FACTUAL INFORMATION

The aircraft departed Auckland for a scheduled flight to Melbourne. Weather forecasts available to the crew indicated that the weather at the estimated arrival time of 2210 UTC would be a variable wind of 5 kts and CAVOK conditions. A trend type forecast (TTF) was issued at 2000 and indicated that the wind was calm with 1 okta of cloud at 1,000 ft.

The forecast was appended with the term NOSIG which indicated that there was no significant change due in the next 3 hours. The automatic terminal information service at 2109 indicated that the cloud now was 6 oktas at 400 ft with lower patches and that the visibility was 4,000 m in fog. The aircraft at this time was past the point at which it could divert to an alternate aerodrome. A special weather report (SPECI) was issued at 2130 and indicated that the cloud would be 7 oktas at 200 ft and that the visibility would still remain at 4,000 m.

These conditions were below the alternate minima for the airport, and the aircraft was not carrying additional fuel to divert. The aircraft landed uneventfully at 2214 after becoming clear of cloud at the minimum altitude with an estimated visibility of 800 m.

ANALYSIS

A report from the Bureau of Meteorology indicated that a number of local conditions contributed to the incident. There was an area of low cloud that was situated over the centre of Melbourne city and this did not extend to the airport.

The dewpoint in the area was considerably higher than normal at 14 degrees. The layer of cloud was thin and was expected to be eroded by the sun. However, it had risen considerably and commenced surface heating. There was also a northerly wind at approximately 5 kts and this was thought to assist in keeping the cloud to the south of the airport and over the city.

However, a weak eddy formed and forced the cloud layer to move to the north and the base to become considerably lower than first forecast. A similar meteorological situation was present a few days after the occurrence and in that case, no cloud formed at the airport.

The Boeing 767 departed Melbourne at approximately 0640 UTC for a scheduled flight to Sydney. The aircraft had an estimated arrival time at Sydney of approximately 0725. The crew had been given the Terminal Area Forecast (TAF) issued at 0000 and the Trend Type Forecasts (TTFs) issued at 0100 and 0430 at flight planning and these indicated that the visibility would be in excess of 10 km and the cloud base would be mostly above 2,500 ft. The routine meteorological report (METAR) which was issued at 0600 indicated that visibility would be reduced to 9,000 m in haze and that the lowest cloud base would be 4 oktas at 1,500 ft.

The crew received this forecast by the aircraft communication addressing and reporting system (ACARS) during the pre-flight preparation at Melbourne. A terminal area forecast was issued at 0606 and this indicated that the cloud base would be 3 oktas at 1,000 ft. Another METAR was issued at 0630 and this indicated that the cloud would be 1 okta at 1500 feet with 7 oktas at 1,800 ft. The visibility on this report was indicated to be in excess of 10 km. A further METAR was issued at 0700 and this indicated that the cloud base would be 3 oktas at 1,400 ft, with further cloud of 7 oktas at 1,800 ft. The visibility, however, had decreased to 8000 m in drizzle. This forecast had the term NOSIG appended to it indicating that there was no significant change expected within the next 3 hours.

A special weather report (SPECI) was issued at 0717, which indicated that the cloud base had lowered to 2 okta at 900 ft and that the visibility had decreased to 6,000 m in drizzle and mist. Coincident with the issuing of this report, a TTF was issued and this forecast was appended with the term INTER which indicated that for periods of up to 30 minutes during the validity period of the forecast the visibility would decrease to 6,000 m and that the cloud base would be 2 oktas at 1,000 ft with further cloud of 7 oktas at 1,500 ft. On arrival in the Sydney terminal area, the aircraft received the automatic terminal information service (ATIS), which indicated that the visibility had reduced to 4,000 m in rain and that the cloud base had lowered to 800 ft.

This weather was worse than that which the crew had expected from the forecasts. The aircraft landed uneventfully at 0724. A report from the Bureau of Meteorology (BOM) indicated that the weather conditions were consistent with the passage of a southerly change that had passed through Sydney approximately 2 hours prior to the aircraft departing Melbourne. Forecasts issued by BOM over the period from 0600 to 0717 indicated a gradual decrease in the visibility and a lowering of the cloud base. All of these forecasts were available to the crew in flight on a request basis.

The ATIS for the period from 0655 to 0713 also indicated that the cloud base was gradually lowering and that the visibility was decreasing. The crew would also have had this information available to them in flight. Forecasts issued by BOM were based on observations from trained weather observers situated at Sydney airport as well as output from a number of weather stations around the Sydney area and throughout the country. It also took into account predictions from the numerous computer programs available to BOM. Because of the time taken to analyse all the available data, there was a time delay in updating the weather forecasts.

The ATIS provided by personnel in the control tower at Sydney airport is based largely on visual observation of the conditions prevailing at the airport at the time. As control over the issuing of the ATIS rested with the personnel in the control tower, changes to the information could have been made almost instantaneously. The time delay experienced could be observed with BOM's issue of another SPECI at 0730, indicating that the cloud base had lowered even further. Even though there were only 13 minutes between the two SPECI reports, it was during this time that the incident aircraft, as well as others, landed.

The aircraft departed Canberra for Merimbula on a VFR flight with a departure instruction to remain on the Tower frequency. A short time after departure, ATC instructed the pilot to call Approach, but there was no reply. The aircraft was then observed to make a right turn to intercept the outbound track to Merimbula although no transponder return was being interrogated. The aircraft was subsequently observed orbiting over the Googong reservoir.

ATC tried unsuccessfully to make radio contact. Shortly after, the aircraft was observed to be returning to the airport for an approach to runway 30. On finals it was evident that the landing gear was not fully extended and a full aerodrome emergency was declared. The aircraft then made a low fly pass and the tower observation confirmed that the landing gear was not down and locked.

The aircraft subsequently made a wheels up landing on the grass parallel to runway 30. When the aircraft came to rest, the sole occupant walked clear of the aircraft. During the maintenance investigation, the landing gear retraction system components were tested and found to be serviceable. The aircraft battery was found to be fully discharged.

On final approach into Melbourne at 2,000 feet an Engine Information and Crew Alerting System (EICAS) message alerted the crew to an asymmetrical leading edge slat problem. The crew selected flaps to 20 degrees and initiated a go-around. After completing all checklist actions, another approach was commenced with the flaps set at 20 degrees. The aircraft subsequently made an uneventful landing.

The asymmetric slat fault indication is a known 767 problem. Boeing is currently working on a solution to this recurring system anomaly.

No text.

The crew of the Boeing 747 (B747) had made initial radio contact with the Brisbane sector controller and confirmed their clearance to climb to flight level (FL) 350. A few minutes later, the controller decided to resolve a traffic conflict by maintaining the B747 at FL330 and instructed the crew accordingly. The aircraft was passing FL320 when the instruction was passed and, when the crew correctly read back FL330, the controller attended to other work-related tasks.

There were numerous thunderstorms in the area and several track diversions and altitude changes had been made by various aircraft. When the controller next checked the radar display, he noticed the B747 was passing FL340. As there were several aircraft in potential conflict, the controller decided to allow the crew to continue with their climb to FL350 and resolve the separation tasks with radar vectors and altitude changes to other aircraft. He did not inform the crew of any perceived error. The crew believed that they may have omitted to set the altitude selector to FL330. There was no breakdown of separation standards.

The pilot had planned a solo training flight from Canberra to Scone, returning to Canberra via Goulburn the next day. The aircraft departed Canberra with full fuel tanks and continued to Scone without incident, where it remained overnight. The pilot reported that on the return flight the main fuel tanks were used for take-off, and the auxiliary tanks were selected after the aircraft was established on the departure route. Approximately 50 minutes later, when the right auxiliary tank indicated almost empty, the right main tank was selected. However, as the left auxiliary tank still indicated some 10 to 12 gallons, feed to the left engine from that tank was continued.

A short time later, just prior to commencing descent from 9,000 ft into Goulburn, the left engine lost power. The throttles on both engines were retarded for descent and the left main tank was selected to supply fuel to the left engine. The left throttle was advanced some time later but there was no response from the engine. The pilot reported that his troubleshooting checks consisted of checking that the fuel selectors were on the main tanks, the mixtures were full rich, the ignition was on and that the auxiliary fuel pumps were selected to LOW. The left throttle was again advanced but there was still no response from the engine. The checks were then repeated, but with the auxiliary fuel pumps selected to HIGH. Again, there was no response from the engine.

The same checks were performed on both engines but neither engine responded when the throttles were advanced. Both mixture controls were moved to idle cut-off for about 30 seconds. Full rich mixture was then selected, but the engines still failed to respond. As the aircraft was now at a low height, the pilot concentrated on finding a suitable forced-landing site. Just prior to landing on an open field, he had to manoeuvre the aircraft to avoid a wire and a ditch which lay across the intended approach path. The aircraft subsequently landed heavily, sustaining substantial damage. The pilot evacuated the aircraft without injury.

The investigation determined that although the main tanks had been ruptured in the accident, sufficient fuel would have been available for continued operation of the engines. However, an inspection of the aircraft found there was no HIGH auxiliary fuel pump switch position, as reported by the pilot. The three switch positions available were LOW, which runs the auxiliary fuel pump continuously at low pressure, a centre OFF position, and an up position, placarded ON. In the ON position, the pump runs continuously at low pressure until low fuel pressure is sensed by the engine fuel pressure switch. This could occur due to an engine-driven fuel pump failure, or running a fuel tank empty. The pump then automatically runs at high pressure, and is retained in that mode by a latching relay.

If a fuel tank is allowed to run dry, and fuel supply to the engine is restored by selecting an alternate tank, the engine will then be supplied with an excessive amount of fuel, resulting in a loss of engine power due to over-fuelling. To correct that condition, the auxiliary fuel pump switch must be moved to OFF to unlock the latching relay. The flight manual instructions for the fuel system are to select the auxiliary fuel pumps to ON for takeoff and landing, and to use full rich mixture and auxiliary fuel pumps on LOW when changing fuel tanks. Further investigation determined that the right engine fuel pressure switch had a very high electrical resistance after activation.

Consequently, electrical current was able to flow in the latching circuit and lock the system to high flow mode when selected to the ON position. The left engine initially lost power due to fuel exhaustion of the left auxiliary tank. When the main tank was selected, and the auxiliary fuel pump switch was moved to ON, the pump mode changed to HIGH and the latching relay engaged. The consequent over-fuelling condition prevented the left engine from being restarted the right engine stopped from a similar over-fuelling condition when the right auxiliary fuel pump was selected to ON during troubleshooting.

In 1988, the aircraft manufacturer issued Service Bulletin MEB 88-3, which modified the auxiliary fuel pump system and operation because of reported failures of engine fuel pressure switches. The modification included changing the auxiliary fuel pump ON placard to read HIGH. The service bulletin was applicable to all 300/400 series aircraft. However, the modification was not mandatory in Australia and resulted in aircraft in the Australian fleet having differing operational specifications. Service Bulletin MEB 88-3 had not been incorporated on this aircraft.

Modified aircraft have appropriate cockpit placarding and flight manual instructions, but the pilot's operating handbooks do not include the amended instructions. This leads to a misunderstanding of systems operation between different aircraft. The pilot's response to the engine failure was not in accordance with the manufacturer's checklist.

SAFETY ACTION

The Bureau of Air Safety Investigation is currently investigating a perceived safety deficiency identified as a result of this occurrence. The deficiency relates to a misunderstanding of operational procedures on Cessna 300/400 series aircraft as a result of the optional modification state of the auxiliary fuel system (incorporation of Cessna Service Bulletin MEB 88-3).

Any recommendation issued as a result of this investigation will be published in the Bureau's Quarterly Safety Deficiency Report.

The pilot was attempting to land at Numurkah airstrip in gusty crosswind conditions. After a normal approach, the pilot reports having difficulty controlling the aircraft during the landing flare. The aircraft speed decreased, the aircraft stalled and landed heavily on the runway. The nose wheel collapsed, damaging the engine cowl and propeller.

FACTUAL INFORMATION

The helicopter was being refuelled on an oil rig. The fuel sample taken from the oil rig fuel tank before fuelling commenced was clear of contaminant. However, the sample taken after fuelling was completed had a different colour and displayed obvious contamination with fine metallic and other particles. The helicopter was de-fuelled, and its fuel system cleaned before it was released to service. The oil rig operator reported that the contaminant had been inadvertently left in the oil rig fuel tanks after the rig had undergone refit in Singapore. The operator also reported that the fuel was regularly tested, although the testing carried out had not discovered the contaminant.

The contaminant was only discovered when the rig's fuel pumps had been run for sufficient time to refuel the helicopter. The time was probably sufficient to stir the contents of the tank and bring the contaminant into suspension where it was delivered to the helicopter. The refuelling procedures used appeared to have been appropriate to ensure the helicopter did not take off with contaminated fuel. However, the procedures were not adequate to ensure good quality fuel was delivered to the helicopter.

SAFETY ACTION

As a result of this occurrence, the Bureau of Air Safety Investigation (BASI) is reviewing the adequacy of aviation fuel quality control procedures on oil rigs. Any safety action arising from this review will be forwarded to the relevant action agencies and will be published in the BASI Quarterly Safety Deficiency Report.

The Piper Arrow was on a flight from Bankstown to Albury under the visual flight rules. At Goulburn, the pilot elected to return to Bankstown due to low cloud. During the return flight, the pilot noticed the engine oil temperature and fuel contents indications drop to zero. He transmitted a PAN call but received no response. Assuming he had suffered a radio failure, the pilot continued to Bankstown and joined the circuit on downwind for runway 29R.

Bankstown tower controllers, observing the aircraft approaching with the landing gear retracted, unsuccessfully attempted to contact the pilot by radio. They then flashed a red light signal to indicate to the pilot he should continue circling. The common crash call and crash alarm were activated, and a full emergency declared. The aircraft subsequently touched down with the landing gear still retracted, skidding along the runway for about 70 metres before coming to a halt. The pilot vacated the aircraft without injury.

The pilot subsequently reported that when he lost radio contact, he assumed he had suffered a radio failure, although the radio continued to generate static. He continued his return and made a non-radio approach to Bankstown. After joining the circuit on the downwind leg, he selected the landing gear down but did not monitor the landing gear warning lights as he was concentrating on a slower aircraft ahead, and looking for light signals from the control tower. He did not see the red signal light and failed to realise the landing gear was not extended until he landed.

A maintenance check of the aircraft did not find any defects associated with the electrical, radio or landing gear systems which may have contributed to the accident.