FACTUAL INFORMATION

A Fairchild Metroliner was one of a number of aircraft conducting training in the Perth terminal area. Three air traffic control (ATC) positions were operating in the tower: surface movement, coordination, and aerodrome control. The aerodrome controller was also conducting "on the job training" (OJT) of a trainee controller. The aerodrome controller conducting the training was suitably rated and had completed an OJT instructor course. Runways 21 and 24 were in use for arrivals and departures. Practice instrument approaches were also being conducted by a number of aircraft. The wind was gusty and there were intermittent showers in the area, with a cloud base of approximately 2,500 ft and a visibility in excess of 10 km. Previously within Perth ATC centre, the approach coordination controller had passed aircraft estimates with the assigned runway to the tower coordinator. In an endeavour to reduce voice coordination between tower and approach elements, a procedure utilising the radar display label was in use. This procedure entailed the approach controller including the assigned runway for arriving aircraft on the "Ops Data" line of the radar display label. Thus, any controller with access to a radar display could read the runway assigned to arriving aircraft.

As part of the procedure, the approach coordination controller ceased to provide updated estimated time of arrivals (ETAs) to the tower coordinator. Aerodrome controllers calculated an ETA from the observed position of aircraft on the radar display and updated tower flight progress strips accordingly. This meant the aerodrome controller had to observe the radar display label for each arriving aircraft to obtain the assigned runway and to calculate an ETA. As a consequence, aerodrome controllers generally changed range scales between the 17 NM and 50 NM scales on a regular basis. The procedure was noted as being easily undertaken in low density air traffic situations. However, when traffic conditions were busy, an aerodrome controller's ability to change range scales, calculate an ETA and note the assigned runway was conditional; being subject to the complexity of aircraft operations at the time and the level of experience of individual controllers.

The trainee controller had very little expertise in the procedure and was developing his skills while undertaking OJT in aerodrome control. The aerodrome control position at the console was designed for single-person operation and there was insufficient space for the trainee and the aerodrome controller to occupy the position at the same time. Thus, while the trainee was undertaking OJT he was required to operate and monitor all aerodrome controller facilities on the console. The aerodrome controller's ability to read the tower radar display labels was limited by the physical location of the display and other facilities on the console. This situation was compounded by the use of a glare shield around the radar display, which was used to reduce the effect of extraneous light on the display. Without the shield, sunlight washed the display presentation and made it difficult for controllers to see aircraft symbols and labels. However, when the shield was in use, a controller could not observe the display without standing adjacent to the console and looking directly over the shield onto the radar display. Consequently, the aerodrome controller had to regularly re-position himself to be able to read the radar display while the trainee controller was operating in the aerodrome control position. The aerodrome controller was unable to readily review information on the flight progress strips or on the radar display. As the traffic sequence developed and became more complex, the aerodrome controller assumed control responsibility from the trainee. The trainee stepped back from the console to enable the aerodrome controller to operate facilities, view the flight strips and monitor the radar display.

The correct reading of the label or re-checking of the label information was critical to the aerodrome controller's task as the information was not provided from any other source. However, the trainee controller had not been advised of the potential problems and was endeavouring to learn the procedure while training in the aerodrome control position. After taking over from the trainee the aerodrome controller did not confirm the runway assignment for any aircraft. He only used the information available to him from the flight progress strips on the console. Runway 24 was annotated on the flight strip for a Boeing 737 inbound from the east of Perth. As he subsequently became engrossed in sequencing and separating aircraft, he did not notice that the assigned runway for the Boeing 737, on the radar display label, was actually runway 21.

The crew of the Metroliner had been cleared to operate in the circuit not above 1,500 ft. After completing a number of circuits, the crew were instructed to make a left circuit for runway 21 and to report sighting a Baron aircraft joining an abbreviated instrument landing system final approach at 5 NM. The crew of the Metroliner did not sight the Baron and were instructed to continue downwind, which took them closer to the boundary between Perth and RAAF Base Pearce control zones (CTR). At about the same time, the supervising controller coordinated unrestricted descent for the Boeing 737 arriving from the east for landing on, what he thought, was runway 24. Due to the proximity of the Metroliner to the boundary between Perth and Pearce CTRs to the north of Perth aerodrome, the approach controller queried the aerodrome controller's intentions for the Metroliner. The aerodrome controller intended to instruct the crew of the Metroliner to make a left orbit on base to sequence the aircraft with the Boeing 737, which he believed was to land on runway 24.

The aerodrome controller then observed the radar display symbol for the Boeing 737 approaching left base for runway 21 and subsequently requested confirmation of the assigned runway from the approach controller. The approach controller confirmed the Boeing 737 was assigned runway 21. The crew of the Metroliner were aware the aircraft was well north of Perth and close to the airspace boundary, but they expected to turn and return to Perth very shortly and were thus unconcerned. The crew of the Boeing 737 reported sighting the Metroliner and were instructed by the approach controller to maintain 2,500 ft. The aerodrome controller instructed the crew of the Metroliner to orbit right to enable the Boeing 737 to continue descent for landing. Once radar separation was established, the crew of the Boeing 737 were instructed to make a visual approach. As the crew of the Metroliner orbited to the right, the aircraft entered the Pearce CTR and conflicted with a Macchi aircraft.

The pilot of the Macchi was conducting a tactical air navigation (TACAN) approach to runway 36 at Pearce and the last reported level was 2,000 ft. The Macchi and the Metroliner were observed by Pearce air traffic control radar to pass with approximately 1.5 NM horizontal separation. Neither crew observed the other aircraft. The crew of the Metroliner and the Macchi continued their flights and landed at Perth and Pearce respectively. There was a breakdown of separation. ANALYSIS Procedures The procedure for notifying assigned runway and calculating estimates for arriving aircraft for the tower had proven to be an effective alternative to the previous voice coordination. However, controllers needed to ensure that the label information was correctly transposed to the flight progress strips and that the radar display scale was changed regularly to check for arriving aircraft. This entailed a degree of skill which had to be developed with experience. The difficulties in using the radar display and the need to occasionally re-check aircraft label information was appreciated by other rated and more experienced controllers. However, this aspect had not been brought to the attention of the trainee controller prior to undergoing OJT. Consequently, he was unprepared for the switching of range scales on the display and the need to confirm the "Ops data" information, on an opportunity basis, to ensure the integrity of the operational information transposed to the flight progress strips. It is possible that the trainee annotated the incorrect runway on the flight progress strip for the Boeing 737 prior to the aerodrome controller taking over the position.

Additionally, because of the difficulties in observing the radar display while the trainee was operating at the position, it is likely that the aerodrome controller was unable to readily confirm the runway assignment of individual aircraft as the trainee annotated the respective flight progress strips. The procedure relied solely on controller skill; should the controller not perform the procedure correctly there was no back-up to ensure the tower received the necessary information. Training The trainee had not received any preparatory brief or training prior to the conduct of aerodrome control OJT. The trainee was rated at surface movement control but he had been absent from Perth aerodrome for a number of months working as a relief controller at Jandakot aerodrome. The lack of preparatory training was in contrast to terminal area trainees who receive 1 week of classroom training and 2 weeks of simulator training prior to undertaking OJT. The lack of preparatory training meant the trainee had to learn most aspects of the task while actually operating in the control position. This made the training difficult because of the different learning skills involved.

The different learning aspects coupled with his unfamiliarity with operations meant that the trainee was unlikely to be able to perform at the optimum level for OJT. Better preparation of the trainee prior to commencing the OJT would have been a defence against the incident occurring. Airspace Due to the proximity of RAAF Pearce to Perth, similar runway alignment and the location of the CTR boundary midway between the two aerodromes, operations in both CTRs can be constrained at times. This makes management of the airspace difficult and can limit the options available to controllers. It also requires a high level of situation awareness by controllers to ensure aircraft are not instructed or cleared in a manner that would have them infringe the Pearce CTR. In this incident the aerodrome controller was aware of the proximity of the Metroliner to the CTR boundary but was limited in his options to ensure descent was available for the Boeing 737. To achieve this he needed the crew of the Metroliner to orbit right. The subsequent instruction caused the Metroliner to enter the Pearce CTR. The crew of the Metroliner were aware that the aircraft was close to the CTR boundary but they believed the controller would have obtained clearance from Pearce ATC for the aircraft to enter the CTR. Consequently, the crew complied with the controller's instruction. The aircraft entered the adjacent CTR without a clearance and conflicted with the Macchi aircraft.

SIGNIFICANT FACTORS

1. The trainee controller was inadequately prepared to undertake OJT.

2. The aerodrome controller had difficulties in observing the radar display while the trainee was operating the control position.

3. The aerodrome controller, after taking over control from the trainee, had insufficient time to confirm the runway assignment of arriving aircraft.

4. The proximity of Perth and Pearce aerodromes and the common CTR boundary limited the options available to the aerodrome controller.

SAFETY ACTION

1. Perth ATS management implemented a traffic management plan in December 1996 which should assist controller traffic management.

2. The Bureau of Air Safety Investigation is evaluating aspects of "on the job training" relating to Perth Tower.

Any forthcoming recommendations will be published in the Quarterly Safety Deficiency Report.

FACTUAL INFORMATION

A Piper Chieftain aircraft, inbound to Launceston from Moorabbin, had been cleared to track via the Launceston (LT) very high frequency omni-directional radio range (VOR) navigation aid to the Nile non-directional beacon

(NDB). The cloud base was approximately 300 feet with fog and visibility of 4000 metres at the aerodrome. The Chieftain was to conduct an instrument landing system (ILS) approach to runway 32 left (32L) due to the weather conditions.

The tower was manned by a single air traffic controller. Prior to the incident the controller had frequently discussed weather conditions with the Melbourne Centre Coordinator, Sector controller and local meteorological personnel. At the same time, the controller had processed a number of aircraft (inbound and outbound) and was monitoring the operations of two vehicles on the aerodrome. A safety officer in one vehicle was conducting runway 32L lighting checks, which required the controller to adjust runway light switches, while the other officer was inspecting grass runway 18/36.

Due to the inbound track of the aircraft and orientation of runway 32L the pilot of the Chieftain had to make a reversal procedure at the Nile NDB prior to commencing the ILS approach. When the pilot reported over the Nile NDB, after tracking from the VOR and before conducting the reversal procedure, the controller instructed the pilot to report 'Nile outbound'. This was queried by the pilot and confirmed by the controller as 'Nile outbound'. The pilot tracked outbound from the NDB, conducted the reversal procedure and reported 'Nile inbound' which was contrary to the report required by the controller. The controller cleared the pilot to make a 32ILS approach, provided airfield information and instructed the pilot to 'Report on final leaving 3,000'.

As the controller completed flight strips, he became distracted and did not observe the Chieftain as it broke out of cloud and approached the runway. The safety officer on runway 32L observed the aircraft on final in the vehicle's rear vision mirror and immediately vacated the runway. The driver did not notify the controller the vehicle had vacated the runway strip.

The pilot had become visual just before the minima and after visually ascertaining the runway was unobstructed continued with the approach and landed without a landing clearance. There was a breakdown in procedures but no breakdown in separation.

ANALYSIS

The controller assumed the pilot report of 'Nile inbound' was the previously requested 'Nile outbound' report, and that the reversal procedure was still to be completed. Consequently, the controller thought there were a few minutes available to clear the printer and compile flight strips before needing to look for the Chieftain on ILS final and instructing the driver to vacate the runway. However, this was not the case, and the aircraft was closer to the aerodrome than the controller appreciated. The controller would appear to have heard what he expected to hear and not what was actually reported by the pilot. The phenomena of 'hearback' is recognised as a problem in aviation communication. However, it is difficult to guard against, especially when an individual is working alone. Had another controller been present the misunderstanding of the pilot's report may have been detected. Alternatively, if the controller had been able to concentrate on controlling and not been required to prepare flight strips, he may have been able to detect the changed report.

The pilot had a clearance for final which authorised the aircraft to leave the lowest holding altitude and to make a circling or missed approach as appropriate. A clearance for final is subject to any further air traffic control instructions and a clearance to land. In this situation the pilot had a clearance limit of the runway and should not have continued with the landing without a specific clearance of 'Clear to land'. But the pilot had the impression that he had a clearance to land and that the controller was well aware of the aircraft's location in the approach. Consequently, the pilot considered further reports unnecessary. This aspect is a failed defence as procedural air traffic control relies solely on pilot reports to arrange aircraft separation in instrument meteorological conditions. The lack of reports would have diminished the controller's situational awareness. Provision of the report of 'On final leaving 3,000' may have provided a cue to the controller that the aircraft was closer to the aerodrome than expected. As a consequence, the controller would probably have given greater priority to the approaching aircraft by ensuring runway availability (by instructing the vehicles to vacate) and issuing a landing clearance.

While the safety officers were obliged to notify the controller when a vehicle had vacated the runway, in accordance with airport operating procedures, the lack of notification was not a significant factor in the incident. Whether notification was provided or not, would not alleviate the responsibility of the controller to make a visual check of the runway to ensure the landing path was not obstructed prior to approving the aircraft to land. The action of the safety officer operating on the aerodrome control frequency while on the runway was an active defence in the incident. The officer heard the radio transmissions between the Chieftain pilot and the controller and was aware an aircraft was in the vicinity preparing to land. Consequently, he was prepared and able to vacate the runway in sufficient time for the Chieftain to land after the aircraft was observed on final. The officer's high level of situational awareness was the major factor that ensured the runway was not obstructed when the pilot landed without a clearance. However, the driver could have assisted the controller to maintain situational awareness by notifying that the vehicle was vacating the runway due to an approaching aircraft.

The pilot stated he became visual just before the minima and observed some vehicle activity near the runway edge. After checking the runway was clear he continued with the approach and landed. The controller did observe the flight of the Chieftain just prior to touchdown. However, after visually scanning the runway, observing the vehicles had vacated and believing that a landing was inevitable, he felt there was no need to distract the pilot with a late radio transmission during a critical phase of flight.

FINDINGS

1. The pilot of the Chieftain did not report at Nile outbound.
2. The controller incorrectly heard or misinterpreted the pilot's Nile inbound report.
3. The pilot did not report on final leaving 3,000 feet as instructed.
4. The safety officer observed the approaching aircraft and vacated the runway strip.
5. The safety officers did not report vacating the runway strip to the controller.
6. The controller observed the Chieftain just prior to touchdown.
7. The pilot of the Chieftain landed without a clearance.

The aircraft had started descent into Darwin when the right engine ITT and torque exceeded their maximum limits. The right power lever was then moved to flight idle but had no effect on the ITT and torque. A precautionary engine shut down was then completed. The flight continued uneventfully, and a landing made at Darwin.

The drive from the fuel pump to the fuel control unit failed.

Factual Information

Following an aircraft arrival to runway 23, a house owner in Lyons St Brooklyn Park complained that several tiles had dislodged from the house roof, some ridge capping had lifted, and a very strong, turbulent wind gust had hit the house.

At about the time of the incident, between 1730 and 1745 CST, five aircraft had made landings on runway 23 at Adelaide; there were no departures from this runway.

The surface weather conditions at the time were a wind of 150 degrees at 10 knots, temperature of 20 degrees Celsius and QNH 1023 mb. An inversion had formed, probably between 500 and 1,000 ft.

The house in Lyons St is located about 150 m to the right of the runway 23 extended centreline and about 1,000 m from the runway 23 touchdown zone.

Analysis

Aircraft landings on runway 23 during this period in order of arrival were a Boeing 737, a Piper Navajoe, a Boeing 727 and two Airbus A 320s. The Navajoe and one of the A320s joined the circuit on a long final approach. The other three aircraft joined from the right at about 3 km and 800 ft.

The radar data indicated that the approaches were identical from 3 km, all five aircraft passing 400 ft at 1.5 km. Beyond this point, estimates of altitude and distance were not accurate.

The house had been subjected to wind turbulence in the past. Although not so severe, previous turbulence had rattled windows and doors. On these previous occasions, the owner reported that the aircraft involved was a Boeing 727.

It is probable that the damage was caused to the house when the wake turbulence or wing vortices from one of the jet aircraft drifted across from a low level and reached the ground at the house's position.

Findings

1. A house at 12 Lyons St Brooklyn Park suffered roof damage.
2. The house is about 1 km from the touchdown zone of runway 23 and 150 to the right of the extended centreline.
3. About the same time, five aircraft landed on runway 23.
4. Each aircraft passed 1.5 km at about 400 ft.
5. The surface wind was 150 degrees at 10 knots.
6. It is probable that the downwash of one of the jet aircraft reached the ground at the house's position.

The pilot was flying the helicopter sideways,15 ft above the ground, as he attempted to herd a bull into a paddock. He felt a shudder through the tail rotor pedals and the helicopter immediately began to rotate to the right. Full left pedal did not stop the rotation. After two and a half turns the pilot was able to land the helicopter, cushioning the touchdown with the collective control.

During the landing the tail rotor struck a fence. A ground witness reported that the tail rotor system appeared intact until the collision. The engine was already stopped when the pilot attempted to shut it down after landing. The fuel tank contained 60 L of fuel, and no water was evident. Wind strength at the time of the accident was reported to be 5 kt.

Sideways movement, shuddering in the tail rotor pedals and loss of directional control to the right are all precursors or symptoms of a condition, in helicopters, commonly known as loss of tail rotor effectiveness. In this state the airflow through the tail rotor is reduced by the sideways movement and the relative wind (the combined effect of the actual wind and the direction of movement of the helicopter). As a result, the tail rotor can no longer supply sufficient thrust to counter the turning effect caused by the rotation of the main rotor system and the helicopter will yaw to the right. If the effect is severe enough, application of left pedal may aggravate the situation.

The evidence indicates that loss of tail rotor effectiveness is the most likely factor in this accident.

FACTUAL INFORMATION

A Boeing 737 aircraft, VH-CZM, inbound to Sydney from Brisbane was cleared to descend to FL190 and instructed to enter a holding pattern at YANGO. YANGO is a holding point to the northwest of Sydney.  A following Boeing 737 aircraft, VH-CZG, inbound to Sydney from Cairns was cleared to descend to FL200 by Brisbane air traffic control (ATC), with a requirement to reach FL200 by YANGO. The crew of VH-CZG then transferred from Brisbane to Sydney ATC and reported on descent to FL200. The Sydney controller instructed the crew of VH-CZG to enter the holding pattern at Yango and to maintain FL190. The crew readback the level and continued descent to FL190. Shortly after, the crew requested confirmation from Sydney control that the previous requirement to reach FL200 by YANGO was still necessary. The requirement was confirmed by the Sydney controller.

During a subsequent scan of his traffic, the controller observed that the radar display altitude label of VH-CZG was indicating FL190. The controller requested confirmation from the crew that the aircraft was maintaining FL200; the level indicated on the flight strip. The crew confirmed maintaining FL190 and stated they had been cleared to FL190 with a requirement to be at FL200 by Yango. The Sydney controller instructed the crew of VH-CZG to climb immediately to FL200 due to VH-CZM being in the holding pattern at FL190. The crew acknowledged and climbed the aircraft to FL200. Radar separation was maintained while vertical separation was less than the standard separation of 1,000 ft. There was no breakdown in separation.

ANALYSIS

The air traffic controller manning the Sydney radar position had recently been re-rated after an absence of approximately two years and was consolidating in the position at the time of the incident. Air traffic was light and he was endeavouring to ensure all aspects of the position were conducted in accordance with current procedures.

Despite the flight strips for both aircraft being annotated with different levels to provide vertical separation of 1,000 ft the controller instructed the crew of VH-CZG to maintain FL190. This was the level assigned to VH-CZM which was already established in the YANGO holding pattern. The controller may have confused the levels notated on the flight strips of VH-CZM and VH-CZG when he acknowledged the initial call by the crew of VH-CZG. This may have caused him to inadvertently transmit to the crew of VH-CZG to 'Maintain FL190' when he actually meant 'Maintain FL200' a level which would have ensured vertical separation between the two aircraft.

While the crew of VH-CZG had only been assigned descent to FL200, the instruction from the Sydney controller to "Maintain FL190" would have strongly suggested to them that the aircraft could continue descent to the lower level.

However, as the controller did not prefix the clearance for the lower level with "Descend to ..." the crew of VH-CZG should have queried the instruction. "Maintain (a level)" is only used to restrict a previously assigned climb/descent level or to confirm a requirement to maintain a current level. The term is not used to assign changes of level. A challenge of the instruction by the crew may have alerted the controller to his error.

By regularly scanning the radar display the controller was able to quickly identify and rectify the situation after observing the radar display altitude label of VH-CZG was the same as VH-CZM.

FINDINGS

1. The crew of the VH-CZM operated in accordance with their air traffic control instructions.
2. The controller inadvertently instructed the crew of VH-CZG to maintain FL190.
3. The crew of VH-CZG misinterpreted the instruction to maintain FL190 as a clearance for further descent.
4. The controller observed that the flight level on the radar display altitude label for VH-CZG was the same asVH-CZM and undertook corrective action.
5. Radar separation was maintained until the standard vertical separation of 1,000 ft was re-established.

The pilot and an approved testing officer were engaged in a commercial licence flight test. An aircraft landing area inspection of Wagin runway 06 and a touch-and-go landing were completed. This was followed by a period of instrument flying. The testing officer then gave the pilot a simulated engine failure, overhead Wagin airfield at 3500 ft. The pilot flew the aircraft to a short base leg for runway 34 at 1000 ft above ground level. Following the pilot's comment that the approach was high, the testing officer instructed him to do something about it. He lowered the nose and allowed the aircraft to accelerate whilst continuing the descent towards the strip. Speed over the runway threshold was faster than optimum and the aircraft landed, following a bounce, 600 m into the 1000 m strip.

As the pilot applied power to go around the testing officer observed powerlines across the departure end of the runway. He instructed the pilot to keep the aircraft on the ground, and they attempted to stop in the runway remaining. The runway surface included loose gravel and de-acceleration was poor. Main wheel skid marks were apparent for 300 m. The aircraft ran off the end of the runway and collided with a raised road surface and a water pipe. The left main and nose landing gear legs collapsed. The aircraft came to a complete stop 100 metres from the end of the runway.

The powerlines were not observed during the landing area inspection because it concentrated on runway 06 and not 34.

The testing officer later indicated that he had tried too hard to introduce some realism into the exercise. He had allowed a simulated situation to continue without ensuring that the aircraft could overshoot safely or stop within the runway confines, should it become necessary. The pilot had not taken independent action because he was under pressure to meet the test requirements and expected the more experienced testing officer to maintain a safety watch.

Factual Information

The aircraft had just completed a landing on runway 21 when the nose wheel collapsed. The engine cowl and propeller struck the ground, and the aircraft came to a stop.

The aircraft had a history of nose wheel shimmy on take-off and landing.

Analysis

The manufacturer's design specifies a Cessna 185 tail wheel as the nose wheel for this aircraft type. The nose wheel is fully castoring with no controls and is free to castor during ground operations. A fixed fairing over the nosewheel prevents any pre-flight inspection of the mounting.

This aircraft has experienced two previous nose wheel collapses. The aircraft owner is considering modifications to convert the aircraft to a tail wheel configuration.

The pilot attempted to land his lightly loaded aircraft uphill on his 425-metre private airstrip with a slight tail wind. Threshold airspeed was slightly fast. The aircraft touched down long. The pilot soon realised that he could not stop the aircraft before the end fence. When he decided to go-around and applied power, the aircraft was too slow to become airborne before the fence. He then closed the throttle, and the aircraft ran through the fence.

Numerous aircraft were inbound to Sydney, including a foreign Boeing 747 (B747) aircraft on a Rivet Three Standard Arrival Route (STAR) and a Fokker (FK50) aircraft on an Oakdale Two STAR. Both aircraft were at about the same distance from the airfield, with the FK50 sequenced to overfly to the east for a left circuit to runway 16L. The B747 was sequenced to remain to the west of the airfield for a right circuit to runway 16R. The FK50 was assigned 7,000 ft and the B747 assigned descent to 6,000 ft, with a requirement to reach that altitude by 14 DME. This was in order to achieve vertical separation before losing lateral separation between the two routes as the B747 was turned downwind, through the track of the FK50.

At 18 DME the B747 was observed on radar by the Approach South controller to be passing through 6,900 ft, and to make an authorised left turn. The aircraft was then expected to track 049 degrees until given radar vectors, but the B747 took up a track of 030 degrees, towards the track of the FK50. An instruction was given to the B747 to turn right immediately onto 090 degrees, and traffic conflict information was passed on the FK50. Because the B747 was initially slow to execute the turn a further instruction to turn immediately was passed. As the B747 turned away from the FK50, the lateral separation had reduced to 2 NM, with a vertical separation of 800 ft. The minimum required separation standard was 3 NM laterally, or 1,000 ft vertically.

The B747 crew were of the understanding that a clearance for a visual approach had been given and hence turned the aircraft onto 030 degrees to position the aircraft for a left circuit for runway 16R. The pilot in command of the B747 had the FK50 in sight and assessed that there was sufficient separation.