraltar non-directional navigation beacon.

The crew of CZA reported maintaining FL370 to the Inverell sector controller at 1322 Australian Eastern Standard Time.

At 1323, CZX was identified to the Inverell sector controller by the previous sector controller. The crew of CZX made radio contact at 1327, and reported that they were maintaining FL370. Between 1327 and 1329, the controller issued arrival and sequencing information to the crew. This information did not change the aircraft's current route or flight level.

The operational shift supervisor commenced a conversation with the Inverell sector controller at 1330 regarding the selection and testing of radio frequency outlets for that sector. The controller was performing other tasks and said that he would call the supervisor back. Between 1330 and 1332, the controller spoke with the crews of seven aircraft, including asking one of them for a radio check of readability after changing the radio outlet settings.

At 1332:30, the controller contacted the supervisor to discuss the selection of primary and secondary transmitters and receivers. During this exchange, the short-term conflict alert activated at 1333:22 due to confliction between CZA and CZX. The supervisor deselected the intercom line without further conversation.

The controller immediately issued a radar vector to the crew of CZX to effect avoiding action. The crew commenced the turn as instructed and received a traffic alert and collision-avoidance system (TCAS) resolution advisory to climb. They carried out the TCAS climb during the turn.

In the next 15 seconds, the controller passed traffic information on CZX to the crew of CZA. The crew of CZX then informed the controller that they were initiating a TCAS climb, and the crew of CZA broadcast that they were descending in accordance with a TCAS resolution advisory.

Analysis of the radar display data indicated that the radar vector took CZX across the flight path of CZA at 1333:46 while maintaining FL370. At that time, CZA was 10 NM north of CZX and also at FL370.

The point of closest proximity was 3.9 NM at 1334:11, when CZA was descending through FL369 and CZX was maintaining FL370. The required separation for that situation was either 5 NM horizontally or 2,000 ft vertically.

Traffic alert and collision-avoidance system (TCAS)

Each aircraft's TCAS functioned in accordance with design parameters. They activated at about the same time, and assisted the crews to respond appropriately.

The first warning the crew of CZX received was an instruction from the controller to turn right immediately. They commented to each other that the controller sounded worried and commenced the turn without delay. As they started the turn they received a TCAS traffic advisory and then a resolution advisory to climb. The pilot in command remembered seeing a "00" indication, which meant the aircraft were at the same height. The crew did not see the other aircraft, due to the turn manoeuvre, but carried out a climb in accordance with company standard operating procedures. They received a "clear of conflict" message after climbing 400 ft. The aircraft remained within a 500 ft excursion of its cleared level.

The crew of CZA had asked for FL370 due to uncomfortable turbulence at FL350. They had been at FL370 for some time and were unaware of the approaching aircraft until the TCAS activated a traffic information advice at the 12 NM extremity of the TCAS display. The pilot in command only had time to consider that advice before the traffic advisory activated and he looked up to see the other B737 straight ahead and in a turn. He heard the controller issue a right turn instruction to CZX and was confident that he could maintain visual separation behind that aircraft. However, when the resolution advisory activated, he commenced a descent in accordance with company standard operating procedures. The aircraft remained within a 500 ft excursion of its cleared level.

Air route and airspace design

Air routes H62 and H29 were subject to heavy air traffic movements and had to cross at some point. Air route design ensured that they crossed in the middle of the Inverell sector while at cruising levels. That design had the effect of producing a common crossing conflict in the centre of the display screen that provided the controller maximum time to recognise and rectify a conflict.

Inverell sector controller

The controller operating the Inverell sector was experienced in the position and reported that the traffic level at the time was moderate. The controller also reported experiencing significant personal factors that would have been likely to cause him considerable stress in the weeks prior to the incident.

The controller had taken several days off duty in the previous two weeks due to illness. He was current in only two of the four positions for which he held a rating, having registered 9.5 hours on the Inverell sector in the past 22 days and 3.5 hours in the last 15 days. The currency requirement was for at least 2 hours every 14 days for each rating.

The Civil Air Traffic Services Operations Administration Manual (CATSOAM) parts 5.2 and 5.3 outline the recency requirements for maintaining a valid air traffic control licence. In practice it was possible for some of the time logged to fulfil the recency requirements for a particular control position to be worked during times of low traffic levels. For example, working a nightshift where several positions were combined may have allowed a controller to credit the hours worked to more than one position.

In the 14 days before the day of the incident, the controller had worked only one shift; a night shift involving combined sectors for which he held ratings. That shift enabled him to meet the recency requirements to exercise the privileges of his air traffic control licence on two of his four ratings, including the control position he occupied at the time of the incident.

During the controller's absence, the Byron Group of sectors was moved to a different aisle within the Brisbane Centre. Although each sector was a stand-alone unit, the sectors each side of Inverell were then different. The controller reported that the repositioning of the sectors made the "feel" of the surroundings seem unusual when he arrived at the console.

When the controller arrived to start work at 1100, he had no immediate console function to perform and decided to read the documentation relevant to his ratings to catch up on the latest changes. While on that break a personal distraction resulted in him having to leave the console. Consequently, he did not take over the sector until approximately 1310.

A short time later (1324:30), the controller gave an instruction to the crew of a Dash 8 that they did not read back correctly. That was not recognised by the controller and was rectified only after an enquiry by the crew.

The controller received the correct coordination in respect of both CZA and CZX. He realised the potential conflict and checked, on at least two occasions, the relative distance of each aircraft from the intersection of the air routes, but did not subsequently take any positive action to ensure separation; even though the aircraft were 10 minutes from that intersection.

The controller was asked by the supervisor to check the radio system and, although initially saying that he had some other tasks to perform and that he would call the supervisor back, he commenced the radio checks soon after receiving the request and during a period of comparatively busy radio activity. The controller reported that he felt under pressure to complete the task without delay. That pressure was largely self-imposed as the supervisor had left the timing of the task to the controller's discretion. However, the controller was aware that the technicians were waiting at a remote location for the check to be performed and he wanted the task completed.

During that time the controller also changed the transmitter and receiver functions on the voice switching and communication system, thinking he was carrying out the changes requested by the supervisor. However, equipment-fail "bleeps" were evident as he was asking a pilot to comment on the quality of the radio transmission. Those bleeps indicated that the selections made were not correct and the controller's discussion with the supervisor confirmed that he had made an incorrect selection.

The controller stated that although he had acknowledged that CZA was at FL370, he was not fully aware that the aircraft was at a non-standard level. He was concentrating on the voice switching and communication system and the potential problems for sequencing aircraft that were soon to enter his airspace.

The controller stated that had allowed his "scan" to be diverted and, when the short-term conflict alert activated, he knew immediately what the problem was and acted to rectify the infringement of separation standards. He was unable to explain either his poor task prioritisation or his memory lapse.

Short-term conflict alert

The short-term conflict alert was designed to activate when the system checks determined that the aircraft were within 1 minute of coming within 4.1 NM and 600 ft. Analysis of the recording medium indicated that the alarm provided approximately 16 NM warning of the point of closest proximity. As the closing speed was approximately 15 NM/min, indications were that the system activated within reasonable tolerances.

Operational shift supervisor

The supervisor was dealing with radio frequency outlet problems on two sectors. He tasked the other sector first, as Inverell was the busier. Radio technicians at Point Lookout found equipment that needed replacement, but the task required the primary transmitter to be unavailable for about 24 hours. Before authorising its removal, the supervisor needed to check that the standby arrangements for the Inverell sector were satisfactory, and contacted the controller to have them checked. When the controller replied that he was busy, the supervisor waited for his return call. Two minutes later, the controller returned the call. The supervisor said that he thought that it was acceptable to the controller to commence conversation, otherwise he would not have returned his call. During that conversation the supervisor heard an alarm and immediately terminated the call. Although he did not know the nature of the alarm at the time, it was the short-term conflict alert that the supervisor heard.

Analysis

Both aircraft crews acted in accordance with company procedures and followed the TCAS resolution advisories.

The incident occurred at a well-known point of conflict within an en-route sector. The experienced controller was operating a control position with which he was familiar and with traffic volume and complexity that should have been well within his ability. However, he did not take action to ensure separation between two aircraft that he had earlier recognised as being in potential conflict. That was most likely the result of a number of predisposing factors, including the effects of stress, limited recency, distraction, and not using a memory aid.

At the time of the occurrence the controller was testing the serviceability of the secondary radio transmitter. That was a routine and relatively unimportant task. However, the controller felt under pressure, largely self-imposed, to complete it without undue delay. Initially the controller was mistaken as to the exact test required and that led to some confusion on his part. Consequently, he was distracted for longer than anticipated and his awareness of the developing traffic situation was compromised.

Correct prioritisation is fundamental to any complex operating task. The controller was faced with several competing demands for his attention. In addition to monitoring the conflicting aircraft he carried out routine tasks such as issuing instructions to aircraft and transferring aircraft to or from other control sectors, and attempted to assist the supervisor with the radio checks.

Inadequate prioritisation committed the controller to remembering the unresolved aircraft conflict in order to deal with it in a timely manner. However, distraction and a subsequent memory lapse left the conflict unresolved until the activation of the short-term conflict alert.

There were other aspects of the controller's performance that, while not significant in isolation, were possibly indicative of a lower general level of performance at the time of the occurrence. When the southbound aircraft entered the controller's airspace, he did not appreciate that the aircraft was at a non-standard flight level, even though he was well familiar with the airways route. Later, approximately 10 minutes before the incident, the controller did not correct an oversight by a Dash 8 crew. Further, when asked by the supervisor to carry out a frequency check, the controller had some difficulty with what was a relatively straightforward task.

Recency

Using low traffic movement periods to combine sectors in order to maintain adequate work levels may result in an individual meeting the formal recency requirements, but not actually having sufficient exposure to a particular sector to warrant meaningful traffic practice. In this occurrence, the amount and type of recent control work that the controller had completed may not have been sufficient to ensure performance to the standard that the recency requirements were intended to maintain; taking into consideration the variation in the level/complexity of traffic during different periods. It is possible that the controller's performance was affected by the relatively short time he had worked in the control position during the previous fortnight.

Stress and performance

At the time of the incident the performance of the Inverell controller may have been adversely affected by stress. Recent significant personal factors may have been likely to cause him considerable chronic stress. The extent to which stress related to non-work factors can affect work performance is often underestimated. Major life events can markedly affect stress levels. In addition, the acute stress of leaving work to attend to an urgent personal matter may also have influenced the controller's performance.

Individuals are often unaware of the extent to which their performance is affected by stress. They may try to "work on" despite problems or difficulties. Individuals may be reluctant to admit, even to themselves, that they are suffering from stress because of a perception that this will be seen as a sign of weakness or failure.

Research has shown that stress can produce errors such as inappropriately delaying necessary actions and forgetting to carry out required actions at a time of high workload or distraction. Stress can result in perceptual and cognitive narrowing, where attention and decision making are focussed on a restricted range of information and tasks. For example, a controller's scan pattern may be disrupted. Stress can also lead to task shedding. This can result in the neglect of crucial matters while time may be spent on tasks of lesser importance. Memory can be significantly inhibited by stress.

In this incident, stress may have reduced the controller's capacity to handle what would normally have been a moderate workload. The controller allowed himself to be distracted by testing the radio equipment, to the detriment of his primary task; that of managing air traffic.

Memory aids

Because the potential conflict remained unresolved, the controller had to keep the task in short-term memory. Omitting to carry out planned actions - a failure of prospective memory - is one of the most common forms of memory lapse. A necessary condition for a memory lapse to occur is that attention is captured by either an external distraction or an internal preoccupation. The use of an appropriate memory aid would have guarded against the separation task being forgotten and a number of such methods were available to controllers.

Summary

A Boeing 737, VH-CZA, departed Maroochydore for Sydney and was tracking southbound on air route H62. The crew planned to cruise at flight level (FL) 350 but experienced turbulence at that level. Several other aircraft had encountered turbulence at various levels up to FL350 but those at FL370 were in smooth flying conditions. Consequently, the crew had obtained approval from air traffic control to climb to non-standard level FL370 well before the aircraft was handed off to the Inverell sector controller.

Occurrence summary

Investigation number 200002060
Occurrence date 23/05/2000
Location 19 km S Gibraltar, (NDB)
State New South Wales
Report release date 02/11/2001
Report status Final
Investigation type Occurrence Investigation
Investigation status Completed
Mode of transport Aviation
Aviation occurrence category Loss of separation
Occurrence class Incident
Highest injury level None

Aircraft details

Manufacturer The Boeing Company
Model 737
Registration VH-CZA
Serial number 23653
Sector Jet
Operation type Air Transport High Capacity
Departure point Maroochydore, QLD
Destination Sydney, NSW
Damage Nil

Aircraft details

Manufacturer The Boeing Company
Model 737
Registration VH-CZX
Serial number 24029
Sector Jet
Operation type Air Transport High Capacity
Departure point Melbourne, VIC
Destination Brisbane, QLD
Damage Nil

de Havilland Canada DHC-6-320, VH-FNU

Summary

The crew of the DHC-6-320 aircraft reported that during descent to Cairns they smelt smoke in the cockpit. They then noticed the battery load meter was at maximum discharge and the right generator light was illuminated. The left generator switch was moved to the OFF position, however, the generator remained on-line.

Subsequent in-flight inspection revealed that the smoke was coming from behind the right cabin roof panel, which had begun to melt and bubble. The pilot in command then contacted Cairns approach control reporting a fire in the cabin and the aircraft was cleared to track direct to Cairns. The co-pilot accessed the cabin fire extinguisher and extinguished the fire. At approximately 4NM from the airport, the co-pilot reported that the fire was an electrical fire and had been extinguished. After landing, the crew stopped the aircraft on the runway and shut down the right engine to allow the fire fighters access to the cabin. Following confirmation that the fire was extinguished the aircraft was taxied to the terminal.

Investigation by the owner's maintenance organisation found that the left reverse current relay had severe heat damage. Other components and wiring near the left reverse current relay were also heat damaged. The reverse current relay was disassembled by the maintenance organisation's engineers, but they were unable to determine the reason for the failure due to the severity of the heat damage.

The Bureau did not conduct an on-site investigation of this occurrence.

Occurrence summary

Investigation number 200001876
Occurrence date 20/05/2000
Location 28 km E Cairns, Aero.
State Queensland
Report release date 24/08/2001
Report status Final
Investigation type Occurrence Investigation
Investigation status Completed
Mode of transport Aviation
Aviation occurrence category Fire
Occurrence class Incident
Highest injury level None

Aircraft details

Manufacturer De Havilland Canada/De Havilland Aircraft of Canada
Model DHC-6
Registration VH-FNU
Serial number 286
Sector Turboprop
Operation type Aerial Work
Departure point Mackay, QLD
Destination Cairns, QLD
Damage Minor

Cessna 402C, VH-NMQ

Safety Action

Several safety deficiencies were identified during the initial phase of the investigation. Local safety action was agreed to and addressed by the operator prior to publication of the final report. Those safety actions were as follows:

Issue of company Maintenance Alert MA/C400/2, which required a detailed inspection and functional check of the fuel selector system of all the operator's Cessna 402C aircraft.

A second Alert MA/C400/2 amendment 1, was issued a short time later after the initial fleet inspection found that the fuel selector valves for a second aircraft were out of synchronisation with the selector indicator. This alert detailed further inspection and rectification instructions.

The operator has undertaken to re-examine the company operations manual with a view to improving the guidance information on fuel planning and reserves.

The operator has agreed to re-examine the need for the cross-feed and shut-off checks. If they are considered necessary, then a procedure to ensure that the desired outcome is achieved will be devised and introduced.

The operator's maintenance controller has introduced procedures to ensure that all trip records are now vetted daily. Any endorsement item considered to have the potential to affect the safety of flight will now be subject to immediate maintenance rectification.

The operator has amended and amplified the incident and accident reporting section of the company Policy and Procedures Manual to better reflect the correct reporting procedures.

The Operator has introduced the Daniel System Australia, Integrated Aviation Software (IAS) in the company's electronic maintenance control system. This software has the ability to monitor and identify repetitive defects. The company Maintenance Control manual Vol 1, section 9 has been changed to reflect the changed procedures.

Analysis

The fuel selector system defect need not have led to a situation in which the safety of the flight was not assured.

The problem was recognised by the pilot early in the flight, and, despite attempts to resolve it, was confirmed when about mid-way. Consequently, without fuel from the right tank, there was insufficient fuel available for the planned flight fuel requirement, let alone the required reserve fuel. The pilot considered that a landing at an alternate aerodrome may have been necessary. However, the need to maintain adequate reserve fuel to ensure safety of the flight was apparently not recognised. As the flight progressed, the quantity of fuel required was assessed only in the context of a direct approach and landing at Darwin.

After landing, the fuel gauge indicated that the left tank contained about 20 lb of unusable fuel. The pilot had therefore made the decision to overfly an alternate aerodrome when in reality, only about 20 lb (or about 6 minutes) of fuel was available.

Maintenance oversight of the aircraft was inadequate. This is evidenced by:

  • the number of reported occurrences of fuel system problems; and
  • the deferral of rectification of 3 such problems because they were regarded as "not major defect".

The repetition of reports and the request entered on the maintenance release to "please report further", suggest that the focus of maintenance rectification action was deficient.

The late reporting of the occurrence to the ATSB and to the operator's maintenance controller arose through breakdowns in the operator's reporting procedures following staff changes.

Summary

History of the flight

The pilot of a Cessna 402C aircraft, VH-NMQ, was conducting a scheduled passenger service from Tindal to Darwin, with 4 passengers.

During the climb to the planned cruise altitude of 8,000 ft, the fuel quantity gauges indicated that fuel was being consumed from the left tank only. At the top of climb, the fuel gauges indicated 100 lb in the left tank and 200 lb in the right tank. Once established in the cruise the pilot tried to balance the fuel load by positioning the left fuel selector to allow the left engine to be supplied with fuel from the right tank.

Approaching the midway point between Tindal and Darwin, the fuel gauges confirmed that the imbalance remained. The pilot assessed that sufficient fuel was available to continue to Darwin. As the flight approached the last suitable alternate aerodrome, the pilot calculated that 13 lb of fuel was needed to reach Darwin from that point, based on the current consumption rate. The left fuel tank gauge showed 40-lb remaining.

The pilot requested and was given a direct approach to runway 29. During the landing roll the left engine stopped, and the right engine stopped soon after the aircraft cleared the runway. The fuel quantity gauges showed 20 lb for the left tank and 200 lb for the right tank.

The operator's ground support staff then towed the aircraft to the terminal area where the passengers disembarked normally.

The aircraft had flown from Darwin for Tindal the previous day with 650 lb of fuel on board. The pilot's navigation log showed that 365 lb was required for the flight from Tindal to Darwin. This comprised 230 lb flight fuel (59 minutes), 100 lb fixed reserve (30 minutes), 25 lb variable reserve fuel, and 10 lb taxi fuel. The navigation log also showed the total fuel on-board at Tindal before departure for Darwin was 385 lb. The pilot reported that before departure from Tindal, the aircraft fuel quantity gauges showed 185 lb in the left tank and 200 lb in the right. The pre-flight check of the fuel cross-feed was normal.

A landing at an alternate aerodrome was not considered necessary, as the pilot had calculated sufficient fuel was available to continue to Darwin, and was conscious of the operating schedule for the aircraft. However, company management stated that there was no instruction or pressure on pilots in this regard.

The pilot held a current commercial pilot licence, a command multi-engine instrument rating a valid medical certificate, and had logged 2,454 hours aeronautical experience. This included 150 hours on Cessna 402C aircraft. The pilot's last proficiency check was on 8 March 2000.

On the day of the occurrence, 5 May 2000, the pilot did submit an incident report to the operator. However, due to a breakdown in the operator's reporting procedures following staff changes, the ATSB was not notified of the occurrence until 15 May. As a result, it was unable to conduct an examination of the fuel selector system fitted to NMQ before the aircraft was returned to service.

Background

The then Civil Aviation Authority of Australia issued Civil Aviation Advisory Publication No: 234-1 (0) in March 1991. It provided advice about the quantity of fixed reserve fuel to be carried, and noted the use of fixed reserve fuel was limited to unplanned manoeuvring in the vicinity of the destination aerodrome. It further noted that fixed reserve fuel would normally be retained in the aircraft until the final landing.

The operator's fuel policy stated that in all foreseeable circumstances an aircraft should complete a flight with at least the fixed reserve fuel intact. The standard fixed reserve for the operator's Cessna 402C fleet was 100 lb of fuel for 30 minutes of flight.

Cessna 402C Fuel Selector System

The two fuel selector controls are attached to the cabin floor between the pilot and co-pilot seats. The selectors enable the fuel selector valves, located behind the engine firewalls, to be positioned to the corresponding tank, crossfeed, or off. The arrow-shaped ends of the selector control handle points to the position on the selector placard that corresponds to the control valve position. Each handle connects directly to a gearbox located under the floor. Cables connect each gearbox to their respective fuel selector valve.

An inspection by the maintenance provider in Darwin confirmed that the pilot had positioned the right fuel selector handle to the right tank. However, the cable connecting the gearbox to the fuel selector valve did not position the valve to the right tank.

The right fuel selector gearbox was removed, the selector gearing adjusted and the selector cable re-rigged. The aircraft was returned to service after the fuel selector system was ground tested.

The Pilot's Operating Handbook for the aircraft contained a description of the fuel system and it's operation. In the Normal Procedures section of the handbook, pilots were cautioned that they should "Feel for (the) detent" when placing the fuel selector at the desired position.

Maintenance action

The operator's maintenance controller was located at Alice Springs. A sub-contract maintenance provider conducted maintenance at Darwin. The operator's maintenance controller was not advised of the occurrence until 23 May, and was therefore unable to specify extra inspection procedures for the operator's Cessna 402C fleet until that time. The maintenance controller issued maintenance alert MA/C400/2 on 25 May. It was applicable to all Cessna 402C aircraft used by the operator and its associate company, and required a detailed inspection of the fuel selector system. These checks were to be conducted at each scheduled inspection until the requirements were included in the operator's System of Maintenance.

During one such check of the fuel selector system on another of the company aircraft, VH-TZH, the screws securing the selector cable lever arm/travel stop to the sector gear were found to be loose. This allowed the travel stop to slide under the base-plate and the sector gear to move beyond the end of its travel, resulting in loss of synchronisation between the selector handle and the selector valve.

The operator immediately issued an amendment to maintenance alert MA/C400/2 that included photographs of the defective fuel selector on TZH. The alert specified that the selector gearbox was to be checked if the fuel selector valve was found not synchronised with the selector indicator. During the investigation the fuel selectors of two other of the operator's Cessna 402C aircraft were found to lack an effective detent.

At the time of the occurrence, the maintenance release for NMQ contained an annotation arising from a previous report of uneven fuel consumption from the left tank. Examination of the fuel system related maintenance documentation for the aircraft revealed 17 events since December 1999. Seven involved reports of the aircraft using more fuel from the left tank than the right when the fuel selectors were correctly positioned. There was one report of greater fuel usage from the right tank. The rectification section for four of the entries stated that the fuel selector system was re-rigged. The reported defect for three of the entries was annotated "not major defect" and transferred to the deferred defects list. The fuel related defect entry for the 18 April 2000 stated "fuel not feeding correctly. Fuel noted to be crossfeeding some fuel in level flight - report further".

The last reference to fuel transfer problems, prior to the incident, was on 20 April. The endorsement on the trip record stated "both engines draw fuel from the left tank only". The rectification section of the record noted the fuel selector valve was removed, lubricated, refitted and pressure tested. The selector cable was re-rigged, and the system operated satisfactorily during ground testing.

Fuel system pre-flight checks

The fuel system pre-flight checks specified in the operator's Cessna 402C Operations Manual differed from the procedures specified in the manufacturer's Pilot's Operating Handbook.

The operator's standard operating procedures required pilots to operate the fuel supply cross feed for 60 seconds to verify normal operation. Also, pilots were to ensure normal operation of the fuel valves by positioning the fuel selectors to the off position to observe a decrease in fuel flow. Following these checks, pilots were to position the fuel selectors to the main tanks.

The manufacturer's Pilot's Operating Handbook did not specify checks for crossfeed operation or positioning the fuel selectors to the off position to observe a decrease in fuel flow.

The pilot did not move the fuel selectors to the off position as part of the pre-flight checks. This was because the Fleet Manager had advised his intention to amend the pre-flight check to delete the requirement. The pilot reported awareness of the need to feel for the detent when moving the fuel selectors, and was confident the detent was achieved during the pre-flight check.

On 16 May the operator issued a memo on the occurrence to its Cessna 402 pilots, and included new instructions for pre-flight operation of the fuel selector. The memo cancelled the previous requirement for pilots to position the fuel selectors to the off position to confirm a decrease in fuel flow.

Occurrence summary

Investigation number 200001827
Occurrence date 05/05/2000
Location Darwin, Aero.
State Northern Territory
Report release date 23/05/2001
Report status Final
Investigation type Occurrence Investigation
Investigation status Completed
Mode of transport Aviation
Aviation occurrence category Fuel starvation
Occurrence class Incident
Highest injury level None

Aircraft details

Manufacturer Cessna Aircraft Company
Model 402
Registration VH-NMQ
Serial number 402C0451
Sector Piston
Operation type Air Transport Low Capacity
Departure point Tindal, NT
Destination Darwin, NT
Damage Nil

Piper PA-31, VH-FMU

Summary

The pilot was tasked to convey six passengers from Brisbane to Miamba, about 165 NM WNW of Brisbane, and return. The aircraft was parked at Archerfield and the passengers were to be picked up at Brisbane Airport.

The aircraft operator had no formal system of recording the amount of fuel remaining in the aircraft at the end of each flight. Normally, the fuel state was discussed between the pilots of the previous and next flights. In this instance, however, the pilot did not speak to the person who last flew the aircraft. The pilot said that it was his practice to use the fuel gauges and the fuel gauge calibration chart to determine the amount of fuel in the aircraft.

The weather forecast obtained by the pilot indicated the 7,000 ft wind as 110 degrees at 15 kts, and the 5,000 ft wind as 120 degrees at 20 kts. The pilot calculated the following flight times and fuel requirements for the intended flights, assuming a fuel consumption rate of 140 lt per hour:

Archerfield - Brisbane 5 mins, 27 lt
Brisbane - Miamba 59 mins, 174 lt
Miamba - Brisbane 59 mins, 174 lt
Total 123 mins, 375 lt

Before the flight, the pilot obtained information about the destination airstrip and decided to have the aircraft refuelled to full main tanks (415 lt). The fuel gauges for the auxiliary tanks were indicating half full. Using the lesser of the aircraft flight manual and the fuel calibration card figures, the pilot assessed that the auxiliary tanks contained 147 lt. From this information, he calculated that the aircraft contained 562 lt of fuel - sufficient for about 4 hours of flight.

Because of air traffic control requirements and the prevailing weather conditions, the flight from Archerfield to Brisbane took about 30 minutes.

During the flight to Miamba, the pilot selected the auxiliary tanks for a period during cruise at 8,000 ft, but did not record the times when these selections were made. Flight duration was about 66 mins.

At Miamba, the pilot estimated (from the fuel gauges) that there was 280 lt fuel remaining for the return flight to Brisbane. At top of climb (7,000 ft) he again selected the auxiliary tanks. Because the gauge indications reduced faster than he anticipated towards empty, he allowed the auxiliary tanks to empty before re-selecting the main tanks. The pilot said that the fuel quantity gauges indicated about 1/8 when the auxiliary tanks were empty. Because of headwinds, the pilot descended to 5,000 ft to try to achieve a higher ground speed but to little effect. At that altitude, the aircraft was in cloud.

At about 30 nm from Brisbane, with the gauges indicating about 1/8, the pilot became concerned that there might not be sufficient fuel to complete the flight to Brisbane. He descended to 4,000 ft, again seeking a better ground speed. The pilot then elected to divert to Archerfield (which was closer) and told air traffic control that he did not wish to descend further until close to Archerfield. Shortly after being cleared to turn towards Archerfield, the left engine surged and then ceased operating. The pilot feathered the propeller and informed air traffic control that he required an immediate landing at Amberley. He did not declare an emergency.

Soon after the left engine failed, the aircraft flew clear of cloud and the pilot saw Amberley aerodrome. He conducted a straight-in approach to runway 15 at Amberley. He manually extended the landing gear with assistance from the passenger in the right control seat. The aircraft touched down about 5,000 ft along the runway and the pilot deliberately steered the aircraft off the runway edge on to grass late in the landing roll. There was no damage to the aircraft. The flight from Miamba had taken about 84 minutes. After refuelling, the engines operated normally.

Investigation revealed that the aircraft actually contained about 477 lt on start-up at Archerfield - 85 lt less than the pilot believed. That included 220 lt remaining from the previous flight, and 257 lt added during refuel. The aircraft operator said that his company used a rate of 150 lt per hour for flight planning purposes. Based on that figure, there was sufficient fuel on board the aircraft for about 190 minutes flight. From the total flight time of about 180 minutes, the aircraft had achieved an actual fuel usage rate of about153 lt per hour (allowing about 30 minutes ground operating time), close to the planning figure used by the operator. It is logical to conclude, therefore, that the engines performed normally during the flight, and that the fuel supply to the left engine was exhausted, causing it to cease operating. It is likely that the fuel supply to the right engine would also have been exhausted within a few minutes if the pilot had not shut the engine down.

The pilot said that, based on the handling notes for the aircraft, 126 lt fuel per hour would be consumed at 65 per cent engine power (31 inches MAP and 2,200 rpm). He had used a rate of 140 lt per hour in planning for the flight, which he considered more than adequate. Although the pilot's aeronautical experience was significant, over 90 percent of it was as a flying instructor on single engine aircraft. The pilot had little experience in charter operations. His experience on the aircraft type was also low.

Occurrence summary

Investigation number 200002018
Occurrence date 23/05/2000
Location 31 km N Amberley, (NDB)
State Queensland
Report release date 01/06/2001
Report status Final
Investigation type Occurrence Investigation
Investigation status Completed
Mode of transport Aviation
Aviation occurrence category Fuel exhaustion
Occurrence class Incident
Highest injury level None

Aircraft details

Manufacturer Piper Aircraft Corp
Model PA-31
Registration VH-FMU
Serial number 31-8212015
Sector Piston
Operation type Charter
Departure point Miamba, QLD
Destination Brisbane, QLD
Damage Nil

Fairchild SA227-AC, VH-IAW

Safety Action

As a result of the investigation into this and related occurrences, the Australian Transport Safety Bureau is currently investigating a safety deficiency involving the Fairchild SA-226/SA-227 type of aircraft that relates to replacement of certain hydraulic piping and ongoing inspection procedures for the hydraulic piping.

Any safety output issued, as a result of this analysis will be published in the Bureau's Quarterly Safety Deficiency Report.

Summary

The crew of the Fairchild Metroliner was conducting endorsement training in the Cowra area when they advised air traffic services (ATS) that their aircraft had a hydraulic failure and they were returning to Bankstown. An uncertainty phase was declared by ATS and the aircraft returned to Bankstown where a flapless landing was carried out.

The subsequent maintenance inspection found a cracked hydraulic pipe in the left landing gear wheel-well that had allowed the loss of the hydraulic oil contents from the power pack. The crack was in the bend radius of the pipe and was not visible to the naked eye. The crack only became apparent with the application of more than 800 psi of hydraulic pressure to the system during the maintenance inspection. The pipe was changed and the aircraft returned to service.

The pilot in command reported that he had experienced two similar hydraulic pipe failures in this aircraft type and the company fleet had five or six similar failures in the preceding twelve months. He also reported that he had a concern that the escaping oil in this situation could have started a wheel-well fire.

The design of the hydraulic system is such that, when a failure occurs in one of these hydraulic pipes, there is no redundancy. As a result, the leak drains the contents from the hydraulic reservoir, which is the common source of hydraulic fluid for both the left and right engine-driven hydraulic pumps. Without hydraulic services, nosewheel steering, anti-skid braking, flap operation and normal landing gear retraction and extension would not be available to the pilot.

Occurrence summary

Investigation number 200001657
Occurrence date 10/05/2000
Location 1 km S Cowra
State New South Wales
Report release date 02/04/2000
Report status Final
Investigation type Occurrence Investigation
Investigation status Completed
Mode of transport Aviation
Aviation occurrence category Hydraulic
Occurrence class Incident
Highest injury level None

Aircraft details

Manufacturer Fairchild Industries Inc
Model SA227
Registration VH-IAW
Serial number AC-600
Sector Turboprop
Operation type Air Transport Low Capacity
Departure point Cowra, NSW
Destination Cowra, NSW
Damage Nil

Boeing 767-338ER, VH-OGA

Summary

On final approach to land at Auckland, the crew of the Boeing 767 observed the right alternating current (AC) electrical bus fail indication. The crew started the auxiliary power unit (APU) to supply additional electrical power and the approach and landing were completed without further incident. The non-normal checklist was actioned on the ground and the right AC bus power was restored. The customer service manager later advised the pilot in command that the emergency lighting illuminated briefly during the failure but then went out. This left the cabin in total darkness throughout the incident until right AC power was restored.

Maintenance investigation could not fault the electrical systems during ground tests. It is likely that there may have been an electrical earth occurring in the right generator feeder wires or terminal. The right generator control unit (GCU) would then have isolated the right generator, leaving the right AC bus without power. A fault would then be automatically detected in either the current to or from the right AC bus and the right bus tie breaker (BTB) would remain open. This would ensure the right AC bus was isolated and could not be powered by the left engine generator. The APU generator came on-line just before touchdown and automatically powered the right AC bus, however not all the engine indicating and crew alerting system (EICAS) messages were cleared with the APU powering the right AC bus. The messages cleared when the non-normal checklist was completed.

On this production series B767 aircraft, if power to the right AC bus fails then all cabin lighting is extinguished. On later production series B767, the left AC bus supplies cabin side wall lighting and the right AC bus supplies cabin overhead lighting, therefore there is some cabin lighting if either AC bus fails.

When the right AC bus lost power so too did the right direct current (DC) bus. However, as no faults were sensed in the DC electrical system the DC tie relay automatically closed to power the right DC bus from left DC bus. This sequence normally takes 11.5 seconds. The emergency lighting circuit, sensing initially there was no DC electrical power, momentarily activated while electrical system switching took place (the 11.5-second changeover). Once the DC tie relay closed the right DC bus became powered by the left DC bus and the emergency lighting then automatically extinguished. The emergency lights are designed not to light up continuously unless all DC electrical power is lost.

As no definite fault could be isolated for the right AC bus failure, the performance of the aircraft electrical system was being monitored by Engineering Maintenance Watch and the company Flight Safety Department

Occurrence summary

Investigation number 200001647
Occurrence date 10/05/2000
Location Auckland, ILS
State International
Report release date 01/08/2000
Report status Final
Investigation type Occurrence Investigation
Investigation status Completed
Mode of transport Aviation
Occurrence class Incident
Highest injury level None

Aircraft details

Manufacturer The Boeing Company
Model 767
Registration VH-OGA
Serial number 24146
Sector Jet
Operation type Air Transport High Capacity
Departure point Sydney, NSW
Destination Auckland, NEW ZEALAND
Damage Nil

Piper PA-34-220T, VH-YSG

Significant Factors

  1. The pilot's fuel planning was based on incorrect fuel content and weight considerations.
  2. The pilot did not include the Sydney Harbour scenic flight in his flight planning.
  3. The pilot diverted from the recommended in-flight fuel management practices.



 

Analysis

The aircraft had been assessed as having 280 L of fuel on board prior to departure from Canberra, based upon the delivery pilot's calculation. ATSB calculations, using the operator's own fuel planning figures, indicated that the amount of fuel consumed during the 1.8 hour delivery flight from Tamworth would likely have equated to a fuel remaining figure of approximately 265 L. In addition, no allowance was able to be determined for the amount of fuel consumed during the engine ground runs at Tamworth, between the time of the last recorded refuel and the delivery flight. Accordingly, the actual fuel tank contents on departure from Tamworth for the delivery flight could not be accurately determined and would probably have been an amount less than the maximum useable.

While planning for the accident flight utilising a computer-based application, the pilot used an incorrect figure (465 L) with respect to total versus useable fuel tank contents (454 L). That led the pilot to include 11 L of fuel, which was unavailable for engine consumption. In addition, he omitted to plan an allowance for fuel consumption during the approximately 12-minute Sydney Harbour scenic flight. Using the operator's fuel consumption planning figure, this may have involved consumption of up to 20 L of additional fuel.

The pilot's use of the computer-based flight planning application included calculations for aircraft weight and balance. His utilisation of the full fuel planning figure in these calculations, instead of the assessed fuel load figure of 280 L, may have led to an erroneous perception that he would be overweight following embarkation of the five passengers at Bankstown, if he added fuel prior to take-off from either Bankstown or Canberra. Had the pilot used the assessed fuel tank content figure, he would have realised that he could have added fuel and remained within aircraft weight limitations.

The manufacturer's fuel usage figures are determined using a recommended technique for leaning of the fuel mixture supplied to the engines. During the accident flight, the pilot deviated from the recommended technique. That deviation would likely have increased the engine fuel consumption.

Technical investigation of the right engine indicated that it was capable of normal operation. Examination of the left engine indicated that it should still have supplied at least partial power in flight. The failure mode of the exhaust valve rocker stud and the bending of the inlet valve pushrod were examined in consultation with a representative of the engine manufacturer and engine overhaul specialists. Although an exact reason for the failure could not be determined, it is possible that the damage may have occurred following the loosening of one or both of the rocker pivot retaining nuts. As the nut(s) loosened during engine operation, the exhaust valve rocker would have lifted, preventing the exhaust valve from opening. During the valve overlap stage of the engine operating cycle, prior to the start of the induction stroke, the ignited and expanding combustion gasses may have been unable to escape through a now closed exhaust valve. That scenario would have greatly increased internal cylinder pressure which could have prevented the inlet valve opening and, consequently, resulted in the bending of the inlet valve push rod as it tried to move the valve. The damage to the inlet valve pushrod was also considered to have been consistent with the rotation of the engine with a "hydraulic lock" situation existing in the cylinder. Such a condition may occur in the cylinder if the engine was over primed prior to an engine start.

It is likely that the engine failures occurred as a result of fuel exhaustion.

Factual Information

The Piper PA34 Seneca III, took off from Canberra for Bankstown under the Visual Flight Rules (VFR). The pilot in command had planned to embark five passengers at Bankstown then to conduct a scenic flight around Sydney Harbour, proceed to Cessnock for lunch and return to Bankstown where he intended to disembark the passengers before returning to Canberra.

The flight to Bankstown was uneventful. On departure from Bankstown, the pilot proceeded southeast to the coast and then north, at low level, along the VFR coastal route to Manly. A delay was experienced at Manly, prior to turning south and entering Sydney Harbour. The Sydney Harbour scenic flight was then conducted, completing a circuit of the Harbour Bridge to Rushcutter's Bay loop. The pilot then returned to South Head and proceeded north via Aeropelican for a landing at Cessnock.

During the return flight, the pilot deviated west of the planned track in order to avoid severe weather that had been indicated on the aircraft's weather radar. Shortly after passing Brooklyn Bridge, which was the start point for the northern inbound VFR track to Bankstown, at an altitude of 2,500 ft, the pilot reported that the left engine suddenly stopped. The pilot reported that he then carried out the engine failure checks from memory, feathering the propeller. Approximately one minute later, the right engine began to gradually lose power, before stopping after approximately a further 20 seconds. The pilot reported that he then pushed all of the engine and propeller controls forward and attempted to restart the engines. That action resulted in the inadvertent unfeathering of the left propeller due to operation of the unfeathering accumulator.

The pilot conducted a forced landing into a treed area beside a main road. The aircraft struck the trees, rotated to the right and impacted the ground left wing forward. Four of the six occupants sustained serious injuries and the aircraft was destroyed.

Pilot information

The pilot was appropriately licensed for the flight and had accumulated a total of 270 hours flying experience with 14 hours on the aircraft type.

Flight and fuel planning

The aircraft arrived on the operator's flight line five days before the accident, after delivery from Tamworth. The technical log indicated that the last recorded refuel had been conducted at Tamworth on 13 March 2000. Between that time and the delivery flight to Canberra the aircraft had undergone a 15-minute dual engine run following a propeller change. The engine run included a period of one to one and a half minutes with both engines at full power. Prior to departure from Tamworth, the delivery pilot reported that he visually assessed the fuel tanks as full. Upon arrival at Canberra, the fuel tank content was assessed as 280 L, using flight time and fuel usage calculations.

The pilot planned the accident flight utilising a computer-based flight planning application. He did not plan for any aerial work at Sydney Harbour, despite having an intention to conduct a scenic flight in that location. Air Traffic Services (ATS) radar data indicated that the scenic flight duration was approximately 12 minutes. In addition, he planned the flight using figures for full fuel tanks, however he used the total fuel tank content figure of 465 L instead of the useable fuel tank content figure of 454 L. Further, he was aware that the aircraft had only an assessed 280 L of fuel on board and considered that that amount was sufficient for the flight from Canberra to Cessnock and for the return flight to Bankstown. He had decided not to add fuel prior to departure from Canberra due to a perception that the aircraft would be over Maximum Take-off Weight at Bankstown or Maximum Landing Weight at Cessnock, after the addition of the five passengers.

Fuel management

The pilot reported that he visually checked the fuel tank contents at Canberra before departure, with fuel just visible through the fuel cap opening. That quantity was then cross-checked with the aircraft fuel gauges and it appeared to agree with the previously calculated total. An attempt was also made to check the fuel tank contents using the aircraft's fuel dipstick, however, that was unsuccessful as the fuel level was well below the lowest graduated scale on the stick. The pilot reported that he again checked the fuel level on arrival at Bankstown. At that time the level was out of sight of the fuel caps, however, the total fuel remaining on the fuel gauges appeared to agree with his assessment. Prior to landing at Cessnock the fuel gauges were again checked, with the pilot considering that the indicated fuel level was what he expected from his mental calculations to allow for the return flight to Bankstown.

ATS radar data indicated a total flight time of 159 minutes from take-off at Canberra to the accident site. That figure did not include any allowance for ground taxi at Canberra, Bankstown and Cessnock. Using the operator's recommended fuel usage planning figure of 100 L per hour, 159 minutes flight time would have consumed 265 L of fuel.

The aircraft manufacturer's recommended engine fuel leaning procedure for cruise flight was detailed in Section 4 (Normal Procedures) of the Pilot's Operating Handbook, which stated:

"For 45, 55 and 65% power the mixture should be leaned to 25 [degrees] F rich of peak E.G.T. [Exhaust Gas Temperature] but not to exceed 1,650 [degrees] F E.G.T.", and "For maximum engine service life, cylinder head temperatures should be maintained below 420 [degrees] F..."

The pilot reported that he had been taught to lean the fuel mixture to the top of the green arc on the exhaust gas temperature gauge. The green arc on the gauge extended from 1,200 to 1,525 degrees F. He reported that during the accident flight he leaned the mixture according to that method, however he had subsequently enriched the mixture one or two graduations below the top of the green arc. Advice from the aircraft manufacturer and other Seneca III operators, indicated that that action may have increased the fuel flow by up to approximately 10 L per hour.

Approximately one litre of fuel was recovered from the aircraft wing tanks, however the fuel tanks had been substantially damaged during the impact sequence. There was no evidence of fuel leakage on the ground. Inspection of the fuel system components revealed no evidence of fuel contamination. The left engine fuel selector control was positioned at the OFF position and the right engine selector was positioned at ON.

Engine and fuel system information

Both engines had recently been overhauled and had flown 57.3 hours since fitment to the aircraft.

During the flight, a short time after passing Aeropelican, the pilot reported that the left engine had required increased throttle to maintain the selected manifold pressure. While taxying after landing at Cessnock the left engine appeared to idle slower than before and the alternator light flickered on and off. The light had extinguished when the pilot increased engine RPM. After vacating the aircraft, the pilot noticed evidence of an oil leak along the outboard side of the left engine cowling, on the underside of the left wing and on the left flap. A check of the left engine oil quantity revealed that the engine had used about half a quart during the flight. Confirmation with the operator by telephone indicated that the oil level was within tolerances. The source of the oil leak was not able to be located. The pilot reported that operation of both engines from departure at Cessnock to the point of engine failure appeared normal.

Technical investigation at the accident site revealed that the left engine number 4 cylinder exhaust rocker pivot was loose and that the rocker had contacted and holed the inside surface of the rocker cover. The forward rocker pivot-retaining stud had sheared flush with the surface of the head with evidence of a fatigue failure on the stud fracture surface. The number 4 cylinder inlet valve push rod had also failed close to the outboard tip of the rod. Detailed technical examination of both engines at an engine overhaul facility noted that the exhaust valve on the number 4 cylinder of the left engine remained closed during rotation of the crankshaft and the inlet valve on the same cylinder only opened a small amount. Both engines were then test run in an engine test cell. The test run of the right engine revealed no condition that would have contributed to the in-flight loss of power. The test run of the left engine indicated that it was capable of operation, however it's performance was affected by the damage evident on the number 4 cylinder's valve mechanism.

Occurrence summary

Investigation number 200001434
Occurrence date 25/04/2000
Location 13 km S Brooklyn Bridge, (VTC Check Point)
State New South Wales
Report release date 10/04/2002
Report status Final
Investigation type Occurrence Investigation
Investigation status Completed
Mode of transport Aviation
Aviation occurrence category Fuel exhaustion
Occurrence class Accident
Highest injury level Serious

Aircraft details

Manufacturer Piper Aircraft Corp
Model PA-34
Registration VH-YSG
Serial number 34-48020
Sector Piston
Operation type Private
Departure point Cessnock, NSW
Destination Bankstown, NSW
Damage Destroyed

Boeing 737-476, VH-TJY

Summary

While on final approach to Brisbane at about 1,000 ft, the crew of the Boeing 737 reported feeling a 'kick' in the rudder pedals accompanied by a minor aircraft yaw oscillation. It was reported that there was no aircraft ahead to cause wake turbulence.

Investigation by the aircraft operator suggested a problem with the rudder power control unit. The aircraft manufacturer recommended that the standby rudder actuator and the standby input rod bearings should also be examined for wear; no wear was evident.

The rudder power control unit had accumulated 3,064 hours since its overhaul by the manufacturer in May 1999. It was removed from the aircraft and forwarded to the USA for examination.

The examination was conducted by the component manufacturer and supervised by the National Transportation Safety Board of the USA on behalf of the Australian Transport Safety Bureau. No discrepancies that may have led to the anomaly in the operation of the rudder power control unit were found. The internal and external components contained no evidence of excessive wear, damage or overtravel and met the manufacturer's standards for in-service units.

In support of the investigation, the National Transportation Safety Board also conducted a performance simulation study based on the actual aircraft configuration data at the time of the incident. The study concluded that the rudder had oscillated.

The investigation was unable to determine why the rudder reportedly oscillated.

Occurrence summary

Investigation number 200001362
Occurrence date 18/04/2000
Location Brisbane, Outer Marker
State Queensland
Report release date 06/02/2001
Report status Final
Investigation type Occurrence Investigation
Investigation status Completed
Mode of transport Aviation
Aviation occurrence category Flight control systems
Occurrence class Incident
Highest injury level None

Aircraft details

Manufacturer The Boeing Company
Model 737
Registration VH-TJY
Serial number 28151
Sector Jet
Operation type Air Transport High Capacity
Departure point Melbourne, VIC
Destination Brisbane, QLD
Damage Nil