1. The pilot in command did not adequately convey his approach intentions to the co-pilot.
2. The co-pilot, while concerned, did not adequately communicate his concerns to the pilot in command.

Although the crew had briefed a night visual approach to Norfolk Island, the unicom advice regarding rain showers in the area was clearly inconsistent with the forecast they held. Investigation of the occurrence identified cockpit crew coordination deficiencies that included ineffective question-and-answer communications, and crewmembers performing tasks in isolation. The crew's performance during the approach might have been more effective had they acted in a manner consistent with recognised principles of crew resource management to ensure that all resources available to the crew were fully and effectively utilised. That the perceptions of the pilot in command and co-pilot were different was in itself an indication that there was a breakdown of communication and coordination between them.

The co-pilot believed that the pilot in command did not clearly communicate his approach intentions during the final approach phase. In effect, the co-pilot was left out of the pilot in command's decision-making loop. This generated a conflict in expectations between the pilot in command and co-pilot that was not resolved. As a result, and in view of the prevailing weather, the co-pilot became concerned for the safety of the operation.

The approach briefing was inadequate. As the crew had flown frequently together during the previous two years, they should have been very familiar with one another's usual practices. The pilot in command could have requested explicit feedback from the co-pilot to assess his understanding of the situation. The crew's familiarity with each other and their destination probably contributed to the brevity of the approach briefing.

The operator had not ensured a uniform standard of crew resource management across the company. Moreover, the safety regulator had not provided operators engaged in multi-crew air transport operations with specific guidance regarding crew resource management. Consequently, the operator was left to assess issues such as syllabus content, standard of training, qualifications of CRM assessors and frequency of assessment.

The pilot continued to fly after consulting a doctor regarding severe headaches with vision disturbances.

The aircraft

A detailed examination of all available parts of the helicopter concluded that the engine was most likely to have been operating and driving the main and tail rotors. Consideration was given to failures that could cause an uncommanded descent or turn in flight. However, the investigation considered it unlikely that an experienced pilot would allow the helicopter to smoothly descend to collision with the ground.

No evidence was found of any pre-existing defect or malfunction to the aircraft that may have contributed to the development of the accident.

The flight path

Under the prevailing conditions, the investigation considered that the track flown by the helicopter was in keeping with the pilot's normal routine. The unusual aspect of the flight path was its inappropriately timed descent.

The helicopter was reported to be in a smooth flight path, with no jerking and no abnormal sounds such as would be produced with an engine power change or sudden control inputs. On impact with the trees, the helicopter was in a steady flight regime, laterally-level and descending at about 1,200 ft/min. This could indicate that the pilot was maintaining his grip on the cyclic control. If the cyclic control was released, it would most likely have fallen to some abnormal position in a short time. This could have caused the helicopter to adopt an unusual attitude very quickly. The conclusion, therefore, is that the pilot may have been able to maintain control in some form or other: from fully conscious, to a situation where he had managed to "freeze" on the controls and simply maintain them in position.

If the pilot was fully conscious, then the flight path may have resulted from a distraction. However, no evidence was found of anyone making a phone call to the aircraft around the time of the accident. In any case, it is unlikely that an experienced pilot flying at low altitude toward rising terrain would allow himself to become distracted for the time involved in this accident sequence (about 15 secs) and not pay any attention to the flight path.

The pilot

The pilot had visited a DAME a few weeks prior to this accident. This was as a result of headaches that had increased in severity and involved vision disturbances. The ENT specialist found a suspected sinus infection and prescribed a course of antibiotics. The investigation found no underlying personal factors, other than the headache problem, that may have played a role in this accident sequence.

There is no evidence that the pilot complied with the relevant regulations, other than his initial referral of the problem to a DAME. CAR 6.16A imposed a condition on the pilot that he not resume flying duties until being cleared to do so. It is likely that the pilot was reluctant to cease his normal duties unless strongly advised. The DAME had not yet reported the pilot's medical situation to CASA. Consequently, CASA were not in a position to take more positive control of the case.

The pilot was flying the Bell Jetranger from Archerfield to his home base at Channel 10 on Mount Coot-tha. The flight continued normally until passing the Channel 9 buildings situated on the southern end of Mt Coot-tha. Approaching that area at 1,000 ft, the pilot apparently intended to track to the western side of Mt Coot-tha for a landing at the Channel 10 pad. When passing abeam the Channel 9 tower at about 850 ft, the helicopter was seen to enter a descending right turn until it struck trees and the ground 600 ft above sea level on the western slopes of Mt Coot-tha. The helicopter collided with the trees in a laterally-level attitude, with no roll rate, a descent angle of about five degrees, and a speed estimated to be its normal cruise speed. A fire erupted immediately and consumed the helicopter. The pilot was fatally injured as a result of the impact.

Accident locality

Mount Coot-tha is the location of the television transmitting towers for the Brisbane area. Channel 9 is the southernmost tower and Channel 10 is the northernmost tower. Each television station has a helicopter landing site and associated flight paths. As a standard practice, each television station pilot was aware of the others' normal approach and departure procedures and avoided the relevant areas during flights.

The pilot involved in this accident had developed a practice of monitoring the movements of the Channel 9 and Channel 7 helicopters: he would check on their presence while flying to or from his landing site, if the intended flight path took him near the other stations. On the accident flight, this could have been achieved by flying to the west of the mountain.

After passing abeam the Channel 9 area at about treetop height, the route to Channel 10 involved crossing a ridgeline. The ridge was about the same height as the mountain and would have required the helicopter to maintain altitude or climb slightly to cross the ridge.

Weather conditions at the time were suitable for the flight. The wind was light, with an easterly to south-easterly tendency.

Personnel information

The pilot held an Airline Transport Pilot Licence (Helicopter) with a current Class 1 medical certificate, with vision correction required. He had accrued considerable aeronautical experience in both aeroplanes and helicopters. Of this, 7,566.2 hours had been gained in helicopters. His most recent competency check had been a Biennial Flight Review conducted on 12 and 13 August 1998. His most recent medical examination for the issue of a flight crew medical certificate had been conducted in June 1998.

Flight and duty time records maintained by the pilot indicated that he flew the accident helicopter regularly. In addition, he regularly flew a fixed-wing aircraft, normally at fortnightly intervals. His flight and duty time sheet recorded having specialist medical appointments on 10 and 14 August.

Little examination of the pilot's activities and eating habits in the few days prior to the accident was possible. The pilot had recently been under some personal emotional stress. However, information received from colleagues and friends indicated that the current life stress did not appear to have impaired the pilot's routine behaviour and functioning.

Aircraft information

The helicopter was a Bell 206B Jetranger, serial number 1946, manufactured in 1976. It was maintained for operations in private, air work, and charter categories involving flights under the visual flight rules by day or night. Up to the morning of the accident flight, it had accrued 4,532.8 hours time in service.

Three months before the accident, an entry in the maintenance required section of the maintenance release stated "high pitched noise / vibration through airframe at speeds 100 kt +". The clearing endorsement indicated that the helicopter was inspected thoroughly and the inspection doors on the right and left sides had been reshaped to fit more snugly. The entry contained a pilot's clearing signature for the subsequent test flight.

The high speed / high frequency vibration problem was reported to have persisted for some years prior to the accident and attempts to positively isolate the problem had been unsuccessful. Its intermittent nature made troubleshooting difficult. The noise was reported to be present in cold air conditions more consistently than in warm conditions. After considerable discussion with pilots who had experienced, or had attempted to induce the problem, the investigation concluded that the problem was most unlikely to have arisen on the accident flight. In any event, the problem was considered to be unrelated to the controllability of the helicopter.

An examination of all maintenance records did not reveal any other aspect considered likely to have contributed to the accident.

Communications

The helicopter was fitted with a mobile telephone and the pilot routinely carried a pager. Answering the phone involved manipulating a panel on the centre console area to the left of, and slightly behind, the pilot. No evidence of any calls made to the telephone or the pager was found.

Recorded voice communications and radar data for the flights to and from Archerfield were examined. The data indicated that the helicopter was flown normally during the flights.

Wreckage and impact information

The helicopter's initial impact was with the top limbs of a dead tree, followed soon after by collision with a large gum tree. The main rotor blades severed the top half of the tree and the airframe shattered the bottom half, allowing the top half to fall beside the tree stump. Loud bangs heard by witnesses were consistent with these impacts and also with the outbreak of the ensuing fire. The helicopter's speed at the time of impact was estimated to be around 110 kts, its normal cruising speed. Its descent angle of five degrees was consistent with a descent rate of about 1,200 ft/min.

Although the fire consumed most of the airframe, there were sufficient parts available for specialist examination. The investigation concluded that the engine was operating at impact, and the engine was driving the rotor systems. Rotor control systems were also intact, as far as could be examined. A trailing edge balance weight at the inboard end of a main rotor blade was not found. Specialist examination determined that the mounting point at the outboard end of the weight had been fractured for some time prior to the accident flight and that the inboard mounting point had also developed a fatigue crack. The investigation could not establish whether the balance weight finally became detached as a result of impact during the accident sequence or prior to impact. Representatives of the helicopter manufacturer considered that the absence of the balance weight would not have affected the helicopter's controllability. The investigation found no pre-existing defects likely to have contributed to the accident.

Helicopter controllability

The investigation considered a number of possible failures that could have been encountered during the flight. Witness reports were consistent with recorded radar data that indicated the flight path was smooth, with no unusual noises or abrupt movements of the helicopter. At impact, the helicopter was laterally-level and at high speed. The cyclic control movement needed by a pilot to transition from level flight to a descent angle of about five degrees (corresponding to a descent rate of 1,200 ft/min) was reported to be about 12 mm. The investigation was advised that an experienced pilot would normally be aware of this amount of cyclic control movement.

Loss of tail rotor control in the cruise was considered and assessed as not being a significant immediate problem, due to the speed of the helicopter at the onset of the accident sequence. The pilot could have turned away from the mountain towards the valley to the left, where appropriate decisions could be made without the need to avoid terrain.

Possible jamming of the hydraulic system associated with the main rotor controls was considered. If a failure caused a control deflection fully one way suddenly, an abrupt flight path deviation and change in aircraft noise would be expected. Similarly, a jammed control should also be likely to produce some abrupt movements, at least initially. Both problems should have been controllable by the pilot as the aircraft hydraulics are designed to be overpowered.

Had engine power loss or surging occurred, these should have produced some audible changes in noise from the engine and the rotors. A turn initiated by the pilot toward the lower ground, climb and slowing of the helicopter would also be expected.

A bird strike, or some other event affecting the pilot, was also considered. If the pilot had been incapacitated to the extent that he was unable to control the helicopter, then the flight path would be expected to have changed in some way. Experienced pilots interviewed during the investigation indicated that if the cyclic control is released it should remain in position for a few seconds and then start to fall in a random direction. The rate of change could increase if the control diverged from the central point. This could produce obvious changes in aircraft noise and flight path.

Medical information

Medical evidence provided to the investigation indicated that the pilot had suffered from a subarachnoid haemorrhage, for which no bleeding vessel could be found, in June 1994. His flight crew medical certificate had been cancelled as a result of that event. A Class 2 medical certificate was issued as a result of a medical examination in June 1995. His next medical examination was in July 1996 and Class 1 and 2 medical certificates without restriction were issued in August 1996. Further routine flight crew medical examinations were subsequently passed in June 1997 and June 1998.

On 28 July 1998 the pilot attended a designated aviation medical examiner (DAME), reporting that he had begun to experience severe, migrainous type headaches with blurred vision and instances of double vision. The DAME considered that there was a strong possibility that emotional stress was the cause. Since the symptoms did not match the normal indications of migraine headaches, the doctor referred the pilot to a neurologist for specialist examination. The DAME also indicated that the patient was a pilot and asked for advice concerning whether he should continue flying (the pilot was keen to continue flying). The pilot initially consulted the neurologist on 5 August and a follow-up meeting was held on 10 August. With no neurological problems evident as a result of that examination, he was then referred to an ear nose and throat (ENT) specialist who diagnosed a severe sinus infection. The specialist prescribed a course of antibiotics. That visit had taken place on 14 or 17 August. No information on the speed of onset of the headaches was available.

The neurologist involved in the 1994 event was the same person involved in the pilot's recent specialist examination, and had concluded that the pilot's current symptoms had not been related to the pilot's previous medical history.

Evidence available to the investigation indicated that the pilot had suffered a headache early in the morning of the accident flight and some Channel 10 staff reported that he did not appear to be well on arrival at work. On the other hand, other people familiar with the pilot had attended the meeting at Archerfield and reported that he seemed normal at that time.

Post-accident advice from the Civil Aviation Safety Authority aviation medicine staff indicated that, based upon information obtained from the medical practitioners, the pilot had experienced a change in his medical condition so that he no longer met the required medical standard.

The limited post-mortem information available did not assist with an assessment of the pilot's physiological state at the time of impact. There was insufficient post-mortem evidence to determine if any neurological anomalies had contributed to the accident.

Medical regulations

Considering that the pilot had experienced recent medical problems and had consulted a DAME, the relevant Civil Aviation Regulations (CARs) were examined. Anecdotal evidence indicated that the DAME could have advised the pilot as to whether or not he could continue with his flying duties. CAR 6.16A also indicated that the pilot was not permitted to fly, pending a resolution of his medical situation.

CASA normally provides each DAME with more detailed guidance on the matters considered significant to aviation. The neurological section described different forms of headaches and the considerations involved with each type. In relation to the type of headache considered likely to involve this pilot, the guidance stated, "Such migraines are characterised by long periods of remission and capricious onset, and may completely incapacitate the sufferer. All cases will be considered on an individual basis."

1. The approach controller changed the approach sequence for TNU.
2. The sight and follow instructions issued to TNU provided insufficient and inaccurate information.
3. The crew of TNU misidentified EWM as CZA.
4. Neither TNU nor CZA was equipped with TCAS.

In the circumstances, the controller's decision to change the assigned runway for TNU was inappropriate and not in accordance with Local Instructions. It resulted in additional workload by way of the vectoring instructions, as well as the setting up of the sight and follow procedure, for TNU. There were other options for reducing workload such as vectoring the traffic survey aircraft away from the area. Ultimately, the decision set the scene for the conflict to develop.

The information the controller gave the crew of TNU concerning the aircraft they were to sight and follow was incomplete. Because the crew was not given its number in the approach sequence, there was no reason for other than one aircraft ahead to be sighted. Also, because of the speed differential between the two aircraft, the azimuth information provided by the controller rapidly became inaccurate. These oversights resulted in the misidentification by the crew of TNU of EWM as CZA.

The report by the crew that the traffic had been sighted, and the immediate acceptance by the controller of this information, was, in effect, one assumption reinforcing another. In other words, the crew said what the controller wanted to hear. This had two main effects. It:

1. confirmed to the crew that they had correctly identified the traffic, effectively implying that there was no other potentially conflicting traffic, and
2. it enabled the controller to divert his attention away from TNU and CZA to the control of other traffic.

Had either TNU or CZA been traffic alert and collision avoidance system equipped, it is likely that either or both crews would have become aware of the confliction before separation standards were infringed. It is also probable that the seriousness of the occurrence would have been less if the TNU cabin crewmember had reported her observation of CZA to the flight deck crew.

1. The pilot was operating the aircraft in instrument meteorological conditions below the approved minimum descent altitude.
2. Low cloud was covering Gregadoo Hill at the time of the accident.

The pilot had received an accurate appreciation of the weather conditions in the vicinity of Wagga Wagga prior to departing Albury. At that stage it would have been apparent that low cloud and poor visibility were likely to affect the aircraft's arrival. Under such conditions it would not have been possible to land from the GPS arrival procedure.

As the reported cloud base and visibility were both below the minimum criteria, it is difficult to rationalise the pilot's transmission that, according to the latest weather report, he would be visual at the minimum descent altitude. This statement suggests that the pilot had already made the decision to continue his descent below the minimum altitude for the procedure and to attempt to establish visual reference for landing.

Based on the report of broken low cloud in the vicinity of the aerodrome, the pilot would have needed to descend to 1,324 ft above mean sea level to establish the aircraft clear of cloud. This is within 50 ft of the last altitude recorded on the GPS receiver.

Due to the difference between the actual and forecast QNH, the left altimeter would over-read by approximately 150 ft. At the time of the occurrence an otherwise correctly functioning instrument would have indicated an altitude of approximately 1,400 ft.

The pilot had probably set the right altimeter to the local QNH prior to departing Albury. As this setting also corresponded to the actual QNH at Wagga Wagga, that instrument would have provided the more accurate indication of the aircraft's operating altitude. However, because of its location on the co-pilot's instrument panel, it is unlikely that the pilot would have included that altimeter in his basic instrument scan.

It was not possible to assess the extent to which illicit drugs may have influenced the pilot's performance during the flight and affected his ability to safely operate the aircraft.

The aircraft operator had been contracted to provide a regular service transporting bank documents, medical pathology samples and items of general freight between Wagga Wagga, Albury and Corowa. On the day of the accident a passenger was accompanying the pilot for the day's flying.

The pilot commenced the flight from Corowa to Albury under the Visual Flight Rules, flying approximately 500 ft above ground level. At Albury he obtained the latest aerodrome weather report for Wagga Wagga, which indicated that there was scattered cloud at 300 ft above ground level, broken cloud at 600 ft above ground level, visibility restricted to 2,000 m in light rain and a sea-level barometric pressure (QNH) of 1008 hPa.

At 1715 Eastern Standard Time (EST) the aircraft departed Albury for Wagga Wagga under the Instrument Flight Rules. The pilot contacted the Melbourne en-route controller at 1728 and reported that he was maintaining 5,000 ft.

Although the aircraft was operating outside controlled airspace, the en-route controller did have a radar surveillance capability and was providing the pilot with a flight information service. However, no return was recorded from the aircraft's transponder and at 1732 the pilot reported that he was transferring to the Wagga Wagga Mandatory Broadcast Zone frequency. This was the pilot's last contact with the controller.

Although air traffic services do not monitor or record the Wagga Wagga Mandatory Broadcast Zone frequency, transmissions made on this frequency are recorded by AVDATA for the purpose of calculating aircraft landing charges. This information was reviewed following the accident.

The pilot broadcast his position inbound to the aerodrome on the mandatory broadcast zone frequency and indicated that he was conducting a Global Positioning System (GPS) arrival. He established communication with the pilot of another inbound aircraft and at 9 NM from the aerodrome, broadcast his position as he descended through 2,900 ft.

Approximately 1 minute and 20 seconds later, the pilot advised that he was passing 2,000 ft but immediately corrected this to state that he was maintaining 2,000 ft. He also stated that it was "getting pretty gloomy" and that according to the latest weather report he should be visual at the procedure's minimum descent altitude. The aircraft would have been approximately 6 NM from the aerodrome at this time. This was the last transmission heard from the pilot.

The resident of a house to the south of Gregadoo Hill sighted the aircraft a short time before the accident. He was standing outside his house and stated that the aircraft was visible as it passed directly overhead at what appeared to be an unusually low height. The aircraft then disappeared into cloud that was obscuring Gregadoo Hill, approximately 350 m from where he was standing. Moments later he heard the sound of an impact followed almost immediately by a red flash of light. The noise from the engines appeared to be normal up until the sound of the impact.

The aircraft had collided with steeply rising terrain on the southern face of Gregadoo Hill, approximately 40 ft below the crest. The hill is 4 NM from the aerodrome and is marked on instrument approach charts as a spot height elevation of 1,281 ft. The estimated time of the accident was 1739.

The pilot and passenger sustained fatal injuries.