The helicopter was spraying chemical over a banana crop when it struck a powerline and heavily impacted the ground. The helicopter caught fire after hitting the ground.

The flight was planned as a one-hour inspection of tourist spots in the Isdell River area, 175 km north-east of Derby. The helicopter departed the base camp at Silent Grove, 15 km south-east of Mount Hart Station, landed once during the flight, and was returning to the base camp when the crash occurred. The crash site was 8 km north of the camp.

All occupants received serious injuries in the crash. One passenger, who appeared to be the most able, walked to the base camp to get help. He arrived there shortly after first light the next day. The wreckage, and the other occupants were found at approximately 1100 on the day after the crash. One passenger died during the night.

FACTUAL INFORMATION

Around lunch time on the day of the accident the pilot invited a number of people for a joy flight later in the day to view whales off the coast near Byron Bay. The flight was expected to take about 45 minutes.

The pilot and one passenger arrived at the aircraft at about 1630 EST, when the owner was repositioning the aircraft in its hangar. The other two passengers arrived about five minutes later. At that time, the pilot was conducting a pre-flight inspection and talking to the aircraft owner. Prior to the pilot completing his inspection, the owner departed. The owner did not note the quality of the pilot's pre-flight inspection, or whether the fuel tanks were tested for water.

The passengers boarded the aircraft and the pilot had them fasten their seatbelts. He then started the engine and taxied to the runway. One passenger considered that the pilot was in a hurry during this period; however, another pilot flying near the strip, and observing the take-off of the aircraft, considered that the aircraft spent an unusually long time at the eastern end of the airstrip. The take-off was towards the west at about 1653.

Witnesses reported that soon after the aircraft became airborne, when at a height of 70 to 100 ft above the strip and with the landing gear almost retracted, the engine rapidly lost power. A passenger reported that the pilot tapped a number of instrument faces, asked himself what was happening a couple of times, and appeared to become panicked. Pilots at the airstrip said that they expected the nose of the aircraft to be lowered but this did not occur. A short time later the wings began to rock before the left wing dropped and the aircraft collided with the ground beside the Pacific Highway. At impact, the aircraft was banked about 90 degrees to the left and was descending at an angle of about 30 degrees. The left wing broke into sections and the aircraft slid backwards to rest against trees beside the road. The four occupants were rescued from the aircraft before it was consumed by fire.  The pilot and the passenger who had been seated in the front seat were fatally injured.

The investigation found that the landing gear and flaps were up, with the left fuel tank selected at impact. The main tanks were reported to be close to full, with the wing tip tanks containing only residual fuel. Water was found in the fuel injection distributor valve on the engine. A small quantity of water was also found in the fuel control unit.

When the aircraft was ferried to Australia about a year before the accident, the ferry pilot noted that water collected in the left fuel tank after rain. The O-ring seals in the fuel caps were recorded as having been changed during maintenance to place the aircraft on the Australian register.

The aircraft had been parked in the open up to a few days prior to the flight and during that time, heavy rain had been experienced at the airstrip. The aircraft had been taxied to the hangar from its previous parking spot, although the owner could not remember which tank had been used. The investigation was unable to determine whether a fuel quality check was conducted prior to this flight.

Information provided by the aircraft manufacturer showed that not all water which entered the fuel tanks would have been extracted by normal fuel drain checks. The residual water remained in the tanks during normal flight manoeuvres but would have the potential to move from the tanks into the fuel lines given the right circumstances.

The pilot had flown from the airstrip on previous occasions, and in the aircraft owner's opinion, was well qualified with about 8,000 hours of flying experience. The owner had flown with the pilot on a number of occasions. Although the pilot held a perpetual Australian private licence for aeroplanes, his medical certificate had expired in July 1994 invalidating the licence.

An emergency locator transmitter (ELT) was fitted to the rear of the passenger compartment.  The owner believed the ELT to be capable of normal operation. The unit was destroyed by the fire and no reports of its activation were received from overflying aircraft.

ANALYSIS

The fuel distributor valve on the engine is the last point in the fuel system before the fuel runs through small pipes into the combustion chamber. Water in this valve would mean that water was being fed to the engine. Depending on the quantity of water in the fuel system, the presence of water would result in rough running or complete engine stoppage. On this occasion the engine stopped.

A suitable forced landing area was available if the aircraft had been landed straight ahead in the direction of the take-off. While the distance available may have been inadequate, it did offer an area suitable for a substantial part of a forced landing.

The aircraft's nose was not lowered from the initial climb attitude, and landing gear and flaps were selected up at impact. This evidence suggests that the pilot did not take appropriate emergency actions intended to control the aircraft after the engine failed.

SIGNIFICANT FACTORS

1. The pilot did not conduct an adequate pre-flight fuel inspection.
2. The engine failed due to water in the fuel system.
3. The pilot did not initiate appropriate actions to carry out a forced landing.
4. The aircraft collided with the ground after it lost flying speed.

The pilot reported a birdstrike during the descent. The strike appeared to cause the left engine to lose power, and it was shut down and secured. Subsequent investigation indicates that the wiring providing indications to the flight crew was damaged and the instrument indications, only, were lost.

An unsafe left main gear indication illuminated when the gear was selected down for landing. Although the standby system indicated three green lights, a single-engine flypast was completed to confirm the gear position.

The crew did not request emergency services. The company agents on the ground at Broome declined to declare a full emergency and local standby only was activated by airport management.

During the later part of the landing roll the pilot lost directional control of the aircraft. He reported brake and steering failure.

The aircraft veered off the sealed runway before the pilot stopped it by application of the emergency brakes.

Ground witnesses reported the loss of directional control may have been the result of the use of reverse thrust on the remaining operating engine.

There were 17 people on board.

Early on Monday morning (the day after the accident) a farmer reported finding the aircraft crashed on his property.

A Flight West crew at Roma had reported that the pilot of an aircraft was having difficulties in finding Roma at about 1835 on Sunday evening. The pilot was asking for a description of the aerodrome in relation to the township and requested that the Flight West crew wait on the ground until their arrival. When the Flight West passengers had disembarked, the crew attempted to contact the aircraft again, but without success.

The aircraft call sign was reported to Flight Service as VH-ECN. When SAR (Brisbane) found VH-ECN at Caboolture and no further reports of a missing aircraft were received, the INCERFA was cancelled.

Last light at Roma was 1804 approximately. The pilot was not rated for night visual flight rules.

FACTUAL INFORMATION

History of the flight

The pilot had commenced mustering at about 0700 EST. At about 1030 the aircraft was refuelled to full tanks during a "smoko" break. The pilot's intention was then to muster about 30 head of cattle which had been separated from the main mob. The pilot had agreed to take a passenger on the flight which was expected to be of relatively short duration. The passenger had been holidaying at the property and was keen to experience a helicopter flight.  The helicopter became airborne at about 1050 and was last heard at 1115. At 1130 a jillaroo realised she could no longer hear the helicopter and began making enquiries on a hand-held radio. When nothing was heard, she began a search on a trail bike and eventually discovered the wreckage and the deceased occupants at about 1420.

Impact sequence

The tail rotor had struck the top branches of a lone 7-metre-high sapling, causing one blade to separate. The tail rotor gearbox then separated, and the main rotor struck the tail boom. The wreckage fell to the ground 30 metres beyond the sapling. The right side of the cabin was crushed. There was no fire.

Wreckage examination

The cabin was crushed on the right (pilot's) side by ground impact. The Perspex bubble was scattered in front of the main wreckage, which was facing south-south-west. The main rotor blades showed evidence of having struck the tail boom and the cabin structure. The tail rotor drive shaft was recovered and showed evidence of torque twisting, indicating that the tail rotor drive shaft was being driven under power when the tail rotor contacted the tree.

The left side of the passenger seat with the seat lap belt attachment, had detached from the fuselage structure and this allowed the passenger to strike the upper door frame during impact. The control systems were examined and appeared to be functioning normally. The engine governor switch in the end of the collective control was found in the off position.

The engine was removed and examined. There were no defects found which would have precluded normal operation.

Weight and balance

The weight and balance of the helicopter was within the limitations published in the aircraft flight manual.

Emergency locator transmitter (ELT)

The Ack Technologies ELT was found undamaged in the mounting bracket at the rear of the engine bay. The arming switch was found in the off position.

ANALYSIS

Civil Aviation Regulation Section 29.10 states that during aerial stock mustering operations, a pilot shall not carry more than one other person and that that person must be essential to the successful conduct of the operation. In this case, the passenger was not essential to the conduct of the operation. The helicopter had been refuelled to full tanks immediately before the flight and, although within the specified weight limitation, it was much heavier with the additional weight of the passenger than the pilot was accustomed to for mustering. Although the pilot was highly experienced, the resultant reduction in performance may have been a factor in his being unable to avoid the collision with the tree.

There were no witnesses to the accident and the final flight path before the collision with the tree could not be determined. However, it was evident that immediately after the tree was struck by the tail rotor, the helicopter was subjected to violent manoeuvres.

The RPM governor is fitted to the engine to prevent decay of rotor RPM when manoeuvring. The aircraft flight manual states that flight is prohibited with the governor switched off, except when there is a system malfunction or for emergency procedures training. The governor switch was found in the off position, but it may have been bumped to this position in the accident sequence. If it was deliberately switched off for the flight, manoeuvring performance of the helicopter may have been reduced.

The possibility that the pilot may have been attempting a precautionary landing for some reason such as an engine malfunction, was considered. There was no evidence found to substantiate such a possibility.

SIGNIFICANT FACTOR

The tail rotor struck a tree, and this precipitated a major structural failure. Why the pilot was unable to avoid the tree could not be determined.

FACTUAL INFORMATION

History of the flight

The pilot had been tasked to ferry the helicopter from Windorah to Cadelga Station in preparation for cattle mustering. Several witnesses saw the helicopter depart at about 1620 EST. It was last seen some time after 1630 as it flew past South Galway Station, 61 km south-west of Windorah.

The track distance was 239 km with an expected flight time of about 1 hour and 40 minutes and an estimated time of arrival of 1800. The helicopter was expected at Cadelga Station before last light at 1829. When it had not arrived, the helicopter was reported missing. An extensive aerial search commenced at first light the next morning. The wreckage of the helicopter was found about 2 km to the right of the direct track and 31 km short of the destination.

Pilot information

The pilot was 20 years of age and had commenced his flying training on Bell 47 helicopters in March 1995. He gained an endorsement on Robinson R22 helicopters in May 1995. He passed his commercial helicopter licence test on 22 June 1995. In February 1996, the pilot passed a check flight with an instructor and obtained mustering training. Since then, he had been employed as a mustering pilot on a full-time basis.

The post-mortem examination report did not reveal any pre-existing physiological problems.

Wreckage examination

The helicopter wreckage was examined initially at the accident site and later transported to a maintenance facility at Archerfield Airport where a more detailed examination was possible. These examinations revealed that the helicopter had impacted in an attitude about 60 degrees nose-low and banked to the right. The skid gear had separated at impact and the remaining structure of the helicopter was severely compressed.

Significantly, the main rotor had ceased rotating, and the engine had stopped by the time the helicopter struck the ground. Sections of the tail boom were scattered to the right of the main wreckage. Heavy items, such as a jerry can full of oil and the tail rotor assembly, were strewn ahead, along track (approximately 245 degrees M). Lighter items, such as paint flakes and Perspex pieces, were scattered downwind, generally to the north of the main wreckage.

Examination of the main rotor blades found that both pitch-link rods were broken in overload and that both blades were bent into an "S" shape. The tail boom had suffered several main rotor blade strikes. The first blade struck on the flat with the top surface, distorting the tail boom and swinging the tail rotor assembly into the main rotor disc.  The cabin was also struck by one or both main rotor blades.

The engine was bulk stripped at an engineering workshop. Nothing was found that would have prevented the normal operation of the engine. Examination of the exhaust manifold showed that the engine was still hot at impact. Both fuel tanks were holed, subsequently only a small quantity of clean fuel was recovered. The helicopter had undergone a periodic maintenance inspection prior to the flight.

Examination of the helicopter wreckage and of the maintenance documentation did not reveal any abnormalities which could have led to a loss of control.

Weather and environmental issues

An assessment of the weather conditions was obtained from the Bureau of Meteorology and local sources. The wind at 2,000 ft was a southerly at 10-15 kts. The surface wind was a light south-south-easterly, less than 5 kts. Cloud was scattered cumulus at 3,000-4,000 ft with patches of higher altocumulus. Sunset at the crash site was at 1808.

Robinson Helicopter Company research

Research undertaken by the Robinson Helicopter Company found that when the main rotor RPM decreases below 75% RPM, it will continue to decrease regardless of input from the pilot. In a short time span, the engine will stall, and the rotor system will stop completely. During flight with rotor RPM in the normal range, the blades are relatively straight due to centrifugal force. As rotor RPM decreases below the normal range, the main rotor blades bend upwards in a permanent set, the pitch link rods break in overload and the droop stop tusks also break or bend. The main rotor blades are then free to swivel at the mast and flail, striking the tail and cabin structure.

Cabin luggage

Any luggage has to be carried in the cabin of the helicopter: in the space under each seat, on the passenger's seat, or on the cabin floor in front of the passenger's seat.

The pilot carried three bulky items in the cabin: a swag (bedroll), a full 20-L jerry can, and a small overnight bag. The rolled-up swag was later measured at 1,060 mm long by 350 mm in diameter. It was held upright by the lap/sash seat belt on the passenger's seat. The location in the cabin of the 540 x 300 x 200mm overnight bag could not be determined. The jerry can containing engine oil was ejected from the helicopter, indicating that it was not restrained and probably lying on the cabin floor in front of the passenger's seat.

The swag and overnight bag were removed from the wreckage before the investigation team arrived on site. Therefore, any evidence of control interference caused by the luggage was lost. The Bureau's records show that since 1989 there have been four other similar occurrences, three of which resulted in accidents. In each occurrence, control was lost either totally or partially, due to inadequately restrained cargo in the cockpit. The 1989 accident was due to the in-flight movement of a swag and other equipment, held only by the passenger's seat belt.

Emergency locator transmitter

An emergency locator transmitter was not fitted to the helicopter, nor was one carried by the pilot.

ANALYSIS

Loss of control

The loss of main rotor RPM and engine stoppage was preceded by an event which resulted in loss of control. This event was sudden and severe enough to prevent the pilot from taking timely corrective action. An assessment of the helicopter's attitude at impact and the wreckage distribution indicated that this event occurred at an altitude in excess of 500 ft.

There was no physical evidence found to indicate that the swag shifted and interfered with the flight controls.

The event which led to a loss of control could not be identified.

SAFETY ACTION

Since 1989 there have been four similar occurrences, three of which resulted in accidents. In each occurrence, control was lost either totally or partially, due to inadequately restrained cargo in the cockpit.

The Bureau of Air Safety Investigation is considering an article for publication in Asia-Pacific AIR SAFETY highlighting the risks to safety from carriage of cargo in the cockpit of the Robinson R22 helicopter.

During climb out from Dauan Island (Torres Strait), the helicopter suffered an engine power loss. The pilot placed the helicopter in autorotation from approximately 300 feet. Having heavily impacted the water, the helicopter sank and rolled inverted. The pilot and three of the four passengers escaped. The remaining passenger remained in the aircraft as it sank. One of the passengers who survived the ditching did not remain afloat and disappeared from the other survivor's view. The aircraft sank and has not been recovered.

At 2005 EST the pilot conducted an instrument letdown at Flinders Island and then notified air traffic services that he was proceeding night VFR procedures along the coast to Killiecrankie, a private airstrip approximately 15 NM to the north. He made an operations normal transmission at 2020 and advised he would call again by 2045. No further transmissions were received. An air and ground search was commenced.

FACTUAL INFORMATION

History of the flight

The pilot was conducting a charter flight from Roma to Quilpie, Windorah and Tanbar Station 100 km south-west of Windorah. The aircraft was refuelled at Windorah with 170.5 L of Avgas and departed Windorah at 1400 EST for Tanbar.  At 1739 the pilot transmitted flight-plan details to Brisbane Flight Service by radio for the flight to Roma. He advised that the flight would be conducted under the visual flight rules (VFR), and that the aircraft endurance was 250 minutes. He nominated a SARTIME of 2100 for his arrival at Roma.

At 1903 the pilot made an "all stations" broadcast 20 NM west of Charleville. He reported inbound on the 270 VOR radial on descent for a practice VOR approach and said that after a missed approach he would proceed to Roma.

Witnesses at the airport saw the aircraft fly overhead from the west. The aircraft was seen to turn right onto a southerly heading and soon afterwards the sound of the aircraft diminished. A bang was then heard and felt through the ground at about 1915. The aircraft wreckage was located the next day by a search party. The aircraft had struck the ground whilst banked vertically to the right with a 45-degree nose-down attitude, and disintegrated.

Pilot in command

The pilot was correctly licensed and endorsed to conduct the flight. Prior to the accident flight, the pilot had only 3.9 hours multi-engine command night experience. Although 6.5 hours single-engine dual night experience was recorded, no multi-engine dual night experience was recorded. The pilot had flown the aircraft in command at night on only three previous occasions. These were on 11 and 12 December 1995 (2.1 hours), and on 4 April 1996 (1.8 hours). The pilot held an instrument rating, but the flight was being conducted under night VFR.

Meteorological conditions

The meteorological aerodrome report (METAR) for Charleville on 16 April 1996 at 1900 hours reported the wind to be a southerly at 6 kts, visibility more than 10 km, and no cloud below 5,000 ft. Witnesses said it was a very dark night with no moonlight.

The VHF omnidirectional radio range (VOR) approach

The pilot broadcast his intention to carry out a practice VOR approach at Charleville when he was 20 NM to the west. The VOR approach is designed to allow an aircraft to descend on specified VOR radials to a specified minimum descent altitude (MDA) in instrument meteorological conditions (IMC). The Charleville runway 12 VOR approach MDA is 1,750 ft or a height above the aerodrome of 727 ft.

To commence the procedure when approaching Charleville from the west, a sector entry is carried out. This entails passing over the aid (VOR) at the initial approach altitude of 2,800 ft and turning right onto a heading of 146 degrees for 1 minute. The aircraft is then turned right to intercept the inbound leg of the holding pattern to overhead the aid. The position of the accident site suggests that this procedure had not been carried out, and that control of the aircraft was lost in the initial right turn from overhead.

Fuel quantity

The pilot refuelled at Windorah before proceeding to Tanbar. The main and auxiliary tanks were filled. No fuel was added to the tip tanks. The main tanks contained 204 L, and the auxiliary tanks 113.5 L, of useable fuel.

Engine instrument indications

During the on-site investigation the engine tachometers were recovered. The left engine tachometer was indicating 2,700 RPM, which is red-line or maximum RPM. The right engine tachometer was indicating 1,600 RPM, which is approximate flight-idle RPM. Both indicators were jammed in position by impact damage.

Fuel selector positions

The left and right fuel selectors and valves were recovered from the wreckage and specialist examination was carried out to determine the selector positions at impact.

The right fuel selector was found selected to MAIN. The selector pin was found secure in the handle although the selector knob was broken off. The pin was positively in the MAIN detent position. The right fuel-selector plate was minimally deformed with two of the three screws securing the plate to its base still in place. This means that the handle probably stayed in contact with the selector plate. That the pin remained, indicates that the selector was in MAIN for the whole impact sequence.

The left fuel-selector handle was found slightly anticlockwise from the MAIN position towards the AUX position. The left fuel-selector handle pin was deformed and depressed into the handle. The handle was thus free to move out of the detents. Examination of the left fuel selector showed deformation of the selector area, the handle subject to impact movement, multiple impact marks around the MAIN position, deformation of the selector plate around MAIN, a clear imprint of the handle above MAIN and an imprint in the plastic base of the selector in the MAIN position. The examination indicated that the left fuel selector was selected to MAIN at impact.

Exhaust pipe examination

Exhaust pipe sections from both engines were examined to assess the temperature at impact. Both exhaust pipe sections exhibited straw/gold coloured heat tinting. A temperature cannot be assigned accurately to a heat tinting colour, since the colour varies not only with temperature but with the time at that temperature. However, the presence of heat tinting does indicate that the exhaust pipes were hot (above 350 deg. C) at the time of impact, and that the engines were operating immediately prior to impact.

Aircraft and engines

Examination of the wreckage did not reveal any defects which may have contributed to the accident. There were no mechanical defects found on the engines which would have prevented the engines from developing normal power.

Emergency locator transmitter

The ACK Technologies emergency locator transmitter (ELT) (which complied with TSO C91a) was found outside the main wreckage unattached. The case was intact, and the three-position (ON-OFF-ARM) function switch was in the centre OFF position. The switch was not guarded and may have been moved in the impact sequence. The unit was functionally tested and found serviceable.

ANALYSIS

At the time of the accident, there was no moon, and the aerodrome pilot activated lighting (PAL) had not been turned on. After passing over the township, which is to the north of the aerodrome, the pilot would have had no visual horizon. The pilot's multi-engine experience at night was 3.9 hours, all of which was in command.

The flight times since the last refuelling at Windorah to arrival overhead at Charleville corresponded to that required to exhaust auxiliary fuel tanks. The pilot was known to have allowed auxiliary tanks to run dry before selecting mains on previous occasions. The fuel supply to the right engine may have been interrupted due to exhaustion of the right auxiliary tank. The operating handbook cautions against using auxiliary tanks in other than level flight due to the possibility of uncovering the tank outlet. Should this occur the engine is likely to lose power, surge and stop. Once the fuel system has ingested air, the engine cannot be restarted until the air is purged and a normal fuel flow restored. The fuel selectors appear to have been selected to the main tanks at impact, but as indicated by the engine tachometer readings, the right engine was not delivering power. This was most likely due to the right engine fuel system having ingested air before the main tank was selected.

The possibility of the pilot carrying out a deliberate asymmetric approach was considered. However, this would seem unlikely due to the demanding nature of the exercise and the pilot's low experience on type at night.

An unexpected power loss while the pilot's attention is concentrated on the flight instruments could be most distracting, even for an experienced pilot. The effect would be for his attention to be immediately diverted to the engine instruments, and then possibly the fuel panel. Cross reference between the attitude and performance instruments is required to perform instrument flight, particularly when there is no visual horizon. This is critical in multi-engine aircraft if an engine fails and asymmetric flight is encountered. Should cross-reference be lost for any reason and the aircraft allowed to get into unbalanced, uncoordinated flight, the aircraft may assume an unusual attitude. The pilot may then become completely disorientated and lose control of the aircraft.

The aircraft attitude at impact suggests that this occurred.

SIGNIFICANT FACTORS

1. The pilot was inexperienced on multi-engine aircraft at night and had not undergone night flying training on the aircraft type.
2. The aircraft carried sufficient fuel for the flight, and it is likely that auxiliary tank fuel was depleted or nearly depleted when the aircraft arrived overhead Charleville.
3. The weather was fine, but with no moon and no visible horizon, was unsuitable for VFR operations at night.
4. The pilot was conducting a practice VOR approach at Charleville.
5. The right engine was not developing power, most probably due to fuel starvation.
6. The pilot lost control of the aircraft for undetermined reasons during a practice instrument approach and the aircraft impacted the ground.