Circumstances:

On the morning of the accident the pilot flew the aircraft from a parking area, which was restricted due to vegetation, to a more open area in order to board the passenger and depart for the day's mustering. The aircraft departed the open area and commenced to climb on a westerly heading. At approximately 300 ft and 60 kts the witnesses heard a sharp crack, all engine and rotor noise ceased, and the aircraft was observed descending at a steep angle. The aircraft collided with the ground about 450 m from the witnesses. The occupants were removed from the wreckage by the witnesses before the cockpit area was consumed by a post-impact fire. The on-site investigation determined that one main rotor blade had separated in flight. The helicopter became uncontrollable following the blade separation. The out-of-balance rotor system caused considerable damage to the helicopter before ground impact. The roof of the cockpit was destroyed, both fuel tanks were torn from the fuselage, and the tail boom and rotor were cut off. The post-impact fire, fed by residual fuel in the fuel lines and engine oil, consumed what remained of the cockpit area. Subsequent detailed examination determined that the main rotor blade, Serial No. 2961, failed as a result of fatigue crack growth in the root fitting of the blade. No fatigue cracks were found in the other blade, Serial No. 2953, fitted to the helicopter.

Fatigue cracking initiated in the counterbore of a hole in the root fitting of the blade. No material abnormalities were present at the initiation sites. Fatigue crack growth was estimated to have occurred over a period of approximately 1100 flights. The blade had been in service for 2257.2 hours, although the retirement life of the blade was 2000 hours. There was evidence to suggest that the clamping force produced by a bolt installed in the hole during bonding was low. It is likely that the clamping force provided by the bolt installed at final assembly was also low. It was considered that the low clamping force was caused by misalignment of the holes in the spar and root fitting and an off-centre and off-axis screw thread. The misalignment caused interference between the spar and the root fitting, and the low clamping force caused a change in the load transfer, at the hole, allowing fatigue to develop under normal service loads. The location of the fatigue crack, in the root area, was covered by a layer of flexible skin which prevented it being detected by the inspection requirements that were in force at the time of the accident. The pilot was under considerable financial pressure and was attempting to earn sufficient funds to purchase new main rotor blades. He had made a practice of recording less than the correct hours in the aircraft documentation and it was likely he was aware that the blades had exceeded their safe life. The pilot apparently had made a conscious decision to overfly the maximum number of permitted hours, possibly based on the knowledge that the blades had been safely tested to twice their approved life. However, had he grounded the aircraft when the blades reached their safe life the accident would not have occurred.

Significant Factors:

The following factors were considered relevant to the development of the accident.

1. There were financial pressures which caused the pilot to continue operating the aircraft beyond the mandatory retirement life of some aircraft components.

2. A manufacturing defect occurred which allowed fatigue cracking to develop during the main rotor blade's normally safe life.

3. There was an inflight fatigue failure of one main rotor blade which led to complete loss of control.

Recommendations:

The Bureau of Air Safety Investigation made two recommendations at the preliminary stage of the investigation.

1. The Civil Aviation Authority should review the retirement time for Robinson R22 main rotor blades using information based on the true service time of the failed blade.

2. The Civil Aviation Authority should develop and implement an inspection technique for the main rotor blades which will detect progressive fatigue failure in the area of the rib root fitting. The Civil Aviation Authority reduced the service life of all R22 main rotor blades to 1000 hours until an adequate inspection system had been put in place.

An adequate inspection system was developed and placed into service. In addition, the manufacturer reviewed the manufacturing process and implemented changes to prevent a similar manufacturing defect to the one that was a factor in this accident.

Circumstances:

Prior to landing in a large paddock, the pilot advised the passengers that the balloon would touch down after it passed a dead tree about halfway along. The passengers assumed the pre-briefed crouched position and braced themselves for the landing. Just after the balloon passed the dead tree one of the passengers observed two power cables running across the flight path five metres in front of the balloon. The pilot was alerted and almost simultaneously, the balloon's suspension cables struck the power line. After contact, the balloon slid along the conductors for approximately 30 metres, pulling them downwards, until the basket contacted the ground. One passenger reported receiving an electric shock causing the pilot to order an evacuation and he and two passengers left the basket. The pilot was later observed reaching towards the balloon from outside the basket. Both the pilot and the passenger, who had remained in the basket, were electrocuted. Fire subsequently destroyed the balloon. The power cables ran across the paddock in a single span with a distance between the poles of 366 metres. The supporting poles were in adjoining paddocks and hidden by trees. The type and layout of the cables made them very difficult to see. Evidence indicates the pilot was not aware of the location of the power line prior to the flight, and he did not see them until just prior to impact and at that point it was too late to avoid them. The retrieval crew were following the balloon and did not arrive at the landing site until after the accident. The power cables made contact with the uninsulated, inadequately bonded envelope suspension cables attached to the two forward corners of the basket. There was some evidence that a lack of bonding may have caused an electrical potential difference across the basket which in turn led to the shock reported by the passenger. It could not be determined which of the metal components the pilot and passenger were touching when they were electrocuted. Injuries received by the pilot and damage to the fabric covered metal parachute vent line, indicated that it was probable that the pilot touched the vent line allowing the current to flow to ground. The fabric cover on the vent line was insufficient insulation for the magnitude of the current involved. Each of the power cables was protected by a fuse and although both these fuses worked, the failure sequence did not prevent the two fatalities. Had both cables been earthed earlier in the sequence, it is probable that the fuses would have blown thus protecting the balloon's occupants. The actions of the pilot and passengers indicated a lack of knowledge of safety practices to be used when in the vicinity of high voltage power cables.

Significant Factors:

The following factors were considered relevant to the development of the accident:

1. The pilot was unfamiliar with the area where he intended to land and he was not aware of all the potentially dangerous obstructions, such as power lines.

2. The pilot did not see the power cables until it was too late to avoid them.

3. Inadequately insulated and bonded balloon components which came into contact with live power cables and a lack of a suitable method of safely grounding the balloon after power line contact.

4. There was a lack of adequate knowledge of safety practices to be used when in the vicinity of high voltage power cables.

Recommendations:

It is recommended that the Civil Aviation Authority, in conjunction with the Australian Ballooning Federation and Commercial Balloon Operators, reassess existing requirements for commercial balloon operations and surveillance of standards and in particular give consideration to

1. Introducing a requirement that operators of fixed duration flights, be required to have adequately surveyed all proposed landing sites within its area of operations.

2. Ensuring as far as practicable, that at least one member of the retrieval crew is either present at the proposed landing site, or in such other position as to be able to brief the pilot on obstacles and assist with rapid deflations and evacuations as required.

3. Introducing a safety education program which provided pilots with advice on electrical contact safety procedures.

4. Redefining the Flight Manual Emergency Landing procedures concerning the briefing of passengers before ground contact, with particular emphasis on orderly basket evacuation and electrical contact safety procedures.

5. Initiating manufacturer approved methods, of reducing the amount of exposed metal and providing electrical bonding of all metal components, to achieve neutral electrical potential difference between any two components.

6. Initiating a research and development program into on-board, electronic, directional, power line detection devices.

7. Initiating a research program into whether or not a pre-touchdown electrical grounding device should be fitted to all balloons

Circumstances:

The aircraft was being flown by the owner/pilot, who was on a flying holiday with three friends. The pilot had experienced navigation difficulties on two occasions while endeavouring to proceed to Schofields. As a result of those occurrences, the Civil Aviation Authority required the pilot to undertake local navigation training before continuing the next stage of the flight. Following the training the pilot departed Schofields at 1724 hours for Scone with a planned time interval of 55 minutes. Calculated end of daylight was 1834 hours. The flight planned track passed over aerodromes at Cessnock and Singleton. From details provided by the survivor, it would appear that the pilot thought he had reached Scone but was unable to locate the aerodrome. It is considered that the pilot probably reached Muswellbrook, a similar sized town near Scone but without an aerodrome. At 1830 50 EST the pilot reported to Sydney Flight Service "Sydney, Charlie Mike Echo, Scone circuit area, cancel SAR please". At about that time the aircraft landed on a curved gravel road at an open-cut mine, located about 40 km from Scone. A witness said that he had spoken to the occupants after the aircraft had landed. They said they had intended to land at Scone but had run out of daylight before they could find the aerodrome. They had decided to return and land on a strip they had seen earlier but were unable to find it and so decided to land on the road. The pilot said they would camp the night and leave early the next morning. He also asked that the witness "keep things as quiet as possible because there could be trouble if the right people were to hear about it". The next morning the aircraft was observed by a number of witnesses to attempt to take-off from the gravel road in the opposite direction to that used for landing. This had involved passing through a steel frame, some 18 metres wide, spanning the roadway and then negotiating a curve of about 50 degrees. After a take-off roll of about 500 metres, in an easterly direction, the aircraft had become airborne for a brief distance before the landing gear struck a low mound of rocks bordering the edge of part of the open-cut mine. The aircraft had then descended steeply, at a relatively low forward speed, before impacting heavily on a wide "step" within the open cut, about 200 feet below the level of the roadway. The weather at the time of the accident was reported to have been fine and cool with a light breeze from the west.

Significant Factors:

The following factors were considered relevant to the development of the accident:

1. Pilot became lost/unsure of position.

2. Pilot unwilling to seek assistance due to perceived possible repercussions.

3. Pilot exercised poor judgement. 4. Pilot elected to use unsuitable area for take-off.

Recommendations:

It is recommended that the circumstances of this accident, and the relevant events leading to it, form the basis of an article on operational decision making and the responsibilities of command. It is suggested that the article be published in the Aviation Safety Digest.

Before commencing the 17th spreading flight for the day, the pilot noted that one fuel tank indicated empty and the other indicated one quarter full. After a normal take-off and turn at about 150 feet above ground level, the engine lost all power. The pilot was committed to a landing in a small paddock with a downhill slope. Touchdown was made in light tailwind conditions, and during an attempt to turn the aircraft to lengthen the landing distance available, the left wing struck the ground. The aircraft partially ground looped, one tyre was rolled off its rim, and the aircraft came to rest within the confines of the paddock. The pilot then physically checked the fuel tank contents and found that only a few litres remained in one of the tanks, while the other was empty. The loss of engine power was caused by fuel starvation. The pilot had not previously flown the aircraft and was not aware of the time the aircraft had been flown since it had last been refuelled. He did not accurately determine the quantity of fuel in the aircraft prior to commencing the operation, nor did he have any method of determining the duration of the flight. He relied solely on the fuel gauges to determine the quantity of fuel in the aircraft.

The pilot had recently been employed by the operator. As he had not flown the aircraft type for some time he was given a check flight a few days before the accident. The need to check the main landing gear uplock rollers was emphasised then and the pilot was aware of the need for this check in any case. His preflight inspection before the accident flight included a check of the uplock rollers. His method of checking the rollers was to approach the landing gear from the front of the wing, move the downlock out of the way with his thumb and rotate the roller using his index finger. The flight was normal until gear was selected down on final approach to runway 14 at Brisbane. Approach flap (15 degrees) had been selected prior to this. When he did not obtain a gear down indication (three green lights) he made a missed approach to check the problem, retracting the gear and flap in the process. Flap was later selected prior to lowering the gear but the pilot found that the flap would not extend. Further attempts were made to lower the landing gear but a gear down indication could not be obtained. The pilot advised that he would land wheels up and he was instructed to fly to Archerfield for this landing. Before landing at Archerfield he received advice on ways to try to get the gear down from the engineers who normally maintain the aircraft. This included pulling positive and negative "g" while selecting the gear down. The aircraft was eventually landed wheels up at Archerfield. Later in the day the aircraft was lifted by crane and transported on a semi-trailer to a hangar. The aircraft was placed on jacks before any attempt to lower the gear was made. Examination of the left main landing gear indicated that the uplock roller was seized. Moisture had penetrated into the roller initiating corrosion and degrading the grease. Periodic regreasing had not been sufficient to displace the mixture of old grease, moisture and corrosion products. The rigging of the left landing gear could not be checked due to fracture of the operating rod. During the investigation it was learned that, with failure of the gear to extend due to the uplock not releasing, further attempts to extend the gear would be futile if the gear system was in either the gear up or gear down positions. A wire cable connecting the locks to the gearbox is under tension when the gear is either up or down. As a result the gear leg can not be moved.

The pilot had planned to complete the flight without an enroute refuelling stop, providing that the headwind component was not excessive. He retained a stock of fuel at the DEPARTURE strip, however this had become depleted, and he was reluctant to delay his DEPARTURE until further stocks arrived the following day. By draining the remaining fuel from a number of near-empty drums, the pilot assessed that he had sufficient fuel to complete the journey. No visual inspection of the quantity of fuel in the tanks was made, although the pilot had placed a hand into the filler neck of the left tank, and was able to splash fuel with his fingers. About 135 minutes after DEPARTURE, while cruising at 9500 feet, the pilot detected a change in engine note and some rough running. He altered the mixture control setting without effect, but following a slight reduction in power the engine again ran smoothly. As it approached the destination the aircraft was given clearances to descend to 5000 feet above mean sea level (amsl). The pilot contacted Adelaide Approach when the aircraft was 31 miles (57 kilometres) from the aerodrome, and was further cleared to descend to 1000 feet amsl. This clearance allowed the pilot to descend at whatever rate he considered appropriate for the circumstances. He later advised that the aircraft was at 1000 feet amsl and about 11 miles (20 kilometres) from the aerodrome when the engine again ran roughly and then lost all power. He changed the fuel tank selection and carried out other trouble checks, but the engine failed to respond. The aircraft was too low for the pilot to be able to glide to land, and he was forced to carry out a ditching in St Vincent Gulf. Neither the pilot nor the passenger in the front seat had the sash portion of their seat harnesses fastened, and both suffered injuries when they struck their heads on the instrument panel. However, all occupants managed to evacuate the aircraft before it sank, but no life jackets had been carried on the flight. The group remained together in the water for some time, but then one passenger swam away and was not seen again. The Distress phase of Search and Rescue procedures was initiated immediately after the pilot transmitted a Mayday call advising of the engine failure. An airborne aircraft was diverted, and was in the accident area within 4 minutes. It was joined within 15 minutes by an RAAF aircraft, and later by helicopters and boats. The search was hampered by approaching darkness, and by the survivors having no coloured flotation support or signalling equipment. Rescue of those remaining together was effected some three and a half hours after the ditching. The aircraft was subsequently salvaged. A detailed examination of the engine and fuel system revealed that the engine had failed from fuel exhaustion. It was considered likely that the pilot had commenced the flight with less fuel on board than he believed was the case. It was also possible that the fuel quantity gauges were giving erroneous indications, however the relevant transmitters were affected by impact and corrosion damage, and their previous serviceability could not be determined.

The aircraft was operating as United Airlines flight UA 815, a direct service from Los Angeles USA to Sydney. The planned flight time was 14 hours 38 minutes, and the aircraft carried fuel for 16 hours 17 minutes. Because of equipment unserviceability and crew duty time limitations, the planned departure time was delayed by some 13 hours. The aircraft had departed Los Angeles at the maximum brakes release weight of 315740 kilograms, which included 153300 kilograms of fuel. The maximum fuel capacity for the aircraft was 154960 kilograms. The flight proceeded without recorded incident and the crew contacted Sydney Arrivals Control at 1844 hours, when the aircraft was 200 nautical miles (370 kilometres) east of Sydney. The controller advised that holding would be required, with an expected duration of 10 to 15 minutes. The aircraft was subsequently cleared to descend from flight level 420 to flight level 210, with a requirement to reach the new level by 60 miles by Distance Measuring Equipment (DME) from Sydney. The aircraft entered a holding pattern at 60 DME at 1908 hours. Sydney Airport was closed to arriving aircraft at 1900 hours, because of thunderstorm activity. This information was relayed by the controller at 1905 hours by virtue of an all-stations call. However, as indicated by the subsequent transmissions, that call was not heard by the crew of the aircraft. The crew made two requests for an expected approach time, and were advised at 1912 hours that the delay was anticipated to be about 10 minutes. At 0923 the crew reported that a fuel emergency situation would be declared if clearance was not forthcoming in 5 minutes, and at 1925 the crew reported a remaining fuel endurance of one hour. They were advised that Sydney was closed to all operations because of thunderstorms. A fuel emergency was declared at 1928, and the crew requested approval to conduct an approach to Sydney. This was refused and the aircraft was directed to the RAAF base at Williamtown, where a safe landing was made at 1949 hours. Fuel remaining after landing was 7250 kilograms. Sydney Airport re-opened for normal operations at 1952 hours. Sydney Operations Control had imposed a requirement for arriving aircraft to carry an additional 30 minutes fuel endurance to allow for the thunderstorms at Sydney. However, the particular storm took over 2 hours to cross the Sydney Airport airspace. The Airport was closed from 1900 to 1918 hours, and from 1924 to 1952. It was evident that the 30 minutes fuel requirement was inadequate under the existing circumstances. United Airlines is one of three overseas based carriers which has been approved by the Department of Aviation to be responsible for its own operational control. This control is exercised from Chicago, USA. For aircraft on the Los Angeles - Sydney route, approval to proceed past a point located to the east of Norfolk Island is dependent on the fuel state of the aircraft and the expected weather conditions at Sydney. The necessary clearance had been obtained in this case. However, by subsequently closing Sydney Airport to all aircraft and denying the United flight the opportunity to conduct an approach, Sydney ATC over-ruled the operational control which the airline had been approved to exercise. When remaining fuel endurance is requested, Australian pilots are taught to provide a figure which relates to the total fuel on board. However, the United Airlines policy is for the pilot to give a figure based on the remaining flight fuel only i.e., the figure given does not include the 30 minutes mandatory reserve. This anomaly needs to be explored and resolved in order that pilots and ATC are aware of the actual fuel situation, so that priorities can be accurately established. The fuel policy of United Airlines provides for a minimum of 17000 pounds (7710 kilograms) to be available on arrival at Sydney. This is sufficient to allow the aircraft to go around from the landing approach in the event of the preceding aircraft blocking the intended runway, proceed to a suitable emergency alternate aerodrome (Williamtown), and land with the mandatory reserve fuel intact. The pilot in command had elected to divert to Williamtown before the last possible time for such a diversion. His decision was probably related to the advice given by ATC that Sydney was closed because of a stationary thunderstorm over the aerodrome.

Reccomendations:

As a result of the investigation of this incident, the Bureau makes the following recommendations.

1. The question of whether overseas carriers should abide by the fuel policy of the state of registry or the fuel policy of Australia should be resolved.

2. The right of ATC to deny approaches and otherwise impose operational control on overseas carriers which have been approved to exercise this control themselves should be examined.

3. The current policy of requiring 30 minutes extra fuel be carried when thunderstorms are forecast to affect aerodromes for periods of up to 30 minutes should be reviewed, in the light of the difficulty of accurately assessing the speed and duration of such storms.

4. That ATC ensure, that information regarding the change in status of an aerodrome has been received by all aircraft affected by that change.

The aircraft was engaged in dual instructional circuit training. On what was probably the fourth circuit, the aircraft was observed to land heavily on runway 35 and then to climb away to the N. When the aircraft reached a height of 400-500 ft a witness heard the engine noise cease abruptly and saw the aircraft enter a nose-high, yawing turn to the left. The left wing dropped as the aircraft approached a W heading and the aircraft entered a spiral dive which continued until it collided with the ground. There was no apparent attempt to recover from the spiral. The weather conditions at the time of the accident were fine and cool with a light wind blowing from the N. An examination of the wreckage failed to reveal any pre-existing damage or fault which might have contributed to the accident. However, the left ignition switch was found to be in the 'off' position and propellor damage was consistent with the engine being stopped at the time of impact. This evidence suggested the left ignition switch may have been inadvertently knocked to the 'off' position during emergency actions following a simulated engine failure, resulting in the engine stopping. The initial heading change after the abrupt cessation of engine noise, was consistent with an attempt to turn back towards the runway. There was no evidence found of either pilot having been incapacitated during the flight. It could not be determined who was controlling the aircraft at the time of the power decrease, or why the aircraft was handled in a manner which led to a loss of control. Recovery action may have been impeded by the shoe of one of the pilots fouling the rudder pedal control mechanism.

Significant Factors:

The following factors were considered relevant to the development of the accident

1. A loss of engine power during the climb after take-off was consistent with a simulated engine failure.

The pilot was conducting superphosphate spreading operations from an agricultural strip. Shortly after the aircraft became airborne on about the fifth or sixth takeoff for the day, the engine failed. The pilot immediately changed tanks and placed the fuel-pump switch in the high/prime position, but the engine did not respond. The pilot dumped the load and attempted a forced landing on steeply rising terrain. Touchdown was heavy, the nosegear collapsed and the aircraft overturned. The engine failure was determined to be caused by fuel starvation owing to fuel system mismanagement. It was established that the pilot had been pre-occupied with a personal problem, which distracted him from the task at hand. As a result, one tank had been allowed to run dry, and although the other tank had been full, a defective microswitch had prevented the fuel pump from operating in the high/prime position.

On returning from a flight in the local area, the aircraft was cleared for a straight-in approach. When the gear was selected down the in-transit light illuminated and stayed on. The gear warning horn sounded and a go-around was made from short final. Following a flypast, the Control Tower confirmed that the wheels were only partially extended. As the aircraft was climbing through about 700 feet, there was a surge of engine power and the aircraft yawed from side to side, mainly to the right. The pilot assumed that the right engine had failed, closed both throttles and made a gear-up landing on the grass alongside the runway. Initial investigation revealed that the gear motor circuit breaker had popped. This was a known fault with the aircraft, although the pilot had not been alerted to it. Although fuel was found in the right main and both auxiliary tanks, none remained in the left main and the left engine system was devoid of fuel. During his pre-flight inspection, the pilot had evidently over-estimated the quantity of fuel in the left main tank. He had limited experience on multi-engine aircraft and had not been formally checked on asymmetric handling procedures for some 4 years. Under the circumstances, he elected not to attempt to maintain height on one engine and concentrated on achieving a safe forced landing.