Loss of separation

The duty runways at Perth were 06 and 03 for departures and 03 for arrivals. There were a number of aircraft inbound, including G-BNLW. Only one, VH-NGA, was taxiing for departure when a wind change occurred necessitating a change of runways. Following coordination between the tower, the approach east controller, the approach procedural controller and the flow controller, the duty runways were changed to runway 21 for departures and runways 21 and 24 for arrivals. During this coordination, the flow controller questioned the need for the change, advising that they had a sequence of seven aircraft (being sequenced for runways 03/06).

After the runway change was effected G-BNLW was advised of the change and to expect vectors for an ILS approach to runway 21. G-BNLW was inbound to Perth from the west and at the time of that advice had about 30 track miles to run. The aircraft was under the control of the approach west controller. VH-NGA was permitted to continue taxiing for runway 03, even though the aircraft was at a position where it would have been just as convenient to taxi for runway 21. Shortly afterwards G-BNLW was given a heading change to position for a runway 21 ILS approach and advised there was 25 to 26 track miles to go.

On departure, VH-NGA had planned to track to the southeast and would therefore be under the control of the approach east controller from shortly after take-off. This controller was responsible for separating VH-NGA from G-BNLW on the ILS approach for runway 21. It was his plan to have VH-NGA make an early right turn after take-off and have the aircraft established on an assigned radar heading (120 deg) prior to the three mile radar separation standard between the two aircraft being infringed.

As G-BNLW got closer, the approach east controller realised that for his plan to work, it was necessary for VH-NGA to get airborne without delay.  This was evidenced in transmissions to VH-NGA while still taxiing including requesting "minimum delay due inbound traffic", "will this be a rolling take-off" and "clear for immediate take-off". Although the approach east controller intended that VH-NGA make an early right turn, he never communicated this to anyone and therefore it was never communicated to the aircraft. VH-NGA made an intersection departure which further reduced the distance between VH-NGA and G-BNLW.

When VH-NGA became airborne, it was obvious from the radar return that the aircraft was not making an early right turn. The approach west controller was still controlling G-BNLW and he realised there could be a loss of separation between the two aircraft unless some preventive action was taken. At this time G-BNLW was intercepting the runway 21 localiser from a heading of 180 deg and was cleared to descend to 1500 ft.

The approach west controller instructed G-BNLW to continue its right turn onto 290 deg but the response from the pilot was "we're fully established now on the ILS". The controller repeated the instruction but again G-BNLW did not comply. The controller did not tell G-BNLW the reason for his instruction. By this time, it was evident that VH-NGA had commenced its right turn so the approach west controller did not persevere any longer in trying to get G-BNLW onto a westerly heading. He then instructed G-BNLW to contact the tower.

Shortly afterwards, VH-NGA contacted the approach east controller airborne and reported ".....right turn 120 climbing 3000 passing 2000". The controller responded asking the aircraft to make a hard right turn. It was estimated that separation between the two aircraft reduced to approximately 1.75 miles with less than 1000 ft vertical separation. The required standard is three miles when there is less than 1000 ft vertical separation.

Analysis

When the change of runways occurred, VH-NGA was taxiing in the vicinity of the terminal. It was not necessary for the aircraft to continue to be processed for a runway 03 departure. Had the aircraft been redirected to runway 21 then this incident would never have occurred.

There was no evidence from the recorded communications or interviews with the controllers involved of any consideration being given to reclearing VH-NGA for a runway 21 departure. It appears that the aircraft was simply permitted to continue taxiing for runway 03. This indicated a lack of consideration/coordination in respect of the departure for VH-NGA.

With VH-NGA processed for a runway 03 departure, the approach east controller accepted responsibility for separation between VH-NGA and G-BNLW. In doing so, he did not then apply any positive measures to ensure that separation was maintained. His plan was simply an expectation that VH-NGA would be on a diverging radar heading before radar separation was lost. This was a misjudgement which made no allowances for anything going wrong such as radio failure or an aircraft malfunction.

When the approach west controller became aware of the deteriorating separation situation between the two aircraft, he issued heading instructions to G-BNLW (i.e. continue the right turn onto 290 deg) which were meant to maintain separation. However, the intent of the instructions were obviously not understood by the crew of G-BNLW because the aircraft did not comply. This lack of compliance was most probably due to the lack of alerting phraseology (e.g. due opposite direction traffic continue the right turn onto...) to convey the urgency of the situation.

Factors

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

1. A runway change occurred at Perth airport from runways 03 and 06 to runways 21 and 24. As a result an inbound Boeing 747, G-BNLW, was vectored for an ILS approach to runway 21 while a taxiing Westwind, VH-NGA, was allowed to continue for an opposite direction departure from runway 03.
2. Although it would have been a simple matter to redirect VH-NGA for a departure from runway 21, for reasons not determined this was not done.
3. With the decision made to depart VH-NGA from runway 03, the approach east controller, who was responsible for separating that aircraft from G-BNLW, did not take adequate measures to ensure that separation.
4. The approach west controller, who was controlling G-BNLW, saw the loss of separation developing and issued heading instructions to G-BNLW to alleviate the situation. However, because he did not use appropriate phraseology to convey the urgency of the situation, his instructions were not followed by the aircraft.

The pilot-in-command of the SF-340A had been issued a take-off clearance from runway 23, with a restriction to maintain runway heading due to an inbound Cessna 402B approaching from the south, to join a right-hand circuit for runway 12.

After take-off, the aircraft was seen to turn right to intercept its planned outbound track, contrary to the take-off clearance instructions.

The turn was observed by both the Tower and Departure Controllers, but too late to effect the required separation. Both aircraft remained in sight of the Tower Controller during the confliction.

While the SF-340A was turning it passed within 1.85km horizontally, and less than 1,000ft vertically above the inbound traffic.

The co-pilot had recently been endorsed on the aircraft type, and the pilot-in-command advised that he had been distracted during taxi and take-off by slower than normal pre-take-off checks due to the inexperience of this crew member and forgot about the after-take-off restriction.

FACTUAL INFORMATION

Sequence of events

The B747 had departed Sydney and was maintaining flight level (FL) 310 on air route A576. The crew had reported passing the position reporting point MASDA at 1642 EST, with an estimate for reporting point KALUG of 1735.

The B737 had departed Alice Springs and was proceeding on A576 in the opposite direction. The crew had reported passing KALUG at 1616 with an estimate for MASDA of 1702. The aircraft was maintaining FL290.

When the B737 crew made their KALUG position report, the crew of the B747 was monitoring a different control frequency and was unaware of the presence of the B737. However, when the B747 crew made their MASDA position report, the crew of the B737 was on the same frequency, but did not hear the report and was unaware of the presence of the B747 on the reciprocal track.

At 1648, the crew of the B737 requested climb to FL330 and was instructed to "stand by" by Air Traffic Control sector 5 (Sec 5). Sec 5 then co-ordinated the request with the Sec 6 controller who concurred with the change of level for Sec 6 airspace. The Sec 5 controller proceeded to calculate a time of passing so that he could assess when a change of level for the B737 could be initiated.

The Sec 5 controller elected to use the ground speed method to calculate the time of passing. He asked each crew, in turn, to report the ground speed of their aircraft. It was at this time that both crews first became aware of the presence of the other aircraft, and that there was a possible confliction.

The controller made the initial calculation mentally and followed this with a written calculation on a note pad. As both calculations produced a result of 1724, he wrote this time of passing on the flight progress strip.

At 1652, Air Traffic Control approved the B737 to climb to FL330, with a requirement to reach FL330 by time 1714. The crew commenced the climb immediately.

The crew of the B747 received an information symbol indicating "Other Traffic" on their traffic alert and collision avoidance system (TCAS) when the B737 was approximately 40 NM ahead. The target information changed to a proximity alert followed by a traffic advisory warning. At this point, the crew decided that if the TCAS proceeded to a resolution advisory the pilot in command would manually fly the aircraft as directed by the advisory. The co-pilot would simultaneously make a radio broadcast indicating the action being taken.

At approximately 1653.20, the TCAS issued a resolution advisory advising "climb, climb". The pilot in command immediately commenced a climb in accordance with this advice. Aware that the other aircraft was also climbing, the pilot in command continued the climb at the best rate to approximately FL321, before returning the aircraft to the approved cruising level, FL310. The co-pilot made a radio broadcast indicating that they were carrying out a TCAS climb at the same time as the pilot in command initiated the climb manoeuvre.

Analysis of the B747's quick access recorder showed that the aircraft gained approximately 500 ft during the time that the resolution advisory was active and then gained a further 550 ft before descending.

The pilot in command of the B747 did not see the B737. However, immediately before the TCAS resolution advisory warning activated, the co-pilot sighted the B737 straight ahead and at a slightly lower level. He lost sight of it as his aircraft entered the climb. At that point, the B737 was passing directly below the B747.

The B737 was climbing at approximately 1,000 - 1,200 ft/min when the crew heard the TCAS climb radio call from the crew of the B747. They realised that their aircraft must have been the conflicting traffic and initiated an immediate descent. The crew looked up to see the underside of the B747 pass over and slightly to the right of their aircraft. The pilot in command levelled the B737 at FL305 and believed that his aircraft may have reached FL308 prior to the descent.

At 1653.30 (approximately ten seconds after the TCAS resolution advisory) the pilot in command of the B747 requested from air traffic control the position of the other aircraft. The Sec 5 controller responded by instructing the B737 to expedite climb to FL330. This instruction was carried out by the B737 crew.

The aircraft had passed with approximately 500 ft vertical separation and little or no horizontal separation. The required separation standard for two aircraft passing in the opposite direction on the same air route above FL 290 is a minimum of 2,000 ft vertical difference.

Air traffic control

During the investigation, several people involved in the training and performance appraisal of the sector 5 controller indicated their concern with certain aspects of his abilities. These concerns were notably in the area of procedural separation standards. Examination of check reports and interviews with supervisors, training staff, and workface controllers revealed that the official air traffic controller performance and check reports related only to some of the perceived deficiencies.

Many supervisors and training officers (including the former "Check" controllers) prepared formal reports for official filing but also retained other opinions in "private" files which were not readily available to management. However, the records that were available showed that his supervisors considered that the controller's performance was deficient in the area of procedural separation standards. When the controller was selected for training on another sector (Sector 6), the use of radar standards, which he had not used in previous ratings, was required. He subsequently failed this rating due to difficulty with procedural separation standards, rather than the radar standards. At the time of his final check for the Sector 6 rating the controller was experiencing personal problems which were not known to management.

There was no system in place to ensure that the responsible manager was specifically informed of a perceived weakness other than some review of the various reports prior to filing. This applied to ongoing performance appraisal as well as initial training and rating reports. As individual managers were often responsible for numerous controllers, the opportunity for them to adequately review all performance and check reports was not always available.

After the controller had failed to gain a rating on Sector 6, management elected to place him back on Sectors 1 and 5, those sectors on which he was rated prior to the Sector 6 training. The decision was made to retrain rather than refamiliarise the controller. Such a decision indicated a greater than normal concern for his ability to obtain the ratings which he had only recently held.

Having made this decision, management then elected not to provide the controller with simulator training to assist in this re-rating program, even though specific programs were available. The training officer was instructed to concentrate on separation standards during normal on-the-job training, especially time of passing and lateral separation problems.

Airborne collision-avoidance systems (ACAS)

ACAS are intended to improve air safety by acting as a last-resort method of preventing collisions or near collisions. The equipment utilises secondary surveillance radar technology to operate independently of ground-based aids and air traffic control. It monitors other aircraft in the vicinity and assesses the risk of collision by interrogating airborne transponders. Non-transponding aircraft cannot be detected. TCAS II, as fitted to the B747, is a version of ACAS which provides resolution advisories (for vertical manoeuvres only) as well as traffic information. The B737 was not equipped with any form of ACAS.

Time of passing calculation

Several attempts were made to replicate the calculation made by the controller for the time of passing. Neither the investigation team, air traffic services specialists, nor the controller concerned could determine how the time of 1724 was derived.

There was a 3 minute 20 second delay between the time the crew of the B737 requested the climb and the time the controller approved the climb. In this period the controller made two independent time of passing calculations, one mentally and one on a notepad. Whilst performing these tasks he was required to respond to nine separate air-ground and co-ordination communications involving 25 interchanges.

Two-way air routes

On 14 September 1995, a new upper air route network was introduced in the Australian Flight Information Region. The change resulted in numerous one-way air routes replacing many two-way routes and was intended, amongst other things, to reduce the potential for incidents such as this. The route being flown by the aircraft involved in this occurrence was not changed to a one-way route.

Flight crew response

TCAS resolution advisory

The pilot in command of the B747 was aware of the separation standard of 2,000 ft which was applicable at FL310. He chose to continue climb beyond that normally expected for a TCAS event, because of the known position and climbing attitude of the other aircraft. He wanted to ensure that his aircraft would outclimb the B737 and took into consideration the fact that any third aircraft in the area should be at least 2,000 ft above his aircraft.

ANALYSIS

This occurrence involved the provision of an airways clearance to permit an aircraft to climb through the level of another aircraft travelling in the opposite direction on the same air route.

Calculation of the closest point of approach indicates that, had the crews of both aircraft not responded following the TCAS activation, the aircraft would have passed with no horizontal separation and approximately 200 ft vertical separation.

Flight crew response - TCAS resolution advisory

The resolution advisory is an indication given by TCAS II to a flight crew that a vertical manoeuvre should or should not be performed to maintain safe separation from other aircraft. If required, such a manoeuvre should average between 300 ft and 400 ft. To prevent possible secondary separation problems, the manoeuvre should not normally exceed 700 ft. In this occurrence, the pilot in command decided to climb his aircraft in excess of this expected range because of the relative positions of the two aircraft.

The crew of the B737 did not see the B747 until after the TCAS broadcast by the crew of the B747, and the only visual acquisition was by the co-pilot of the B747 some seconds before the TCAS resolution advisory after he had received traffic advice from the TCAS. It is therefore reasonable to say that, had the B747 not been equipped with TCAS, neither crew would have seen the other aircraft in time to commence effective avoiding action.

Time of passing calculation

The separation standard based on the time of passing calculation was correct. It was the calculation itself that was in error.

The controller was confident that his calculations were correct because the time written on his notepad agreed with his mental assessment. However, he had been distracted by the amount of air-ground and co-ordination exchanges required in such a short period of time. This level of distraction would have resulted in a high mental workload and may have contributed to the erroneous time of passing calculation.

Management strategies

Having decided to re-rate the controller on Sectors 1 and 5 instead of refamiliarising his previous ratings, management indicated a greater than normal degree of concern for his performance. Information stored in "private files" would have given management a better insight into these perceived procedural deficiencies.

A re-rating program was devised to assist the controller, but this plan did not require simulator training for procedural separation standards. Such training may have provided a more comprehensive preparation for his rerating. Although adequate simulator exercises were available, finding appropriate training staff for the simulator proved difficult.

SIGNIFICANT FACTORS

1. The aircraft were travelling on a two-way air route in opposite directions.
2. Management strategies for the early detection and remedy of perceived deficiencies in controllers' competencies were ineffective.
3. The controller miscalculated the time of passing of the aircraft.
4. The TCAS on the B747 provided the only catalyst for avoiding action by the aircrew.

SAFETY ACTION

During the course of the investigation the Bureau of Air Safety Investigation issued the following interim recommendation to the then Civil Aviation Authority on 6 June 1995:

"IR 950117

"The Bureau of Air Safety Investigation recommends that the Civil Aviation Authority:

"(i)   mandate the fitment and use of an Airborne Collision Avoidance System (ACAS) in all aircraft engaged in Regular Public Transport (RPT) operations;

"(ii) consider the requirement for the fitment and use of a suitable ACAS in other aircraft engaged in the carriage of passengers for hire or reward;

"(iii) mandate the activation of any SSR transponder in aircraft so equipped at all times when airborne, irrespective of the class of airspace in which the aircraft may be flying;

"(iv) mandate the standard of ACAS equipment to be carried in each aircraft classification;

"(v)   set a timetable for the introduction of ACAS equipment; and

"(vi) ensure that air traffic services officers are given adequate and timely education and continuation training in the capabilities and operational impact of ACAS equipment."

The following response was received from the Civil Aviation Safety Authority on 21 August 1995:

"I refer to Air Safety Interim Recommendation IR950117 regarding Airborne Collision Avoidance Systems (ACAS).

"In response to the reported air miss occurrences referred to as justification, in part, for this recommendation, the Civil Aviation Safety Authority has commenced an examination and analysis of both BASI and AA incident data in order to identify and better understand the nature of these deficiencies within the Australian aviation system. Significant differences between BASI and AA data, in both number and classification of reported incidents, has complicated and restricted this process to date.

"Apart from analysis of risks in a non ACAS environment, CASA also recognises the need to assess the impact of ACAS on future major projects such as TAAATS and implementation of AACS. The Authority also intends to review any potential disbenefits that may arise from ACAS, including spurious activations and effects of resolution advisories (RAs) in complex air traffic situations.

"Industry and Public consultation will precede any adoption of requirements for aircraft to carry ACAS. The Authority continues to monitor world trends and is currently reviewing recent UK safety studies on ACAS." The Bureau has classified the response as OPEN and has initiated further correspondence on the matter.

Additionally, Interim Recommendation IR950218 was issued to the Civil Aviation Safety Authority and Interim Recommendation IR950219 was issued to Airservices Australia, after the 1995 reorganisation. These were issued on 11 January 1996:

"IR950218

"That the Civil Aviation Safety Authority in conjunction with other operators of TCAS-equipped aircraft, evaluate the initial and recurrent training requirements for aircrew of TCAS-equipped aircraft.

"The evaluation should ensure that the risks associated with excessive vertical deviations in response to TCAS RAs are identified."

The Civil Aviation Safety Authority responded on the 12 February 1996 as follows:

"I refer to your interim recommendation number IR950218 regarding the incident involving Boeing 747, G-BNLR and Boeing 737-476, VH-TJK, on 4 May 1995.

"Summary

"The authority accepts the Interim Recommendation.

"Background to response

"Details of the exact nature and extent of training currently undertaken by TCAS equipped operators will be determined to establish whether or not such training adequately addresses the risks associated with excessive and inappropriate pilot responses to RAs."

The Bureau has classified this response as CLOSED/ACCEPTED.

"IR 950219

"That Airservices Australia evaluate the TCAS initial and recurrent training requirements for ATS personnel. This evaluation should ensure:

"(i) thorough and consistent practical knowledge is achieved throughout ATS; and

"(ii) TCAS RA encounters are practised in simulator sessions."

Airservices Australia responded on the 22 February 1996 as follows:

"I am writing in response to your Interim Recommendation IR950219 which recommends that Airservices Australia:

"evaluate the TCAS initial and recurrent training requirements for ATS personnel.

"Preliminary evaluation of the level and extent of TCAS training provided to Air Traffic Services personnel indicates that training on TCAS was generally consistent across Air Traffic Services centres.

"The training provided consisted of local team development supported by a video presentation with accompanying booklet detailing TCAS procedures.

"The ATS Operational Training Unit is currently developing a comprehensive package for the purpose of evaluating TCAS training and awareness. The purpose of this evaluation is to assess both the level of initial training and the current appreciation of TCAS.

"Future TCAS training will be dependent upon the results of this evaluation. The focus of any proposal will be on improving and enhancing training delivery to ensure that both initial and recurrent training needs are clearly identified and addressed.

"A variety of instructional methodologies and strategies will be considered during the evaluation process. Any proposal for simulation of TCAS RA encounters will also be dependent on the results of the evaluation and the feasibility of successfully simulating such encounters."

The Bureau classified this response as CLOSED/ACCEPTED.

Note: In December 1996 CASA issued a Legislative Instrument Proposal which proposed the mandatory carriage of TCAS equipment in some classes of aircraft and addressed further the recommendations IR950218 and IR950219.

Final recommendations

The Bureau of Air Safety Investigation made the following recommendations on 1 November 1996:

"R950227

"That Airservices Australia further review the upper air route structure to create, wherever practicable, one-way air routes within the Australian Flight Information Region."

"R950228

"That Airservices Australia review the training given to ATC and Flight Service staff in coping with distractions in the workplace and consider improvements in the techniques to reduce the impact of distractions on staff performance."

"R950229

"That Airservices Australia:

"(i) develop strategies that encourage all officers who carry out performance appraisals to notate all concerns in an adequate and timely manner on the approved proforma;

"(ii) develop strategies for improved management oversight of appraisal reports, especially in the area of "below expected performance" assessments; and

"(iii) ensure that specific training programs, including simulator and desk-top exercises, are developed to address perceived deficiencies in each of the skill and knowledge items contained in any appraisal report."

Factual information

The F-27 aircraft had been carrying out circuit training and had landed on runway 05. The crew accepted the Aerodrome Controller's (ADC) offer to backtrack on runway 05 for further circuits and they reported ready for takeoff at 1002.32 CST. The TB-10 aircraft was making an approach to runway 12 for a touch-and-go landing. At 1003.05, the ADC cleared the aircraft for the touch-and-go and also passed departure instructions to that aircraft. Co-ordination between the ADC and Approach controller resulted in an agreement to sequence the F-27 so that it would land behind four jet aircraft arriving for runway 05. The ADC informed the F-27 crew of the sequencing difficulty and, at 1003.15, said there would be about a 30-second delay before he could issue a take-off clearance. At 1003.45 the pilot of the B737 contacted the ADC whilst on approach for runway 05. At 1004.58, the ADC cleared the F-27 for takeoff.

The crew elected to hold position as they had seen the TB-10 on its approach to land but had not seen it depart after the touch-and-go. From the position of the F-27 part of runway 12 was hidden from view behind buildings. Consequently, the F-27 was held by the crew with power on and ready to commence the takeoff once the TB-10 was clear of the intersection. The B737 crew saw the F-27 lined up on runway 05 and slowed their aircraft lowering, 40 degrees of flap to assist in the sequencing process. At 1005.04 the ADC informed the B737 crew to expect a late landing clearance. At 1005.12 the F-27 commenced its take-off run approximately two minutes after receiving the advice of the expected delay of 30 seconds. The TB-10 was airborne but still not clear of the runway intersection and at this time, the B737 was on short final. At 1005.30, as the F-27 was becoming airborne the ADC cleared the B737 to land. At the same time the B737 crew informed the ADC that they were commencing a go-around which they had started just prior to receiving landing clearance. The go-around decision was based on doubts that sufficient separation existed between the B737 and F-27, especially if the F-27 was to reject its takeoff. The B737's go-around was flatter than the normal climb profile so that the captain could maintain visual contact with the F-27 and to ensure a safe manoeuvring speed. At 1005.50, the ADC instructed the F-27 crew to maintain 500 ft and to turn left onto a heading of 300 degrees. As the F-27 was commencing its left turn, the B737 was turned right about 20 degrees to ensure continued sighting of the F-27.

Analysis

Aerodrome Controller

The ADC elected to process the F-27 by backtracking it on runway 05. The sequence chosen by the ADC was time-critical as it depended on the TB-10 crossing the intersection before the F-27 could take off, and the B737 was dependent on both these aircraft departing before it could land.

The controller realised that the sequence would be tight but he did not consider an alternative plan even though several options were available. Although after line up the F-27 crew were told to expect about a 30-second delay the actual delay was one minute and 43 seconds. When the crew of the F-27 were issued with their take-off clearance the TB-10 was still not clear of the runway intersection which led to a considerable erosion of the time base on which the ADC had based his original plan. Radar analysis indicates that the take-off clearance was issued to the F-27 approximately 20 seconds before the TB-10 had cleared the runway intersection. Paragraph 23 of the Manual of Air Traffic Services (MATS) 6-3-4 specifies the separation standard applicable to a situation in which two aircraft are departing from different runways stating that aircraft B [F-27] shall not be permitted to commence takeoff until aircraft A [TB-10] has crossed the intersection. Paragraph 26 of MATS 6-2-3 allows the controller to issue a take-off clearance prior to the prescribed separation existing if in the opinion of the controller no collision risk exists, and there is reasonable assurance that separation will exist when the aircraft commences its take-off roll. In this case, as the F-27 crew were expecting an immediate departure on receipt of take-off clearance, the ADC could not have reasonably expected the aircraft to delay its take-off roll. The investigation found that no uniformly accepted interpretation of the meaning of paragraph 26 existed between ATS management and some operational staff.

The B737 crew were told to expect a late landing clearance and the ADC issued this clearance between 1005.31 and 1005.33. Radar analysis indicates that the B737 was either just short of, or passing the runway threshold when the landing clearance was given. The F-27 became airborne at 1005.35. Paragraph 32 of MATS 6-3-5 states that aircraft B [B737] shall not be permitted to cross the runway threshold until aircraft A [F-27] (less than 136,000 kg MTOW) is airborne after takeoff. If the B737 was short of the runway threshold at 1005.35, then the clearance complied with the MATS instruction. If however, the B737 was passed the runway threshold at that time, the runway separation standard would have been breached. The ADC believed that had the B737 landed, the separation standard would have been achieved and that the F-27 would be airborne prior to the B737 crossing the threshold. Radar analysis could not determine if the separation standard had been achieved. The standard requires that an aircraft below 136,000 kg MTOW should be airborne before a landing aircraft of any size crosses the runway threshold. Therefore, as an example, an aircraft could be airborne at 70 kts and 500 ft along the runway when a landing jet aircraft crosses the threshold at up to 150 kts. The separation standard would then be breached if an unexpected event forced the departing aircraft to discontinue takeoff and reoccupy the runway. An ADC is required to take into account any unexpected manoeuvres by aircraft when issuing a clearance. In this occurrence, there was a 2-4 second period in which the standard may have been achieved. The B737 crew doubted that the standard would be achieved. Once the B737 had commenced its go-around, the ADC expected the climb to be similar to a normal B737 departure and believed this would solve any separation problems that may arise.

However, the climb profile was more shallow than expected, and the ADC considered further action was required to guarantee separation. Instructions were issued to both crews to ensure that vertical and lateral separation were maintained. Crew of the F-27 After lining up and holding on runway 05, the F-27 crew were given traffic information on the TB-10 which was on final approach for a touch-and-go on runway 12. They sighted the aircraft and monitored its descent and touchdown. The ADC had asked the F-27 crew to be ready for an immediate departure as soon as the TB-10 was airborne. When the take-off clearance was issued, they increased power but held position until they had sighted the TB-10 and this delayed the commencement of their takeoff for approximately 12 seconds. Crew of the B737 When on final approach for runway 05, the B737 crew became aware of the F-27 occupying the runway and aircraft speed was reduced as much as possible. On short final, the ADC issued a late landing clearance expectation and the crew briefed for a possible go-around. The B737 captain was concerned that if the F-27 rejected its takeoff or had a malfunction shortly after lift-off and landed again, there would be insufficient room for the B737 to land safely. As the B737 approached the threshold, the crew could see that the F-27 was still not airborne and well within the distance they would need to land and they initiated a go-around about the same time as they received a landing clearance. To regain safe climb speed and maintain visual contact with the F-27, the captain elected to climb at a sufficiently nose-down attitude until his aircraft was safely clear. The captain considered that the F-27 was sufficiently close to his aircraft that a risk of collision existed if he lost sight of that aircraft.

Findings

1. The ADC misjudged the time required for the TB-10 to clear the runway intersection.

2. The ADC issued a take-off clearance to the F-27 when the required runway separation standards were not met. (MATS provides the ADC with this discretion.)

3. The discretion shown by the ADC in finding 2 above, was inappropriate.

4. The F-27 crew could not see the TB-10 during the ground roll portion of its touch-and-go landing on runway 12.

5. The F-27 crew had good reason not to commence takeoff at the time the ADC issued the clearance.

6. The ADC cleared the B737 to land in the belief that the appropriate runway separation standard would be achieved.

7. The B737 crew executed a go-around manoeuvre when they judged that the runway may not be available for the safe operation of their aircraft.

8. At the time of the B737 go-around, it was unclear (to the investigation team) if the runway separation standard would have been maintained.

9. The B737 climbed at a more shallow angle than that expected by the ADC.

SIGNIFICANT FACTORS

1. The ADC placed undue pressure on himself by initiating a backtrack by the F-27.

2. The decision by the ADC to delay the F-27 until the TB-10 had completed its touch-and-go placed the three aircraft in a position whereby they each required the use of the runway intersection within a very short period of time.

3. The runway separation standard applicable in this case is such that it allowed the ADC to issue clearances whereby the B737 crew were unable to guarantee the safe landing of their aircraft.

4. The ADC did not have a fall-back plan for the sequence he initiated.

SAFETY ACTION

As a result of the investigation the following local action was taken. Air Traffic Services management clarified the interpretation of MATS 6-2-3 para. 26 to the satisfaction of the investigation team and ensured that all Adelaide Tower staff were in no doubt as to their responsibilities in this regard. Additionally, the Bureau of Air Safety Investigation issued the following interim recommendation to Airservices Australia on 29 May 1996.

IR960046 The Bureau of Air Safety Investigation recommends that Airservices Australia, in conjunction with the Civil Aviation Safety Authority, reviews the runway separation standards for landing aircraft in relation to departing aircraft. This review should consider what separation standards may be required to cover contingencies where slower departing aircraft may impede the safe passage of faster landing aircraft. Airservices Australia responded on 29 July 1996. IR 960046 Airservices, in conjunction with CASA, is currently reviewing runway separation standards for landing aircraft in relation to departing aircraft. This review includes an examination of the ICAO standards in relation to departing aircraft, as well as current international practices. The review will also attempt to assess the impact and safety benefit of the introduction of more restrictive practices on runway capacity at specific locations.

It is intended that the review be completed by the end of August.

Response Status: Closed - Accepted.

The Civil Aviation Safety Authority responded on 17 June 1996. In accordance with IR 960046 the separation standards in relation to a faster landing aircraft behind a slower departing aircraft will be reviewed. (Note: the Report states that there is a 2 - 4 second period where the standard may or may not have been achieved: were the tolerances of the radar taken into account in making that assessment?)

Response Status: Closed - Accepted BASI Note: The radar tolerances were taken into account in the investigation.

VH-NJG departed unrestricted on the Runway 19 Kilcoy Standard Instrument Departure (SID). The next departure off Runway 19 was VH-XFZ departing on a heading of 220 degrees. The departure restriction was to accommodate an aircraft arriving from the west for runway 14.

The Departures Controller intended to turn the trailing aircraft (VH-XFZ) further right shortly after contact was made. However, there was a Secondary Surveillance Radar (SSR) correlation problem with VH-XFZ in that its transponder did not operate until the crew recycled the selector. When VH-XFZ showed up on the SSR, it was less than 2 NM behind VH-NJG. However, the crew of VH-XFZ reported that they had VH-NJG in sight since take-off.

The Departures Controller had assumed that the ADC would provide separation of two minutes or 3 NM for departing aircraft if their tracks diverged by less than 30 degrees.

As the crew of the trailing aircraft had visual contact with the preceding aircraft throughout their take-off and initial climb, there was no risk of collision.

Analysis

Under the circumstances, the Departures Controller's operational technique did not provide separation assurance. His assumption that the ADC would apply separation was the more significant factor in the breakdown of separation.

Safety Result

Following the incident Local Operating Instructions were amended to define more clearly the responsibilities of both the ADC and Departure Controller in relation to initial departures.

VH-HSP departed Sydney five minutes before VH-TJP. Both aircraft were tracking via Wollongong and T18 to Eildon Weir. On first contact with sector control, VH-HSP was cleared to climb as planned to flight level (FL) 350. VH-TJP was also planned at this level but on first contact with sector control, VH-TJP was cleared to climb to amended FL 330. At this stage, VH-TJP was closing on VH-HSP at 110 knots and was 13 miles behind.

A very inexperienced trainee was operating the position under the supervision of a rated controller. The controller was distracted from his supervision task for a short period and when he returned his attention to the screen, he noted that VH-TJP was now only five miles behind VH-HSP and 200 feet below.

The controller told the trainee to turn VH-TJP onto 150 degrees. The trainee instructed VH-TJP to turn right onto 150 degrees but immediately corrected this to turn left onto 150 degrees. The crew then queried the direction of turn. The controller took over from the trainee, told the crew to turn left and gave them traffic information on the other aircraft. Separation reduced to a minimum of three miles at the same level.

Both the controller and the trainee indicated they were not very familiar with the performance of VH-HSP and that the climb performance achieved had been less than they anticipated. The controller expected that VH-HSP would reach flight level 350 before longitudinal separation was lost. Separation would then be maintained vertically which was the reason VH-TJP was assigned amended FL 330.

Significant Factors

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

1. Inadequate knowledge of the performance of VH-HSP.

2. The controller did not maintain an adequate monitor of the situation to ensure that the separation strategy devised, actually achieved its aim.

The aircraft was cleared for a visual approach to the duty runway (RWY) 11 and joined a 7km final from a right base position. The crew were advised by ATS of several military fighter aircraft taking off from RWY 29.

The captain advised ATS of his position and that a traffic confliction was about to occur as he saw two of the six fighter aircraft pass above and to the left of his aircraft. The last of the section of fighters passed within 1,000 metres of the aircraft at the same level.

Subsequent investigation revealed that a request for the use of a non-duty runway (RWY29) had been agreed to previously for the military aircraft, but there had been a delay with their departure. The radar controller failed to note that a non-duty runway was marked on the flight strips. Reflected radar returns were being experienced on the Surface Movement Radar and the controller misidentified these reflected returns as those of the military aircraft, and being relatively inexperienced cleared the military aircraft to take off in confliction with the approaching aircraft.

OE-LAV, a Boeing B767 aircraft (flight number LDA2) enroute from Melbourne to Singapore at Flight Level 350 (FL350) suffered a reduction of oil pressure in one engine. The problem could not be rectified in flight and the engine was shut down.

The crew notified Perth ATC of the problem, advising that they were commencing an emergency descent to FL240 and required a diversion to Darwin. They were advised of opposite direction traffic, HS-TMB, a DC10 aircraft (flight number THA991), at FL330 which should be passing their present position at that time.

They were initially cleared to FL340, but the crew advised that their TCAS equipment had shown they had passed THA991 and were now diverting right of track and requiring an immediate descent to FL240.

THA 991 was instructed to descend to FL290, but as ATC could not provide a positive separation standard between the two aircraft instructed LDA2 to maintain its own separation while descending to FL240 and track direct to Darwin.

VH-WNI departed Wagga for Canberra with a clearance to maintain 4000 feet visual due to VH-PVX inbound to Wagga on a reciprocal track at 6500 feet. The pilot of VH-PVX called at three miles west of Tumut (approximately the CTA boundary) at 6500 feet. She was later issued with a clearance to descend to 4500 feet. The clearance was read back correctly, and mutual traffic was passed to both aircraft.

Later, VH-WNI reported crossing the Hume Highway. VH-PVX was then asked to report distance to run to the Hume Highway and altitude. The pilot responded, "two miles to run and leaving 4300 feet on climb". The two aircraft would have been passing at that time with less than 500 feet vertical separation.

Later, when VH-PVX was approaching at Wagga, the pilot was instructed to descend to 2500 feet and overfly for a right circuit runway 23. This was to maintain separation with other circuit traffic. VH-PVX was observed approaching the aerodrome from the east and in response to a request from air traffic control, the pilot reported her altitude as 2000 feet. She was instructed to climb back to 2500 feet.

The pilot of VH-OGN advised that whilst on descent approaching their cleared altitude of 5000 feet via a Hopla 1 STAR for runway 16, they received a TCAS proximity traffic warning with no altitude readout. The intruding traffic was sighted and was reported to be similar to either a Cessna 182/206, tracking from left to right and about one and one half nautical miles to their left. VH-OGN was levelled at 5400 feet and banked to the right.

The other aircraft maintained its track and passed approximately half a nautical mile behind VH-OGN. The captain of VH-OGN said that while this was occurring, the TCAS changed to a "yellow traffic with aural warning". He estimated that the Cessna was between 4800 and 5000 feet.

Although the Cessna was squawking code 2000, there was no altitude readout on air traffic control screens. Attempts to contact and identify the Cessna were not successful.