Assessment Summary Sheet for UKAB Meeting On 19 Jun 2013

Transcription

1 Assessment Summary Sheet for UKAB Meeting On 19 Jun 2013 Total Risk A Risk B Risk C Risk D Risk E Airprox Reporting (Type) Reported (Type) Airspace (Class) Cause ICAO Risk ERC Score A320 (CAT) A319 (CAT) UAR (C) The A320 crew believed they had been cleared to FL250, read back FL250. The controller was unable to detect the incorrect read back and the A320 descended into conflict with the A319. C A319 (CAT) EMB190 (CAT) Lon TMA (A) ATC released the A319 into conflict with the EMB190. C PA28 (Civ Trg) C182 (Civ Trg) F borough ATZ/Lon FIR (G) Odiham APP vectored the C182 into conflict with the PA28. C Hawk T2 (Mil) Hawk T1 Formation (Mil) Valley MATZ (G) In the absence of TI, a conflict resolved by the Hawk T2 instructor. C Parachutist (Civ Club) Rockwell RC114B (Civ Pte) Tilstock Para Site (G) The RC114B pilot flew through a promulgated and active parachuting site and into conflict with a parachutist, who he did not see. C Robin A120T (Civ Club) SU29 (Civ Pte) Lon FIR (G) An apparent non-sighting by the SU29 pilot. B CE560XL (Civ Comm) PA28 (Civ Club) F borough ATZ (G) ATC did not achieve the safe integration of traffic. C Vigilant T1 (Mil) Sport- Cruiser (Civ Pte) Lon FIR (G) Effectively a non-sighting by the Vigilant crew. C A319 (CAT) PA38 (Civ Club) Liverpool CTR (D) The A319 crew was concerned by the presence of the PA38 to the north. E Viking Glider (Mil) GA7 Cougar (Civ Pte) Lon FIR (G) A non-sighting by the GA7 pilot and a late sighting by the Viking crew. B Tornado GR4 (Mil) DG-808C (Civ Pte) LFA 14 /Scot FIR (G) A conflict in Class G airspace. B Typhoon FGR4 (Mil) Tucano T1 (Mil) Vale of York AIAA (G) A conflict in the Vale of York AIAA. C 2

2 UKAB EVENT RISK CLASSIFICATION TRIAL (V3) Question 1 Question 2 If this event had escalated into an accident outcome, what would have been the most credible outcome? Category 1 Multiple fatalities on ac with >19 seats or MTOW >5700kg or significant 3 rd party risk. What was the effectiveness of the remaining barriers between this event and the most credible accident scenario? Effective Limited/ Partial Minimal Not effective Category 2 Multiple fatalities on ac with 5-19 seats Category 3 Multiple fatalities on ac with <5 seats or multiple injuries or ac damage Category 4 No accident outcome. Possible operational implications (eg diversion) 1 Note: The Category scaling is for trial purposes; it does not reflect CAA or MoD policy or risk appetites. The reference to 19 seats and MTOW of 5700kg is for coherence with the requirement to fit TCAS. If the number of ac occupants is known it will be used to determine the Category; otherwise the aircraft seating capacity will be used.

3 AIRPROX REPORT No Date/Time: 2 Feb Z Position: 5156N 00324W (13nm NNW BCN) Airspace: UAR (Class: C) Reporting Ac Type: A320 A319 Operator: CAT CAT Alt/FL: FL250 FL340 (Saturday) (Night) Reporting Ac Weather: VMC NR VMC NR Visibility: >10km Reported Separation: Nil V/3nm H Recorded Separation: 100ft V/2 4nm H BOTH PILOTS FILED Nil V/1-1 5nm H 33: : : : : : :19 33: :11 A320 33: : C/L UL9 Radar derived Levels show Mode C 1013hPa 33:03 32:51 33: :35 32:23 32:15 BCN C/L UN : NM :27 31: A319 C/L UN : PART A: SUMMARY OF INFORMATION REPORTED TO UKAB THE A320 PILOT reports en-route to Manchester, IFR and in receipt of an ATS from London on MHz, squawking 7474 with Modes S and C. Heading 360 at 320kt and descending to FL250 another ac s strobe and nav lights were seen in their 2 o clock range 3-5nm at about their level. A TCAS TA alert was received and they reduced their ROD to 400ftpm and the other ac passed 3nm to their R and behind. He assessed the risk as low. THE A319 PILOT reports en-route to Dublin, IFR and in receipt of an ATS from London squawking 1436 with Modes S and C. Level at FL340 at 390kt having been cleared direct LIPGO, a TCAS TA alert was received on an ac crossing their track in a leisurely descent. It was sighted at about 3-4nm range and was seen to cross and descend through their level at about 1-1 5nm range. He informed ATC and an immediate L turn was given before clear of conflict was received 30sec later; no TCAS RA was received. The other ac appeared not be on their frequency, he thought, and ATC advised that a report would be filed. He assessed the risk as high. THE S5/6/8/9/23/35/36 TACTICAL CONTROLLER reports taking over the W bandboxed Sector at 1730 with the A319 maintaining FL340 routeing MEDOG-LIPGO and the A320 N bound at FL360 destination Manchester. The 2 ac would cross at approximately 90 E of MADOG. Because the A320 would need descent shortly, he cleared the flight to FL350 on top of the A319. The crew did not read this instruction back so he later repeated the descent clearance to FL350; this clearance was read back correctly, he thought. Whilst controlling other traffic he noticed STCA flashing white (low severity) between the 2 ac but he was not concerned as this is a regular occurrence when 1 ac is approaching 1000ft either on top of or underneath another. He was also confident that the A320 crew had read back the correct clearance and he continued to control other traffic on the sector. He did not notice the Mode S SFL readout of the A320. A short time after the A319 crew queried the presence of the A320 which was now in the A319 s 1 o clock position approximately 2 5nm away and 300ft above but flying away. He gave avoiding action to the A319 flight to re-establish 5nm separation and when clear he instructed the flight to resume own navigation to LIPGO. He did not give any avoiding action instruction to the A320 flight as its track was the best in order to achieve the 1

4 required 5nm. STCA did not flash red (high severity) until the A319 queried the A320 s presence. Minimum separation was 300ft/2 5nm. THE S5/6/8/9/23/35/36 PLANNER CONTROLLER reports the W End Sector was all bandboxed. At a position just N of DIKAS the A320 flight was cleared to FL350 on top of crossing traffic, an A319, at FL340 which was routeing MEDOC-LIPGO. The A320 crew read back FL350 and shortly afterwards the ac labels merged. The A319 crew reported traffic level with them at range 1 5nm so she pulled the labels of the 2 ac apart and couldn t believe the A320 was now also at FL340. The Tactical controller issued avoiding action to the A319 flight to turn L heading 250. Minimum separation was 300ft and 2 4nm. ATSI reports that the Airprox occurred at 1733:25 (UTC) within Class C CAS, 7nm ESE of waypoint MEDOG, between an A319 and an A320. The A319 was W bound IFR at FL340, inbound to Dublin and in receipt of a RCS from London Control on frequency MHz. The A320 was N bound IFR at FL360, inbound to Manchester and in receipt of a RCS from London Control on frequency MHz. London Area Control (LAC) Swanwick Local Area Group West (LAG (W)) known as West End were operating as London Control in a combined bandboxed mode, combining sectors 5, 6, 8, 9, 23, 35, and 36, with cross-coupled combined frequencies; 133 6, , , , , and MHz. The area covered by the combined sectors is shown below in Figure 1. (Figure 1 Area covered by the combined LAC(W) sectors). Swanwick LAC utilise the interim Future Area Control Tools Support (ifacts), which uses Trajectory Prediction, Medium Term Conflict Detection, and Flight Path Monitoring, to provide controllers with decision-making support and to assist in managing workload. LAC MATS Part2, GEN-46, paragraph , states: The ifacts tools are to be used, in conjunction with the radar display, for the detection of conflictions and assessment prior to issuing clearances. All data, including Tactical Data, must be input into the tools. When entering Tactical Data electronically, the principal of Enter As you Speak, Read As you Listen should be applied, in order to ensure the accuracy of the data entry and pilot read back. Team members shall check for system conditions and error messages, including the Flight Messages Window as part of their routine scan and ensure that other team members are informed as appropriate. 2

5 The combined West End sector was manned by a Tactical (T) and Planner (P) controller. The T controller had taken over the position at 1730:00, 3min prior to the incident and the P controller had been in position since 1700:00. Both controllers were experienced on the sector. The T controller is a Local Area Supervisor (LAS) with additional watch management responsibilities, required to work a minimum of 14hr on operational position within the preceding 30 day period in order to maintain operational competence. The operational hours for the 2 controllers is shown below: Month T controller Planner controller December hrs 38mins 41hrs 12mins January hrs 43mins 32hrs 23mins CAA ATSI had access to RT recordings, area radar recordings, the written reports from the pilots of each ac, together with written reports from the T controller, P controller and the watch management investigation report. CAA ATSI interviewed the two controllers involved and was able to view a replay of ifacts recording. A frequency occupancy analysis (clock busy chart) was provided by the ANSP for the period 1700 to 1759 and showed that the sector workload increased at 1731:00, just prior to the incident (Figure 2). Figure 2 Clock busy chart for LAG(W) for the period UTC. At 1728:04, the A320 squawking 7474 was N bound at FL360 and the A319 squawking 1436 was W bound at FL340. The distance between ac was 42 8nm as shown in Figure 3. (Figure 3 Swanwick MRT at 1728:04) 3

6 At 1730:45, the T controller having just taken over the position, transmitted, (A320 c/s) London you can descend to flight level three five zero and then simultaneously entered 350 as the new Cleared Flight Level (CFL) into ifacts. However there was no reply from the A320 crew and the controller transmitted to another flight (AC1). ATC AC1 (AC1 c/s) set course to SOMAX Set course to SOMAX (AC1 c/s) At 1731:03, ifacts generated a Selected Flight Level (SFL) alert to show that the A320 s CFL (350) entered into ifacts differed from the SFL (360) and the alert is shown on the bottom line of the Target Data Block (TDB) as a white (360) in figure 4. (Figure 4 ifacts alert showing A320 CFL as 350 and SFL as :03) At 1731:14, the T controller transmitted again, (A320 c/s) descend flight level three five zero and the A320 pilot responded, Descend flight level????? five zero (A320 c/s). The A320 s SFL then changed to 250 and, at 1731:28, ifacts generated a white (250) alert showing that the SFL did not agree with the CFL entered into ifacts (Figure 5). (Figure 5 ifacts alert showing A320 CFL as 350 and SFL as :55) The P controller was engaged in an operational phone call from 1731:15 until 1731:27 and did not hear the read-back. 4

7 The T controller indicated that he was aware of the first alert (360) but not the second (250). The controller was confident that the A320 pilot had given a correct read-back FL350 and felt assured that separation would be maintained. CAA ATSI analysed these transmissions and made the following observations: a) the two instructions transmitted by the T controller to the A320 giving descent clearance to FL350 were clearly distinguishable. b) the read back from the A320 on a single replay could easily have been mistaken for FL350. c) neither the controller nor pilot used the phonetic TREE to distinguish between 2 and 3. d) by slowing the speed of the recording, varying the tone and looping the track, CAA transcription were clearly able to identify the read-back as FL250. e) CAA ATSI did not have access to any cockpit voice recordings and were therefore unable to evaluate the transmissions received by the A320 crew. At 1731:34, the T controller continued talking to other flights. AC2 London Control good evening (AC2 c/s) climbing FL 150 SUPAB ATC (AC2 c/s) London roger climb to FL260 AC2 Level 260 (AC2 c/s) ATC (AC3 c/s) contact Brest on AC (AC3 c/s) Bye Bye ATC Goodbye AC4 Good evening (AC4 c/s) climbing FL350 ATC (AC4 c/s) London route direct SUPAB At 1732:15 low-level Short Term Conflict Alert (STCA) activated. The A320 was passing FL354 (1400ft vertical separation) and the distance between the 2 ac was 11 4nm (Figure 6). At the same time AC4 queried its routing. AC4 Please repeat that point where we are cleared to navigate (AC4 c/s) (Figure 6 STCA activates at 1732:15 vertical separation is 1400ft) The T controller indicated that he had observed the STCA alert and reported that it wasn t uncommon for such alerts to be generated when ac approached their cleared level. The T controller remained confident that the A320 would stop descent at FL350. At 1732:23 whilst responding to another flight, the T controller repositioned the label of A320 (FL353) as shown in Figure 7. 5

8 (Figure 7 STCA alert after the T controller moved the A320 label :23) ATC AC4 ATC AC5 ATC Are you routeing via KORUL Standby (AC5 c/s) descend FL220 level VATRY 220 to be level by VATRY (AC5 c/s) (AC4 c/s) Say again At 1732:35, the A320 s Mode C indicated FL350 and the distance between the 2 ac was 9nm. AC6 ATC London (AC6 c/s) climbing altitude FL - climbing FL130 heading 290 degrees (AC6 c/s) climb to FL150 what s your requested level At 1732:52, as the A320 passed FL347 (700ft vertical separation) ifacts generated a White CFL Deviation Alert on the top line of the A320 s TDB, indicating that the A320 had deviated from the CFL by more than 200ft. Radar showed the distance between the 2 ac was 7 1nm (Figure 8). (Figure 8 - STCA, the CFL deviation alert and incorrect SFL alert :52) At the same time the ifacts Separation Monitor, situated on the bottom LHS of each of each controller s situation display showed interactions between the 2 ac. The top red/black flashing interaction is generated when separation is predicted to be lost within 3min and the bottom white/black flashing interaction is generated when the system recognised that there is uncertainty in the outcome. These are circled yellow in Figure 9. 6

9 (Figure 9 ifacts separation monitor interactions :52) At this point the P controller was engaged in the coordination of an AT75 ac into the sector, routeing from Birmingham to join CAS at MOSUN. The P controller was busy inputting route details into ifacts and the P controller s situation display showed the labels of the 2 ac overlapping, as shown in Figure 10. (Figure 10 Overlapping labels on the P controllers situation display :52) AC6 ATC Climb FL150 and requesting FL360 (AC6 c/s) Copied A TCAS simulation tool (see report below) suggested that each ac received a TCAS TA at 1733:02, as the A320 passed FL346 and crossed the track of the A319 from L to R, when the distance between the 2 ac was 5 7nm. The A320 pilot s written report indicated that at this point he adjusted the ROD from 1000fpm to 500fpm then to 0fpm. At the same time 1733:03, AC7 contacted the sector. AC7 was positioned 67nm S of BHD on the Southern edge of the T controller s situation display as shown in Figure 11. AC7 London good evening to you it s the (AC7 c/s) descending FL300 towards Berry Head 7

10 (Figure 11 NODE radar showing the relative position of AC7 at 1733:03) ATC (AC7 c/s) hello route to EXMOR for the Bristol 2B correction Bristol 2D At 1733:11, separation was lost between the 2 ac as horizontal distance reduced to 4 8nm whilst the vertical separation was 400ft (Figure 12). (Figure 12 the T controller s situational display at 1733:12) AC7 Route to EXMOR for the 2D and don t suppose there are any rugby fans amongst you over there. [The T controller responded with a short non-standard comment.] At 1733:21, the P controller, having completed coordination, then separated the 2 overlapping labels on the situation display. The A320 was passing FL343 and the range between the ac had reduced to 4nm. At 1733:24, the T controller initiated a telephone call to LAG N Sector and at the same time the A319 pilot reported, London er (A319 c/s) we ve got a contact at 200ft about 2 miles ahead of us. At 1733:27 the T controller highlighted the TDB of the A319 (Figure 13). 8

11 (Figure 13 ifacts T controller's display :24) At 1733:33, high level (RED) STCA activated and the T controller immediately responded, (A319 c/s) roger turn avoiding action turn left immediately heading 250 degrees (crossed transmission from A319) it s actually out of our way but er he s level with us right now about a mile and a half two miles. The controller replied Copied. The vertical distance was 200ft (Figure 14). (Picture 14 - High Level STCA activated at 1733:33) [UKAB Note (1): Minimum horizontal separation of 2 4nm is shown on the 2 radar sweeps at 1733:43 and 1733:47, the A319 maintaining FL340 as the A320 is shown descending through FL342 and 341 respectively.] At 1733:47, the T controller asked the A320 pilot, (A320 c/s) just tell me your cleared level and the pilot replied, Cleared level two five zero (A320 c/s). The T controller still had the telephone line open to LAG N and advised them to disregard, terminating the telephone call. The A319 flight was then cleared to resume own navigation to LIPGO and advised that the other ac was clear of them. At 1734:23, separation was re-established as horizontal distance increased to 5nm. 9

12 At 1736:00, the T controller handed-over the position to an oncoming controller. At 1737:34 the A319 pilot confirmed that he would be making an Airprox report. The P controller handed over the position to an oncoming controller at Later the T controller indicated that he considered it to be a normal working day; there were no distractions and he was operating the combined bandboxed sector with a light to medium workload. The T controller observed the SFL 360 alert after the initial missed call but did not see it subsequently change to 250. The T controller was convinced that the A320 crew had given a correct read back of FL350 and even after hearing a replay of the recording remained certain that this had been the case. The controller was interested to know about the circumstances that led to the A320 crew mishearing 350 and indicated that there had been some general discussion on the unit about the difficulty in distinguishing between 2 and 3 and the emphasis of using phonetics such as TREE. When the T controller observed the STCA and noted that the vertical separation was 1400ft, he commented that this was a routine occurrence when ac approach their cleared level and was confident, because of the read back, that the A320 would level off at FL350. He was then absorbed talking to other flight and had not observed the CFL deviation alert or interaction shown on the separation monitor, which indicated that the A320 had descended more than 200ft below its cleared level. In discussion, the T controller did not consider that these alerts were sufficiently visible or eye catching and suggested an alternative such as flashing red text. The ATSU watch investigation report suggested that consideration be given to providing the T controller with an ifacts refresher session together with a UCE. ifacts was fully implemented in November 2011 and the T controller was asked if he considered that working the minimum number of hours to maintain competency levels was sufficient to maintain his familiarity with the ifacts. The T controller indicated that whilst he and other supervisors would prefer to work more operational hours this was not always possible, but regarded himself as experienced and the number of hours worked as sufficient to maintain competency. The P controller confirmed that the day had been normal with no distractions. Traffic levels were low to medium and the complexity and workload consistent with bandboxed operations. In discussion, the P controller indicated that it was sometimes a balance to ensure that workload was sufficient to maintain the concentration levels. The P controller did not consider the range of the situation display in bandboxed configuration was excessive for the levels of traffic. When questioned about random or nuisance STCA alerts, the P controller indicated that these did occur but were not really an issue, provided that the alert was monitored, taking appropriate action as required until the situation was resolved. With regard to ifacts alerts, the P controller commented that these were not easily noticeable especially when things started to happen quickly and in discussion suggested perhaps an alert flashing red and also flashing red on the strip bay. As a result of this incident a number of actions have been taken by the ANSP: a) A review of the Separation Monitor to examine the saliency of alerts. b) A two stage flashing SFL alert (in line with the development of Swanwick AC ifacts track data block human factors review) is being developed with the aim of making the information more prominent. c) A defensive controlling package is being produced which will include reference to the passing of TI in situations where STCA has, or is likely, to activate. d) The ANSP human factors group are conducting a review of the risks associated with supervisors achieving minimum hours on radar over a prolonged period. 10

13 Analysis of the RT recordings showed that at 1730:45, the controller transmitted a clear instruction to the A320, (A320 c/s) London you can descend to Flight level three five zero. In the absence of a response from the A320 crew, the controller made a second clear transmission at 1731:14, (A320 c/s) descend Flight Level three five zero. The response from the A320 pilot Descend Flight Level????? five zero (A320 c/s) was unclear and could easily have been mistaken for FL350. The T controller indicated that he believed that the A320 pilot had given a correct read-back of FL350 and still believed that this was the case. An analysis of the recording by CAA transcription unit showed that whilst the read-back could easily have been mistaken for FL350, it was FL250. CAA ATSI considered that given the nature and quality of the incorrect read back, the T controller could reasonably have accepted it as being correct. In consultation with CAA FOI, the A320 operating company were asked to comment on the flight deck perspective and possibility of distraction or workload factors that may have contributed to the first missed first call and then the incorrect read back by the A320. The operating company had not completed their own investigation but provided the following response from the crew: First Officer (PNF): I remember the event pretty well. I heard a clearance of descend FL250 which I read back over the radio, the captain set on the FCU what I read back and so we started to descend. At first I did wonder to myself that the clearance was a little early for our descent profile and to a lower FL than expected with so many track miles to go however I remember reading the clearance back very clearly and that my response was not challenged by ATC. Therefore I rightly or wrongly thought that it must have been correct. I should have in hindsight challenged my gut instincts with ATC. It is very possible that I miss heard the clearance but I am 100% sure that I read back the clearance that I thought I heard very clearly. After we realised that we had an Airprox event the captain and I had a discussion it [sic], the captain also thought he had heard FL250 and that was backed up by my response to ATC and so what he set on the FCU. Captain (PF): I've no issues or anything to contradict [FO's] recollection of events. Like he, I too thought the clearance to FL250 was early, but not unusual in my experience in that airspace, we'll often get FL330 by EXMOR, but not on this occasion. As [FO] said, we had a lot of track miles and I'm sure I mentioned it in the ASR (MOR) I selected -1000fpm because of this. As we got the TA I backed this off first to -500fpm and then to 0fpm until clear of the other traffic when I resumed the -1000fpm ROD to FL250 and [FO] as PNF, along with the A319 crew, queried the original clearance, the new controller, who had taken over, asked us what we had been cleared to, and confirmed our continued descent. The increase in workload, just prior to the Airprox, together with the effect of the combined bandboxed sectors and the scale and range of the controller s situational display were considered by CAA ATSI. The LAG(W) sectors sector configuration is shown below. Time Sector configuration 00:00 06:20 Operating as 5/6/8/9/23/35/36 all bandboxed 06:20 06:35 Operating as 5/23 and 6/8/9/35/36 06:35 11:18 Operating as 5/23 and 6/9/36 and 8/35 11:18 14:20 Operating as 5/8/23/35 and 6/9/36 14:20 00:00 Operating as 5/6/8/9/23/35/36 all bandboxed The LAC MATS Part 2, page MORS-50 and MOPS-46, state: 11

14 The Local Area Supervisor (LAS), in consultation with the sector team, shall decide when a sector can be bandboxed. The LAS shall also use flow and staffing information to make the decision about bandboxing. Ensure all sector team members are informed that bandboxing will take place. It is recommended that the LAS should use flow information and information from the sector team operating a bandboxed sector to pre-empt the need for staff to be recalled. The Traffic Load Prediction Device (TLPD) histogram charts for 0630, 0930, 1330 and 1630 on the 2 Feb 13 are shown in Figure 15 below: Figure 15 TLPD histogram charts for 0630, 0930, 1330 and An interrogation of the TLPD showed a reduction in predicted traffic levels after With these predicted traffic levels the afternoon LAS (W) agreed to combine all of the sectors at Network Management subsequently confirmed that these predicted traffic levels would be manageable in bandboxed configuration. Anecdotal evidence suggested that this was quite early in the afternoon for all LAG (W) sectors to be combined but not unheard of for the predicted traffic level. The sectors remained combined for the remainder of the day with TLPD predicting that a split would not be necessary. 12

15 During further discussion, the operational controllers regarded the traffic levels on the combined sectors as being light and that the range of the situational display was normal for that mode of operation. The number of operational hours on console completed by the T controller was considered to be the minimum required to maintain competence. The LAC MATS Part 1, 8/2/1 2.1(b), states: Air Traffic Controllers must comply with the unit competency scheme. NATS Unit Competency Scheme, Page 14, states: Paragraph 11: Rostering The Operational Resource Team will provide individual controllers with a monthly roster which, as far as reasonably possible, provides sufficient rostered duties to satisfy currency requirements. Swanwick: GSs with radar validations and LASs with tactical validations should be rostered, as far as is reasonably possible, for a minimum of 4 radar duties per calendar month. P controller only LASs should be rostered a minimum of 3 P controller duties per month. Page 15 Para 12 Maintaining Currency - Minimum Monthly Hours In order to maintain competence controllers shall be required to maintain a minimum level of currency in their tasks. The achievement of the relevant number of hours below will not in isolation be considered sufficient to maintain competence. Time spent operating as an OJTI or under training for an additional UE/rating will not count towards the currency requirement. Controllers hours should be reasonably spread over the month and they must record their hours in an approved medium which must be produced upon request. The operational hours of the T controller during the 2 previous months was regarded by NATS and the CAA to be the minimum required to satisfy competency requirements. However ATSU Watch investigation report suggested that consideration be given to providing the T controller with an ifacts refresher session together with a UCE. It was not possible to determine why the A320 crew missed the first clearance (FL350) which at this point was entered into ifacts by the T controller. This resulted in ifacts generating an alert to show that the CFL (350) did not agree with the SFL (360). This was noticed by the T controller who transmitted the clearance again (350). The second transmission was misheard by the A320 crew and an analysis of their read back transmission determined that it was difficult to distinguish between FL350 and FL250. The T controller was, and remains, convinced that the read back of 350 was correct. It was unfortunate that neither the T controller or the A320 crew made use of the phonetic term TREE which may have helped distinguish between 2 and 3. CAP 413, Chapter 2, Page 1, Paragraph 1.1, states: the use of non-standard procedures and phraseology can cause misunderstanding. Incidents and accidents have occurred in which a contributing factor has been the 13

16 misunderstanding caused by the use of non-standard phraseology. The importance of using correct and precise standard phraseology cannot be over-emphasised. Once the read back error passed undetected, ifacts changed the SFL alert to show that the CFL 350 did not agree with the SFL 250. However this alert went unnoticed by the T controller, very likely because once the first alert was highlighted and noticed by the T controller, the second alert would not have been as apparent, especially as the T controller then started to interact with other flights. LAC MATS Part 2, GEN-115, paragraph 8.4.3, states: The checking of SFL is not a mandatory task for the controller, although it is encouraged for early identification of possible level busts. The display of SFL is not a substitute for RT read back, which remains a mandatory controller task. STCA will trigger when an ac is operating outside the parameters set for STCA, and a contributory factor can be the high ROD of an ac, which may be the first indication of a potential level bust. Some STCA alerts are considered by controllers to be spurious and the P controller indicated that these are not really an issue provided that appropriate measures were taken to monitor the situation, taking appropriate action as required. The Manual of Air Traffic Services (MATS) Part 1, Section 1, Chapter 5, page 18, paragraph 21.1, states: In the event an STCA is generated in respect of controlled flights, the controller shall without delay assess the situation and, if necessary, take action to ensure that the applicable separation minimum will not be infringed It is likely that the trajectory and ROD of the A320 (1000fpm) with traffic crossing 1400ft below infringed the STCA parameters and triggered the low level alert. After the STCA was generated (1732:15) the T controller was convinced that the A320 would level off at FL350 and assessed the STCA to be spurious. As the T controller spoke to another flight he deconflicted the TDB labels, which was probably an automated response and he did not notice the SFL alert showing the A320 had selected the incorrect level. The T controller was then absorbed talking to other flights (5 transmissions) and did not observe the A320 as it passed FL350 (1732:38) or the ifacts (2 transmissions) deviation alert (1732:52) when the A320 passed FL347 (2 transmissions). Then at 1733:03 AC7, 67nm S of BHD and at the southern edge of the T controller s situation display, called. It is likely that at this crucial point the controller s focus of attention was on AC7. At 1733:24 the A319 reported the A320 was crossing 2nm ahead and 200ft above. The T controller s conviction that the 2 ac were safely separated likely lulled him into a false sense of security. This perception and expectation bias probably caused the controller not to monitor STCA alert at a crucial point when Mode C showed the A320 descend below its cleared level and when ifacts generated the deviation alert. The T controller had just taken over the combined West End sectors 3min prior to the Airprox. Team members are required to check ifacts system conditions and error messages as part of their routine scan ensuring that other team members are informed as appropriate. However the P controller was engaged in a coordination task, leaving just the T controller to monitor the combined LAG (W) sector at the time of the STCA alert and encounter. The range of coverage of the bandboxed sectors, together with the increase in workload and the relative positions of the other ac, likely contributed to diverting the T controller s focus of attention away from the STCA and ifacts alerts. The T controller had completed the minimum number of hours required to maintain operational competency, the subsequent watch investigation report recommended that consideration be given to providing the T controller with ifacts refresher training. The policy of combining the sector in quiet periods is determined by the predicted traffic levels and staffing information. Although the predicted traffic levels indicated that traffic levels would be manageable in bandboxed configuration, CAA ATSI consider that had the sector remained split for longer in the afternoon period, the T controller would 14

17 have had more time to monitor the 2 ac, with the possibility that adjacent sector team members may have provided a prompt or warning to alert the T controller. Both the T controller and P controller commented on the fact that the ifacts deviation alert was not attention grabbing and suggested that a more prominent flashing red would be more appropriate together with a red warning on the strip bay. The nature and quality of the A320 crew s incorrect read back, not unreasonably, caused the T controller to accept it as being correct (FL350). The T controller did not notice that the A320 s SFL was indicating FL250. When STCA activated the T controller was convinced that it was spurious and that the A320 would level off at FL350 with vertical separation assured. As a result the T controller did not monitor the 2 ac labels or observe the Mode C readout of the A320 as it descended through FL350. The T controller did not notice the deviation/interaction alerts generated by ifacts at the crucial point, when the A320 continued its descent and into conflict with the A319. CAA ATSI considers that the T controller s expectation bias resulted in him not giving sufficient priority to the monitoring of the STCA or to ifacts at a time when alerts were being generated. Recommendations CAA ATSI is content with the actions already underway by the ANSP. It is recommended that the ANSP, in the course of their future development of ifacts, give consideration to making deviation alerts more prominent and noticeable. It is recommended that the ANSP review the guidance for bandboxed operations with a view to taking into account the combined sectors ability to sufficiently monitor and interact with generated alerts. NATS TCAS PERFORMANCE ASSESSMENT Mode S Downlink no TCAS RAs were recorded via Mode S downlink. InCAS Alert Statistics A319 Alert Time Alert Description Altitude (FL) Intruder Range (nm) Vertical Sep (ft) 1733:02 TRAFFIC ALERT A320 Alert Time Alert Description Altitude (FL) Intruder Range (nm) Vertical Sep (ft) 1733:02 TRAFFIC ALERT CPA CPA Time Horizontal Sep (nm) Vertical Sep (ft) 1733: Minimum Lateral Separation Min LatSep Time Horizontal Sep (nm) Vertical Sep (ft) 1733: Minimum Vertical Separation Min VertSep Time Horizontal Sep (nm) Vertical Sep (ft) 1734: Greatest Erosion of (5nm/1000ft) Standard Separation Time Horizontal Sep (nm) (% of Std) Vertical Sep (ft) (% of Std) 1733: (49%) 162 (16.2%) 15

18 Eurocontrol s automatic safety monitoring tool (ASMT) did not record any RAs relating to this encounter. InCAS simulation based on Clee Hill single source radar data suggested that each ac received a traffic alert (TA) at 17:33:02. The time to closest point of approach (tau) required for an RA at this altitude is 35sec. In simulation the minimum value of tau observed was 32sec, however no RA was issued, because the horizontal miss distance filter (HMD) was active. The HMD is designed to minimise nuisance RAs. Above FL100, unless the horizontal and vertical separations are predicted to be simultaneously less than 2nm and 750ft respectively an RA would be considered as nuisance. In this case, the predicted (and actual) horizontal miss distance was approximately 2 5nm and therefore no RA was issued. PART B: SUMMARY OF THE BOARD'S DISCUSSIONS Information available included reports from the pilots of both ac, transcripts of the relevant RT frequencies, radar video recordings, reports from the air traffic controllers involved and reports from the appropriate ATC authorities. Although it was clear to Members that the initial trigger element to this Airprox was the RT exchange between the A320 crew and the LAC West End Tactical controller regarding the ac s descent clearance, the comprehensive ATSI report had uncovered several other elements which resulted in a lengthy discussion by the Board. A CAT pilot Member commented that the descent clearance to FL350 issued twice by the Tactical controller had been clear but the quality of the A320 crew s read back received by the Tactical controller had been unclear and the Tactical controller was unable to detect the incorrect read back. An ATCO Member noted that when the Tactical controller issued the descent clearance for the first time he had used non-uk standard phraseology by inserting the word to ahead of the cleared flight level. This word should always be omitted when clearing a flight to a FL but should be used when altitudes or heights are involved. The Board also discussed the apparent widespread lack of use, in the experience of Members, of the phonetic pronunciation of the number tree which may have highlighted to the crew the correct tens of thousands of feet in the cleared FL issued or made the crew s read back clearer. The use of the word to in the first transmission ahead of the FL may have influenced the A320 crew s assimilation of the descent instruction was to FL250; however, it is equally possible that the first descent clearance went unheard by the crew as they did not reply. The correct phraseology was used on the second transmission, which was acknowledged by the A320 crew. Without any further information it was unclear why the A320 crew had not replied to the first descent clearance or why they had misheard their cleared level of 350 which was read back as 250. Pilot Members advised that there may have been valid operational reasons - cockpit noise or crew carrying out a briefing (Top of Descent); however, for whatever reason, the crew perceived FL250 to be their cleared level and, as this was not challenged by the controller, the ac was descended into conflict with the A319 which had caused the Airprox. After the A320 had started to descend there were a number of elements that could have broken the chain of events. The Tactical controller had seen the discrepancy between the CFL and SFL when he first entered the CFL 350 into ifacts and before the A320 crew changed their SFL; however, the subsequent change to FL250 went unnoticed. A pilot Member remarked that the colour used to highlight the discrepancy appeared not to be outstanding enough and that perhaps another form of attention getting, such as a flashing alert, would be more appropriate. A controller Member commented that at ScACC a CFL/SFL discrepancy would generate a flashing alert to the controller. The NATS Advisor informed Members that the 2 systems were indeed different but that HF work is ongoing, with respect to optimising the use of flashing alerts as warnings particularly when STCA is available. Putting aside that it was not mandatory for controllers to check for CFL/SFL discrepancies 16

19 a controller Member advised that it would be impractical in a busy TC sector - Members were concerned that the frequency of STCA alerts had apparently created a mindset that, more often than not, these alerts were spurious. The NATS Advisor informed Members that STCA parameters are adjustable and, through fine tuning over time, trigger levels have been set to give a warning in enough time for a controller to assess the situation and take action to resolve a conflict; both NATS and CAA are content with the STCA parameter set-up. In the firm belief that the A320 was descending to FL350 and would level-off (expectation bias), the Tactical controller had repositioned the A320 s label but he did not assess the deteriorating situation and turned his attention to other sector traffic. As the workload increased, the A320 s descent through FL350 and the ifacts Deviation alert also went unnoticed to both Tactical and Planner controllers. After the A319 crew had informed the Tactical controller of the A320 crossing ahead he highlighted the A319 s TDB and then, immediately after STCA triggered a high-severity (red) alert, he issued the flight with an avoiding action L turn. However, by then the ac had crossed, with the CPA occurring shortly afterwards with the flights diverging. A CAT pilot Member questioned the wisdom of clearing the A320 flight to descend just 1000ft at this range from its destination since every clearance takes up some of the available time on the RT and provides opportunities for errors. CAT pilot Members noted that both of the A320 pilots had thought that the descent clearance, erroneously believed to be to FL250, was earlier than that required for their flight profile and that with the benefit of hindsight, they should have questioned it. Members also discussed the A320 crew s action of reducing their ROD in response to a TCAS TA, which was contrary to the published guidance. The TA should be looked upon as a heads-up/get-ready warning that action may be needed in anticipation of an RA. Adjusting the flight path in response to a TA may cause TCAS to recalculate the optimum resolution at a critical stage as the ac are approaching the protective bubble point when an RA would be generated. That said, the A320 crew had spotted the A319 s lights prior to the TCAS TA, and the A319 gained visual contact after receiving a TA; both crews then monitored their flight paths, the A320 crew assessing that remaining above the other ac was their safest option. The radar recording shows the A320 crossing 4nm ahead and 300ft above the A319. Taking all of these elements into account, the Board concluded that any risk of collision had been effectively removed. Assessing the safety barriers, the Board acknowledged how difficult it had been for the controller to detect the incorrect read back. However, none of the procedures, systems or warnings that might have alerted him to the developing conflict were successful, leading the Board to conclude that the suite of ATC barriers was ineffective in this incident. What remained was the aircrews SA from their TCAS and, providentially in Class C airspace, visual sightings with TCAS RAs in reserve. Given the crews SA, and with every prospect that the robust barrier of TCAS RAs would have been effective, the Board assigned an ERC score of 50 to the Airprox. PART C: ASSESSMENT OF CAUSE AND RISK Cause: The A320 crew believed they had been cleared to FL250 and read back FL250. The controller was unable to detect the incorrect read back and the A320 descended into conflict with the A319. Degree of Risk: C. ERC Score:

20 AIRPROX REPORT No Date/Time: 28 Feb Z Position: 5150N 00011W (6nm NW BPK) Airspace: LTMA (Class: A) Reporting Ac Reported Ac Type: A319 EMB190 Operator: CAT CAT Alt/FL: 5000ft NR (QNH) (QNH) Weather: VMC CLAC VMC NR Visibility: NR >10km Reported Separation: 400ft V/3nm H Recorded Separation: Nil V/4 3nm H OR 1700ft V/1 1nm H NR LUT 0723;31 A45 Luton ~5nm A319 LUTON CTR SFC-3500ft LTMA 3500ft+ LTMA 2500ft+ A59 23:51 A49 A50 23:59 A50 A50 24:07 A48 24:11 A53 CPA 24:27 24:15 A55 24:15 A45 24:11 A46 A42 24:07 A52 A51 23:59 A50 A50 23:51 A50 EMB ;31 A50 LUTON CTA ft LTMA 3500ft+ 0 1 BPK ~5nm NM PART A: SUMMARY OF INFORMATION REPORTED TO UKAB THE A319 PILOT reports outbound from Luton, IFR and in receipt of a RCS from London on MHz, squawking 3454 with Modes S and C. Climbing to 5000ft, heading 110, he thought, and flying into sun at 240kt, ATC told them to descend to 4000ft which was flown manually with AP off. Traffic was sighted turning in level flight in their 2 o clock range 2-3nm about ft higher and a TCAS TA was received at the end of the manoeuvre. The frequency was congested and blocked transmissions from other flights led to a later call to ATC than desirable although this was not unusual at the time of day. He assessed the risk as high. THE EMB190 PILOT reports outbound from London/City, IFR and in receipt of a RCS from London, squawking an assigned code with Modes S and C. Both he and the FO considered this not to be an Airprox. Just S of Luton flying at 250kt they were given avoiding action heading by ATC, a L turn, and they were visual with the other ac and didn t get close at any point. There were no TCAS TA or RA alerts/warnings and he believed the avoiding action turn was overkill. He assessed the risk as none. THE LTC NE DEPS RADAR CONTROLLER reports working the NE/LAM/LOREL position bandboxed. The TC NE workload was high with a complex situation at BPK owing to E ly operations but it was very quiet into LAM so she assessed that LAM did not need to be split off as this would not have affected the complexity of the traffic. The EMB190 was released by the Coordinator from London/City on a BPK SID followed by the A319 being released from Luton on a CLN SID. The EMB190 was airborne as the A319 was released and she climbed the EMB190 to altitude 5000ft under a Heathrow BPK departure before she became busy elsewhere with departures; the EMB190 was missed in her scan potentially owing to garbling. As the EMB190 approached BPK the Coordinator requested that she climb the EMB190 to altitude 6000ft at the request of LTC NW. The A319 flight checked-in whilst it was garbling with a Stansted inbound flight from the NW. She 'idented' the flight and stopped it off at altitude 4000ft; however, the crew informed her that they were already level at altitude 5000ft. She climbed the EMB190 flight to altitude 6000ft and then gave avoiding action to the A319 flight by descending it to altitude 4000ft followed swiftly by issuing an avoiding action turn to the EMB190 onto heading 290. She passed TI to the EMB190 crew who reported visual and then passed TI to the A319 crew. Prescribed separation was lost, 700ft/2 6nm. 1

21 THE LTC N COORDINATOR reports the sector loading was moderate. Ironically, prior to this incident, he had discussed the technique of stopping a Luton CLN or DVR SID flight from RW08 at 4000ft to give flexibility against the climb of London/City BPK departures. He had released the EMB190 flight from London/City on a BPK SID and subsequently released the A319 flight on a CLN SID from Luton. His general mode of operating was to ask the Sector Controller (SC) for each release, which he believed he did but could not be sure. He had coordinated the EMB190 with LTC NW at 5000ft and written this on NW s fps. At this time he had not written this coordination on the NE SC s fps. Essex Radar telephoned him on the NW Coordinator line to coordinate a Luton inbound that was going through the gate but high over the line. He pointed this out to the NW SC (student and OJTI) and the mentor asked him to organise a higher acceptance level into their sector. He subsequently coordinated 6000ft and annotated this both on the NW and NE fpss. It was at this time that the Essex SC used the priority line to point out the potential confliction between the EMB190 and the A319. He believed the NE SC had just noticed the confliction and was already taking steps, including avoiding action, to resolve it. ATSI reports that the Airprox was reported by the crew of an A319 following avoiding action being given by LTC NE against an EMB190 at 5000ft in Class A airspace, 8 5nm SE of Luton Airport. The A319 was operating IFR on a flight from Luton to Germany on a BPK5U departure and was in receipt of a RCS from LTC NE on frequency MHz. The EMB190 was on an IFR flight from London/City to Scotland on a CLN7C departure and was in receipt of a RCS from LTC NE on frequency MHz. Traffic departing from London/City RW09 on a BPK5U climbs to altitude 3000ft and turns L to intercept the 150 radial towards BPK. Traffic departing from Luton RW08 on a CLN7C climbs to altitude 5000ft and turns R at LUT NDB to intercept the 337 radial towards BPK. Figure 1 shows the interaction between the two SIDs. Figure 1. The LTC NE sector comprises NE (Deps), LOREL and LAM sectors combined as one position. 2

22 CAA ATSI had access to written reports from both pilots, written reports from the LTC NE controller and N Coordinator together with area radar recordings and RT recordings of the LTC NE frequency. The Luton METARs are provided for 0720 and 0750 UTC: EGGW Z 02012KT 360V NSC 01/ = and EGGW Z 02010KT 9000 NSC 01/01 Q1030= Prior to the Airprox the LTC NE controller had a London/City BPK departure against a Luton CLN departure. The controller and the N Coordinator had a conversation about stopping the CLN departure at 4000ft. The LTC NE controller did not consider it necessary to stop the CLN departure at 4000ft and the situation was easily resolved without incident. The LTC NE controller and the N Coordinator then had a conversation about the general technique of using intermediate levels to give flexibility to traffic climbing on the BPK departure from London/City. At 0714:40 London/City Tower telephoned the N Coordinator to request a release on the EMB190 on a BPK departure (climbing to 3000ft); the EMB190 was released. At 0717:00 the pilot of another ac that was not directly involved in the Airprox contacted the LTC NE controller. The pilot was given a heading of 015 which was read back correctly. Some moments later this ac was observed to be heading 105 and the LTC NE controller corrected the situation. Later the LTC NE controller recalled wondering if an incorrect read back had been missed but did not feel unduly distracted by the issue. At 0718:00 Luton Tower telephoned the N Coordinator to request a release on the A319 on a CLN departure (climbing to 5000ft). The N Coordinator asked Luton Tower to standby and there was a short pause before the A319 was released. The Coordinator s normal practice was to check with the controller before issuing a release; however, he could not specifically recall doing so in this case. The LTC NE controller also could not recall if a release had been issued. Later the LTC NE controller reported not having fully assimilated the presence of the A319 into the overall situation. At 0718:50 the crew of the EMB190 contacted the LTC NE controller climbing to 3000ft. The controller climbed the EMB190 to 5000ft as the anticipated level for coordination. The Coordinator agreed 5000ft for the EMB190 but did not write the level on the fps. The LTC NW controller requested that the EMB190 be co-ordinated at 6000ft due to a Luton inbound that was high through the Luton gate. The new agreed level was annotated on the fps by the Coordinator. At 0723:31 the A319 flight contacted the LTC NE controller, London (A319 c/s) passing four thousand six hundred feet climbing altitude five thousand feet Clacton seven Charlie. The radar display label for the A319 was garbling with another ac inbound to Stansted. The EMB190 was 8 3nm SE of the A319, tracking NNW. The LTC NE controller replied, (A319 c/s) squawk ident maintain alt-stop climb altitude four thousand feet. The crew of the A319 replied, Okay ident and we re just levelling altitude five thousand feet (A319 c/s). The LTC NE controller then instructed the crew of the EMB190 to, climb now altitude six thousand feet expedite which was read back correctly (0723:51). The 2 ac were on conflicting headings 5 5nm apart. [UKAB Note (1): Minimum vertical separation occurs at 0723:59, both ac at altitude 5000ft at a range of 4 3nm.] The A319 crew was then instructed, (A319 c/s) avoiding action descend immediately altitude four thousand feet. During this transmission low level STCA activated, followed by high level STCA 3sec later. The A319 crew read back (0724:00), Descend immediately altitude four thousand feet (A319 c/s). 3

23 The LTC NE controller gave avoiding action to the crew of the EMB190, instructing them to turn L immediately heading 280 and advised they had traffic in their 12 o clock. The pilot of the EMB190 replied, We re visual two eight zero degrees (EMB190 c/s). At 0724:11 the 2 ac were 2 6nm apart with the EMB190 at 5300ft and the A319 at 4600ft. TI was passed to the crew of the A319 on the EMB190 and the pilot of the A319 replied that they were visual. At 0724:15 separation had been restored the A319 was at 4500ft and the EMB190 was at 5500ft. [UKAB Note (2): The CPA occurs at 0724:27 as the EMB190, climbing through altitude 5900ft, passes 1 1nm SW of the A319, which is descending through altitude 4200ft.] Later the LTC NE controller stated that on the day, the traffic was steady but did not require splitting. The controller who relieved the LTC NE controller agreed that the position did not require splitting as did the Coordinator; however, the initial watch investigation indicated that a split may have helped the RT loading. Both the controller and the Coordinator stated that the area around BPK can be complicated when on Easterlies, partly due to the proximity of the Stansted RMA. When the LTC NE controller gave avoiding action the emphasis was on separating the ac vertically to allow time to assess the traffic in the vicinity of the Stansted RMA and ensure that any headings issued did not put the traffic in further confliction with traffic being vectored for Stansted. The LTC NE controller stated that, in hindsight, stopping the A319 at 4000ft would have prevented the confliction. Neither the LTC NE controller nor the Coordinator could recall if a release on the A319 had been approved by the LTC NE controller. As the release was requested just after the LTC NE controller had been required to take action on another ac that had turned onto the wrong heading, it is possible that the distraction prevented the LTC NE controller from incorporating the A319 into the overall plan. When the A319 became airborne from Luton the radar label was garbling with other traffic inbound to Stansted and the potential confliction against the EMB190 was not immediately apparent on the display. Although the crew of the A319 reported passing 4600ft the controller did not seem to absorb this information and the A319 was initially instructed to stop climb at 4000ft. When the controller became aware of the confliction appropriate action was taken to resolve the situation. The Airprox occurred when an A319 was allowed to depart unrestricted on a SID that climbed to the same level as that already occupied by an EMB190 on a conflicting track. It is unclear whether or not the LTC NE controller and the Coordinator discussed the release on the A319 prior to the Coordinator issuing a release to Luton Tower; however, the LTC NE controller reported not assimilating the presence of the A319 into the overall traffic plan (possibly due to the distraction caused by an earlier situation). The confliction between the 2 ac did not become apparent to the LTC NE controller until the A319 was airborne and levelling at 5000ft and in conflict with the EMB190 which was also at 5000ft. 4

24 PART B: SUMMARY OF THE BOARD'S DISCUSSIONS Information available included reports from the pilots of both ac, transcripts of the relevant RT frequencies, radar video recordings, reports from the air traffic controllers involved and reports from the appropriate ATC authorities. A controller Member advised the Board that this incident had occurred in a busy piece of airspace where tactical coordination was imperative for the safe and expeditious movement of traffic departing Luton, Stansted, Heathrow and London/City. The normal SOP for BPK departures from London/City is for the NE controller to climb the ac above the SID level of 3000ft as soon as possible to 5000ft, the normal coordinated altitude with LTC NW. As this climb impacts on the CLN SID from Luton climbing to 5000ft, it was essential for the NE controller and N Coordinator to ensure that appropriate steps/measures were taken to deconflict the 2 flights; the norm is to restrict the Luton departure to 4000ft. However, the required coordination did not take place as the A319 was released from Luton into conflict with the EMB190 which had caused the Airprox. Neither the NE controller nor the N Coordinator could remember if the NE controller had approved the release; certainly the NE controller had not taken the A319 into her traffic plan. The confliction only became apparent to the NE controller when the A319 flight made initial contact on frequency, with separation at 8nm; however, she did not assimilate the crew s report passing 4600ft when she instructed the flight to stop at 4000ft. After the A319 crew reported levelling at 5000ft she cleared the EMB190 to climb expeditiously to 6000ft, the coordinated level agreed with LTC NW, before giving the A319 crew an avoiding action descent to 4000ft. This was all achieved before STCA activated, which changed quickly from low to a high severity alert as the A319 crew acknowledged the avoiding action. The A319 crew saw the EMB190 as they actioned the descent prior to a TCAS TA being received. The NE controller then gave the EMB190 crew an immediate L turn away and TI. The EMB190 crew reported visual with the A319 with separation now 700ft and 2 6nm before standard separation was restored seconds later. Although not ideal, the NE controller had quickly reacted to the deterioration situation when the potential confliction was noticed. She had issued instructions which were acted upon swiftly by both crews which had resulted in a minor loss of separation. These combined actions were enough to allow the Board to conclude that any risk of collision had been effectively removed. Assessing the safety barriers, although the loss of separation had occurred as a result of controller error, the controller then took action before the STCA activated to recover the situation; Board Members concluded, therefore, that effective ATC safety barriers remained. Both crews had visually acquired each other s ac whilst following avoiding action instructions. While visual sightings are providential rather than systemic in Class A airspace, both aircrews had SA from their TCAS with the prospect that the robust barrier of TCAS RAs would have been effective. Since robust aircrew barriers and ATC barriers remained, the Board assigned an ERC score of 50 to the Airprox. PART C: ASSESSMENT OF CAUSE AND RISK Cause: ATC released the A319 into conflict with the EMB190. Degree of Risk: C. ERC Score: 50. 5

25 AIRPROX REPORT No Date/Time: 14 Mar Z Position: 5114N 00048W (2 5nm SSW Farnborough - elev 238ft) Airspace: ATZ/LFIR (Class: G) Reporting Ac Type: PA28 C182 Reported Ac Operator: Civ Trg Civ Trg Alt/FL: 2000ft 1600ft QNH (1017hPa) QFE (1002hPa) Weather: VMC CLNC VMC CLBC Visibility: 30km >20km Reported Separation: Nil V/<0 5nm H Recorded Separation: 200ft V/0 2nm H 200ft V/1nm H Odiham 5nm MATZ PA :55 A22 58:11 A21 CPA 59:11 PA28 A19 C182 A21 58:35 A20 58:51 A19 58:51 A21 59:03 A21 Farnborough Elev 238ft 59:03 A :35 A21 NM ATZ LTMA 3500ft+ 58:11 A21 Radar derived Levels show altitudes as Axx LON QNH 1017mb C :55 A22 PART A: SUMMARY OF INFORMATION REPORTED TO UKAB THE PA28 PILOT reports flying a dual CPL training sortie from Bournemouth, with the student (HP) in the LH seat, VFR and in receipt of a BS from Farnborough LARS W on MHz, squawking 0430 with Mode C. The visibility was 30km in good VMC and the ac was coloured blue/white with anti-collision and nav lights switched on. They had just departed the Blackbushe cct and flown through the Farnborough ATZ, passing W of the O/H, as requested by ATC. Heading 180 at 100kt level at 2000ft QNH 1017hPa the student was finalising diversion planning as they overflew Farnham [4nm S Farnborough]. He, the instructor, saw another ac, a high-wing single-engine type, about 0 5nm away and called out traffic 11 o clock same level L to R turn L now. The student turned immediately and a few seconds later they passed behind a Cessna by <0 5nm at approximately the same level. The Cessna did not appear to change heading or height at any time. He assessed the risk as high. THE PA28 COMPANY S CFI comments that the reporting pilot is a very experienced commercial instructor. In this very busy area of airspace, despite a very good service offered by Farnborough Radar, all of their instructors and students are briefed to keep a very good lookout whilst operating in this area. THE C182 PILOT reports acting as Commander/Examiner with another pilot operating as P1S on an IMC rating revalidation flight and in receipt of a TS from Odiham Approach on 131 3MHz, squawking an assigned code with Mode C. The visibility was >20km flying 2000ft below cloud in VMC and the ac was coloured white/black/red with anti-collision beacon and strobe lights switched on. Odiham Approach was vectoring them for a LH radar cct for an ILS/DME approach to RW27. Whilst on base leg, TI was given on traffic to the N but nothing was seen. Further TI was passed when they were on a closing heading and still nothing was seen. When established on the LOC about 6DME heading 275 at 110kt and 1600ft QFE 1002hPa further TI was passed on traffic to the N of them crossing R to L at a similar level. He briefed the HP to be prepared for him to take control of the ac for any avoiding action. Whilst he scanned the area to the N he saw a low-wing, single-engine ac approximately 1nm away and 200ft below making a L turn to pass behind their ac. He informed Odiham Approach that they were visual with the other ac and that it was turning to avoid them. He was unaware which ATSU the conflicting ac was working or what coordination had taken place. He assessed the risk as low. 1

26 THE FARNBOROUGH LARS W CONTROLLER reports that during a busy period he was asked to call Odiham. Odiham asked about his 0430 squawk, the PA28, tracking S and he informed them that it was at 2000ft. Odiham mentioned that they had a 3650 squawk on an ILS. The PA28 flight was under a BS routeing to the W of Farnborough outside the Odiham MATZ to avoid Farnborough ILS traffic. The Odiham controller did not ask for any coordination regarding the PA28. THE ODIHAM PAR CONTROLLER reports a U/T Approach controller was being screened on RAD/DIR while a C182 was conducting ILS approaches to the duty RW under a TS in BLUE Wx conditions. The C182 was on base leg heading 360 when an ac tracking S through the Farnborough ATZ was observed and was called by the RAD U/T controller. The C182 was established on the LOC at 7nm with the conflicting traffic still tracking S at a similar level. This traffic was called 3 times by the RAD U/T controller at various ranges. He was listening on the RAD frequency awaiting a handover for an ILS monitor of the C182. The C182 pilot called visual with the other ac at approximately 0 5nm and said he would turn to avoid but then didn t as the other ac started to turn away first. Both ac were seen on the PAR display as being <0 25nm apart at a similar level. ATSI reports that the Airprox occurred at 1059:10 UTC, 2 5nm SSW of Farnborough on the boundary of the Farnborough ATZ, between a C182 and a PA28. The Farnborough ATZ comprises a circle radius 2 5nm, centred on the midpoint of RW06/24 and extending to a height of 2000ft above aerodrome level (elevation 238ft); except that part of the circle situated N of the M3 Motorway. The C182 was operating VFR on a local flight to and from Lasham and was conducting an ILS approach to Odiham RW27 as part of an IMC re-validation with an examiner. The C182 was in receipt of a TS from Odiham Approach on frequency 131 3MHz. The PA28 was operating VFR on a CPL training flight and was departing the visual cct at Blackbushe, returning to Bournemouth via Midhurst. The PA28 was in receipt of a BS from Farnborough LARS(W) on frequency MHz. The UK AIP, page AD 2.EGLF-6 (13 Dec 2012), states: Pilots wishing permission to transit the Farnborough ATZ should in the first instance attempt to obtain that permission from Farnborough ATC on MHz during the notified operating hours of that frequency. At the time of the Airprox a Farnborough inbound was being provided with a Surveillance Radar Approach (SRA) to RW24 and in addition the Odiham RTC RW27 had been notified as being active. A Memorandum of Understanding (MoU) exists between Odiham and Farnborough. Paragraph 4.8, states: Odiham Radar Training Circuit. Odiham are to inform Farnborough when Odiham RTC is active and when activity ceases. In order to integrate with Farnborough arrivals and departures, Odiham Director s traffic in the Runway 27 radar training circuit will routinely operate at or below 1600ft QFE (2100ft Farnborough QNH) when east of a line North/South through Odiham. CAA ATSI had access to the RT recording of Farnborough LARS(W), the area radar recording, written reports from the 2 pilots, the Farnborough LARS(W) controller report together with the Farnborough unit report. The workload and traffic levels of the LARS(W) controller were assessed by CAA ATSI as medium-heavy. (Note: The CAA transcription unit reported that, due to a line fault believed to be at the Odiham end, the quality of incoming calls from Odiham was poor. An earlier call was reported as unreadable. Farnborough had earlier reported the fault to Odiham.) The Farnborough METAR is provided: 2

27 EGLF Z 36007KT 9999 SCT020 04/M02 Q1017= Odiham had previously notified Farnborough Approach that the Odiham RTC was active with an ac squawking This had been notified to the LARS(W) controller. The PA28 flight was departing the visual cct at Blackbushe and contacted Farnborough LARS(W) at 1055:32 reporting, (PA28 c/s) is a Piper Warrior from er Bournemouth to Bournemouth currently erm departing the circuit at Blackbushe heading towards Midhurst and we re on er altitude two thousand feet. The controller replied, (PA28 c/s) Q N H one zero one seven er Basic Service squawk zero four three zero. This was acknowledged correctly and the controller responded, (PA28 c/s) transit the Farnborough A T Z is approved to the west of the threshold please there s an S R A at six miles for runway two four. The PA28 pilot acknowledged, er transit to the west of the er overhead and er copied traffic (PA28 c/s). The PA28 turned to enter the Farnborough ATZ at an altitude of 2000ft and changed squawk at 1056:54 as it tracked S ly 2nm NNW of Farnborough. The C182 was 5nm SSE of Farnborough squawking 3650 tracking N at an altitude of 2100ft and on L base for the Odiham ILS RW27. For illustrative purposes the ATZ for each aerodrome has been overlaid below (figure 1). Figure 1 Area MRT radar at 1056:54 The Farnborough controller s workload increased. A call was received (1057:32) from Fairoaks regarding a VFR departure and during the conversation another flight contacted LARS(W). At the same time (1057:40) Farnborough Approach answered a call from Odiham Approach requesting TI on the 0430 squawk. The Farnborough Approach controller was conducting an SRA and advised Odiham that LARS(W) would call them back. At 1058:04, the C182 (2200ft) had turned onto a heading to intercept the Odiham RW27 LOC and the PA28 was passing 0 9nm SW of Farnborough (2200ft). The distance between the 2 ac was 3 4nm. At 1058:12, LARS(W) contacted Odiham and asked, "I hear you want to know about the zero four three zero" and Odiham replied, "Yes, he's continuing on that track is he"; LARS(W) responded, "Yeah, he is, he's a Warrior". Odiham acknowledged, "He's a Warrior", and LARS(W) responded "OK". Odiham stated, "Ours is just established on the ILS." LARS(W) again responded, OK. Odiham ended the call at 1058:34 reporting, Odiham Approach. The LARS(W) controller was then busy with other traffic and the RT was continuous until after the Airprox. 3

28 The 2 ac continued to converge and at 1058:48, the range between them was 1 3nm with a vertical separation of 100ft. At 1059:10 the distance between the ac had reduced to 0 2nm (CPA). Radar showed that the PA28 had turned L followed by a R turn to pass 0 2nm behind the C182 at 1059:14 (figure 2). Figure 2 - Area MRT radar recording at 1059:14 At 1110:52, the PA28 pilot reported changing to Solent on MHz and LARS(W) instructed the PA28 flight to squawk Later the Farnborough ATSU indicated that the LARS(W) controller would have notified or coordinated with Odiham had the PA28 routed through the Odiham MATZ. The Farnborough Manual of Air Traffic Services (MATS) Part 2, APR-33, paragraph (3), states: Coordination for aircraft wishing to transit the Odiham MATZ is required: a. If the aircraft is non-transponder equipped and identified to Farnborough. b. Whenever the Odiham instrument pattern is active. c. When other Odiham traffic indicates a potential confliction. d. For any aircraft wishing to penetrate the Odiham ATZ. Farnborough ATSU reported that the LARS(W) frequency was extremely busy and there was IFR traffic inbound to Farnborough. It was therefore likely that the LARS(W) controller had prioritised protecting the Farnborough traffic and was coordinating with Approach at the time. The LARS(W) controller had assessed the risk to the PA28 under a BS as a lower priority and with more time, generic advice or a warning may have been passed. When the range between the 2 ac was 3nm, the Odiham controller asked LARS(W) if the PA28 was continuing on present heading. This was confirmed by the LARS(W) controller but no coordination was requested by Odiham. The C182 was being vectored by Odiham on a TS and was passed 3 warnings by the Odiham controller. The Farnborough/Odiham MoU, paragraph 3.4 states: Notwithstanding the definitions given within these documents and the limitation on the need for tactical coordination given within this LoA, timely, effective coordination will be the basis for the safe operation of ATC. 4

29 MATS Pt1, Section 1, Chapter 10, Page 3, Paragraph 5.2, states: Outside controlled airspace, controllers are individually responsible for deciding whether they need tactical coordination, and to initiate such requests as appropriate. Therefore, unless specified in MATS Part 2, controllers should not rely on other controllers to initiate tactical coordination. The LARS(W) controller was prioritising a busy workload and there was no requirement to monitor the PA28 on a BS. The LARS(W) controller had advised the PA28 to route W of Farnborough due to the inbound SRA traffic. The PA28 was in transit through the Farnborough ATZ and may not have been aware that the Odiham RTC was active. Nevertheless the PA28 pilot remained responsible for his own separation using see and avoid. CAP 774, Chapter 2, Page 1, Paragraph 1 states: A Basic Service is an ATS provided for the purpose of giving advice and information useful for the safe and efficient conduct of flights. This may include weather information, changes of serviceability of facilities, conditions at aerodromes, general airspace activity information, and any other information likely to affect safety. The avoidance of other traffic is solely the pilot s responsibility. Basic Service relies on the pilot avoiding other traffic, unaided by controllers/fisos. It is essential that a pilot receiving this service remains alert to the fact that, unlike a Traffic Service and a Deconfliction Service, the provider of a Basic Service is not required to monitor the flight. Paragraph 5: Pilots should not expect any form of traffic information from a controller/fiso, as there is no such obligation placed on the controller/fiso under a Basic Service outside an Aerodrome Traffic Zone (ATZ), and the pilot remains responsible for collision avoidance at all times. However, on initial contact the controller/fiso may provide traffic information in general terms to assist with the pilot s situational awareness. This will not normally be updated by the controller/fiso unless the situation has changed markedly, or the pilot requests an update. A controller with access to surveillance-derived information shall avoid the routine provision of traffic information on specific aircraft, and a pilot who considers that he requires such a regular flow of specific traffic information shall request a Traffic Service. However, if a controller/fiso considers that a definite risk of collision exists, a warning may be issued to the pilot. The PA28 and C182 flew into close proximity whilst operating in Class G airspace, where the pilot ultimately remains responsible for their own traffic avoidance. The C182, in receipt of a TS was passed 3 warnings by the Odiham controller. The LARS(W) controller s workload was high and there was no requirement to monitor, or pass a warning to the PA28 on a BS. The PA28 pilot sighted the C182 and resolved the conflict by turning to pass behind. BM SPA reports that this Airprox occurred 5 6nm E of Odiham at 1059:10 between a C182 and a PA28. The C182 was operating VFR, conducting radar vectored ILS approaches to Odiham in receipt of a TS from Odiham RAD. The PA28 was operating VFR on a CPL training sortie and was in receipt of a BS from Farnborough LARS W. All heights/altitudes quoted are based upon SSR Mode C from the radar replay unless otherwise stated. The Airprox was reported by the Odiham Talkdown controller. RAD was operating bandboxed with DIR and manned by a trainee, who was already endorsed in the DIR position, and an instructor. At the time that this report was completed, the unit had not provided an occurrence report from RAD; however, the unit did conduct an investigation that was based on interview with the RAD trainee and instructor. Analysis of the radar replay and tape transcript determined that the C182 was the only ac in receipt of an ATS from RAD throughout the incident sequence; consequently, BM SPA contends that taskload and complexity for RAD were low. 5

30 Prior to the start of the incident sequence, at 1036:27, Odiham RAD had contacted Farnborough APP to advise them that the C182 was joining the Radar Training Circuit (RTC) squawking for 2 approaches to Runway 2-7 which was acknowledged by Farnborough. This liaison call was in accordance with the MoU between Farnborough and Odiham which states that Odiham are to inform Farnborough when the Odiham RTC is active. The incident sequence commenced at 1057:33 as RAD instructed the C182 flight to turn left heading three one zero degrees, report localiser established which was acknowledged by the C182 pilot. At this point, as depicted in Figure 1, the PA28 (SSR 3A 0430) was 5nm NW of the C182 (SSR 3A 3650). Figure 1: Incident geometry at 1057:33 Immediately after the C182 pilot acknowledged the turn, at 1057:44 Odiham RAD rang the Farnborough LARS landline, which was answered by Farnborough APP. Odiham RAD requested, traffic information, overhead Farnborough squawking zero four three zero and Farnborough APP advised that they would, get LARS to call you. The landline call was then terminated. The Farnborough-Odiham MoU specifies that: a. SSR 3A code 0430 is utilised by Farnborough LARS W for transit traffic. b. There are 2 direct landlines between Odiham and Farnborough, each with a distinct function; one is to be utilised for coordination and information requests in respect of Farnborough APP traffic and the other is to be used for coordination and information requests in respect of Farnborough LARS W traffic. Immediately after terminating the landline call to Farnborough APP, at 1057:55 RAD provided TI to the C182 flight on the PA28, advising, (C182 c/s) as you turn, traffic North, three miles, tracking south, similar height. Figure 2 depicts the incident geometry at this point, placing the PA28 3 8nm NW of the C182. The C182 pilot acknowledged the TI stating that they were good V-M-C, looking, nothing seen. At 1058:10, RAD provided updated TI on the PA28 to the C182 pilot, advising that the previously reported traffic now North, 2 miles, tracking South, similar height, which was acknowledged. The position report element of the TI was again inaccurate, in that the PA28 was 3nm NW of the C182. 6

31 Figure 2: Incident Geometry at 1057:55. CAP774 Chapter 3 Para 5 Guidance Material states that: Controllers shall aim to pass information on relevant traffic before the conflicting aircraft is within 5NM, in order to give the pilot sufficient time to meet his collision avoidance responsibilities and to allow for an update in traffic information if considered necessary. Immediately after receiving the C182 pilot s acknowledgement of the updated TI, at 1058:20 Odiham RAD answered a landline call from Farnborough LARS W who initiated the liaison by asking RAD, you wanted to know about the zero four three zero?. RAD replied, Yeah, is he just continuing on that track is he? LARS W stated, Yeah he is, he s a Warrior. Radar followed-up, He s a Warrior? and LARS W asked Okay? Radar then added, Oh, he s (the C182) just established on the I-L-S and LARS W replied, Okay. Radar then terminated the landline call at 1058:28. At 1058:33, RAD provided a further update of the TI to the C182 flight, describing the PA28 as, now right one o clock, one and a half miles, tracking South, similar height, which was acknowledged; Figure 3 depicts the incident geometry at this point, placing the PA28 2nm NW of the C182. Figure 3: Incident Geometry at 1058:33. RAD provided a further update to the TI at 1058:50, advising the C182 pilot that the PA28 was, now right one o clock, half a mile (radar replay shows 1 1nm), crossing right to left, similar height. Immediately after this transmission ended, the C182 s pilot advised Radar at 1059:00, (C182 c/s) in fact, yep, got him now and err I ll avoid, which was acknowledged by RAD. At this point, the PA28 was 0 7nm NW of the C182, tracking SSE ly, indicating 2000ft; the C182 was tracking W ly established on the LOC, indicating 2100ft. Figure 4 depicts the incident geometry at this point. At 1059:04, the C182 pilot advised Radar that the PA28 was, just passing us to the right, he s going 7

32 behind, well clear. The radar replay shows that at this point the PA28 was 0 3nm NW of the C182, maintaining a SSE ly track indicating 1900ft. The PA28 can then be observed making a hard turn to port at 1059:10, passing 0 2nm NE of the C182 and indicating 200ft below. This point represents the CPA, 5 6nm E of Odiham. Figure 4: Incident Geometry at 1059:00. Based upon the PA28 pilot s report, they sighted the C182 immediately prior to initiating the L turn observed on the radar replay at 1059:10, with the radar separation according with the pilot s assessment of less than 0 5nm. The unit s investigation determined that, given that the C182 was within the RTC and that the trainee RAD already held a DIR endorsement, the instructor was not monitoring the trainee controller throughout the incident sequence. The statement by the C182 pilot at 1059:04 describing the PA28 s avoiding action indicates that there is a timing discrepancy of 4-6sec between the radar replay and the Odiham RT transcript. However, it has not been possible to determine this conclusively and, given the minor nature of the discrepancy, it has no bearing on the determination of cause. Moreover, given the relative speeds involved, it makes little difference to the separation that existed at the point that the C182 pilot visually acquired the PA28. The incident occurred following RAD s instruction to the C182 pilot to turn onto the LOC at 1057:33 into conflict with the PA28; it was caused by an error in RAD s scan of the surveillance display, which did not detect the PA28 prior to issuing the turn. Whilst RAD then provided TI on a number of occasions during the remainder of the incident sequence, the original error was compounded by the inaccurate TI passed by RAD to the C182 pilot at 1057:55 and 1058:10. Whilst BM SPA acknowledges that the trainee held a DIR operating endorsement, a further contributory factor to the incident was the instructor s error in judgement in permitting the trainee to operate unsupervised. The unit has raised concerns over the passage of the PA28 through the Odiham RTC, at cct height, whilst in receipt of an ATS by Farnborough LARS W. BM SPA has some sympathy with Odiham s concerns; however, the PA28 was in receipt of a BS and the onus was on Odiham RAD to effect coordination with Farnborough, if required, and to ensure that they did not knowingly introduce a risk of collision. That said, given the level of cooperation required by the proximity of the 2 units and that Farnborough APP were advised of the activation of the Odiham RTC, it would be a reasonable expectation for LARS W to have passed some form of warning to the PA28 s pilot; however, this information was not available to BM SPA and will be determined by ATSI. OUTCOMES 8

33 ATM STANEVAL conducted an Intervention Workshop for the members of the Odiham Training, Examination and Standards teams, to provide guidance on how to recognise the requirement to intervene and how to carry out that intervention. BM SPA, in conjunction with the ATM and ASACS Force Commands, will issue guidance on the requirement for controllers tasked with training or examining controllers to maintain vigilance, irrespective of the extant operating endorsements held by the trainee/examinee. The unit s investigation has made a number of recommendations relating to the supervision of controlling staff and the content of the unit training package. PART B: SUMMARY OF THE BOARD'S DISCUSSIONS Information available included reports from the pilots of both ac, transcripts of the relevant RT frequencies, radar video recordings, reports from the air traffic controllers involved and reports from the appropriate ATC authorities. Members noted that there was ample opportunity for either ATSU to have taken action to have prevented this Airprox. Well over 20min prior to the CPA, Odiham RAD had contacted Farnborough APP informing them that the Odiham RTC to RW27 was active with the C182 and passing the ac s squawk code. It was unclear if this information was passed on to the LARS(W) controller but Members thought that it could have been issued to the PA28 pilot as generic TI owing to the subject ac s conflicting flightpaths; there is no onus on the controller to pass specific TI to pilots when under a BS. It was just after the Odiham RAD controller had turned the C182 flight, under a TS, onto a closing heading to intercept the LOC that RAD noticed the PA28 and telephoned Farnborough. Whilst awaiting a call back from LARS(W), Odiham RAD passed inaccurate TI twice to the C182 pilot before LARS(W) telephoned. However, the opportunity was missed to effect coordination and establish a deconfliction plan as only TI was exchanged between the controllers before termination of the call. LARS(W) then became busy and did not monitor the PA28 s progress whilst Odiham RAD did not to take action to break the confliction, instead electing to pass further TI to the C182 pilot. After receiving this updated TI, the C182 pilot saw the PA28 at about 1nm and stated that he would avoid. Meanwhile, the PA28 instructor saw the C182 in their 11 o clock and, although having right of way, told his student to turn L to pass behind. This sighting appears to be simultaneously with that by the C182 pilot who, on seeing the PA28 s turn, decided to continue on the ILS. These visual sightings and actions taken had ensured that any risk of collision had been removed. However the Board unanimously agreed that the Airprox was the result of the Odiham RAD vectoring the C182 into conflict with the PA28. In assessing the effectiveness of the safety barriers remaining, the Board agreed that the ATC barriers had not been completely effective. Although the PA28 pilot was under a BS which worked as required and the C182 pilot s SA was good from the increasingly accurate TI given by RAD that eventually cued the pilot s acquisition of the PA28, RAD vectored the C182 into conflict with the PA28. However, both pilots saw each other s ac at about the same time and the PA28 pilot took positive action to remove any risk of collision. As the incident occurred within Class G airspace where pilots are responsible for maintaining their own separation through see and avoid, the pilots sightings and actions were effective safety barriers and the Board assigned an ERC score of 2. PART C: ASSESSMENT OF CAUSE AND RISK Cause: Odiham RAD vectored the C182 into conflict with the PA28. Degree of Risk: C. ERC Score: 2. 9

34 AIRPROX REPORT No Date/Time: 18 Mar Z Position: 5315N 00432W (RAF Valley MATZ elev 36ft) Airspace: Valley MATZ (Class: G) Reporting Ac Reported Ac Type: Hawk T Mk2 Hawk T Mk1 Operator: HQ Air (Trg) HQ Air (Trg) Alt/FL: 1000ft NR QFE NR Weather: VMC CLBC VMC CLBC Visibility: 10km NR Reported Separation: 400m Recorded Separation: NK <0.5nm PART A: SUMMARY OF INFORMATION REPORTED TO UKAB THE HAWK T MK2 PILOT reports recovering to RAF Valley RW13 with a 'radar to visual' approach for a visual run-in and break (VRIAB). He was the student and PF, occupying the front seat and operating under VFR in communication with Valley ATC [Valley TWR, MHz]. The black ac had external lights selected on, as was the SSR transponder with modes A, C and S selected. The ac was fitted with TCAS II. He had been flying towards the A/D from the direction of Holyhead mountain [8nm NW of the A/D] with a 3-ship formation of Hawk T1 ac also positioning for a VRIAB, approaching from the W over the bay. He descended below cloud approximately 5nm from the RWY extended C/L, with a TS from Valley DIR, and when visual with the A/D switched to TWR. No TI had been called whilst on DIR frequency. Before he had an opportunity to transmit, due to busy RT, the TCAS gave a TA. The rear seat instructor saw the Hawk T1 formation at an estimated range of 1.5nm and directed the student pilot to 'look right' for the traffic. When no action was taken the instructor took control and manoeuvred the ac to the R to deconflict. He assessed the risk of collision as Low. THE HAWK T MK1 PILOT reports recovering to RAF Valley, leading a 3-ship formation of Hawk T1 ac. He was the instructor, occupying the rear seat, with the student, who was PF, in the front. The black ac was operating under VFR in VMC on TWR frequency with external light selected on. The SSR transponder was also selected on with Modes A and C. The ac was not fitted with an ACAS. He had initially contacted Valley APP to begin a descent to recover to RW13 with a TS. When in sight of the A/D he opted for a visual recovery. TI was called at 3nm and at 2.5nm on traffic conducting a straight-in PFL to RW13, he thought. The formation switched to Valley TWR to request join when approximately 3nm S of the A/D on the reciprocal runway heading. He briefly discussed in-cockpit the requirement not to turn in to initials until sure that the PFL traffic was clear. Approaching initials from the S heading 030, perpendicular to the RWY, none of the formation members were visual with the straight-in PFL traffic. TWR then reported traffic in the formation's R 2 o'clock, converging. No traffic was seen until the formation number 2 called 'Look Left'. He saw traffic in the L 10 o clock at a range of approximately ½nm converging and the rear-seat Captain took control. As the converging aircraft was seen breaking R away from the formation, the lead and number 3 remained straight and level and the formation number 2 broke upwards. 1

35 He assessed the risk of collision as High. [UKAB Note(1): The RAF Valley weather was reported as follows: METAR EGOV Z 04010KT 9999 FEW015 BKN070 06/02 Q0992 BLU NOSIG] THE RAF VALLEY DIR reports that he vectored the Hawk T2 pilot for a radar-to-initial (RtoI) recovery for RW13. The pilot was given a standard RtoI, to be sequenced behind another Hawk conducting a straight-in PFL (SIPFL). The Hawk T2 pilot was turned on to a heading of 110 at 8½nm and the controller informed TWR of the joining traffic. The Hawk T2 pilot reported visual with the A/D and switched to TWR frequency. He assessed the severity of the occurrence as Low. THE RAF VALLEY APP reports that he was screening on APP and controlling the Hawk T1 formation that recovered from the E of Valley for a radar-to-visual approach. The formation was cleared to descend to altitude 4000ft initially, to stay above the climb out lane. As the formation descended through about 5000ft, the leader changed intentions to a visual recovery; he was given own navigation and further descent. When about 4nm SE of Valley, he reported visual and switching to TWR, at which point there were no conflicting tracks. He assessed the severity of the occurrence as Low. THE RAF VALLEY ADC reports instructing a U/T controller during what had been a quiet session. At about 1545 a formation of 3 Hawk T1 ac called to join and was given a standard join instruction for RW13. At this point the visual cct was clear with instrument traffic, a straight-in PFL (SIPFL) at 4nm to touch and go and depart. After the SIPFL ac had departed, a formation was cleared for takeoff with the joining formation still outside initials. Using the Hi-Brite equipment, he observed a track on the deadside of RW13 [S of the A/D], about 9nm out indicating 1200ft, which he believed to be the Hawk T1 formation, and another track about 4nm W tracking NW and also indicating 1200ft. He pointed the W ly track out to the U/T controller although at that point it was not a factor for the visual cct. The W ly track then turned R 90 towards initials, as the track believed to be the Hawk T1 formation was about 3nm from initials, on the deadside. TI was passed immediately to the Hawk T1 formation, "[Formation C/S], traffic believed to be you has traffic right 2 o'clock, 2 miles crossing right-left", he thought. The number 2 of the Hawk T1 formation then called, "traffic left climb immediately, [Formation C/S] 2 is out yo-yo", he thought. He stated that the formation became visual from the VCR at this point, with the number 2 ac climbing vertically and the other 2 ac turning away to the L. The unknown track appeared to turn away to the R. He assessed the severity of the occurrence as High. THE RAF VALLEY SUP reports that he was in position in the ACR. An experienced controller was manning DIR with trainees in position on APP, TWR and PAR. He was monitoring the TWR and PAR frequencies and was aware that TWR was beginning to get busy. He was not directly monitoring DIR frequency but did hear DIR tell the Hawk T2 pilot to continue with TWR when at about 7nm final to RW13. He expected to hear the Hawk T2 pilot call TWR and be informed of the other joining ac. He heard TWR call traffic to the Hawk T1 formation but did not hear the Hawk T2 pilot check-in. At this point, the Hawk T1 formation called visual with the Hawk T2 and took avoiding action. [UKAB Note(2): The R/T transcriptions are reproduced as follows: RAF Valley DIR: 2

36 Time From To Speech Transcription Remarks 15:46:13 Hawk T2 DIR [Hawk T2 C/S] heading 250 level 6000 ft. 15:46:17 DIR Hawk T2 [Hawk T2 C/S] roger. 15:46:19 DIR Hawk T2 [Hawk T2 C/S] descend to height 3000 ft. 15:46:20 Hawk T2 DIR Descend to height 3000 ft, [Hawk T2 C/S]. 15:46:22 Other ac1 DIR [Other ac1 C/S] level 2000 ft. 15:46:25 DIR Other ac1 [Other ac1 C/S] roger. 15:46:26 PAR DIR Talk-down free, contact valley northwest 6 miles squawking :46:32 DIR PAR Unintelligible on the procedure, Traffic Service, Straight-in PFL. 15:46:34 PAR DIR Identified Stud 7. 15:46:36 DIR Other ac2 [Other ac2 C/S] contact talk-down Stud 7. 15:46:38 Other ac2 DIR Stud 7 [Other ac2 C/S]. 15:46:47 DIR Hawk T2 [Hawk T2 C/S] descend to height 2000 ft. 15:46:51 Hawk T2 DIR Descend to height 2000 ft, [Hawk T2 C/S]. 15:46:58 DIR Other ac1 [Other ac1 C/S] turn right heading 010 degrees. 15:47:01 Other ac1 DIR Right 010 degrees, [Other ac1 C/S]. 15:47:12 DIR Hawk T2 [Hawk T2 C/S] turn right 040 degrees. 15:47:15 Hawk T2 DIR Right 040 degrees [Hawk T2 C/S], descending to 2000 ft. 15:47:45 PAR ADC 3 miles, [Other ac2 C/S] touch and go. Radar Clearance 15:47:47 ADC PAR [Other ac2 C/S] cleared touch and go, Radar Clearance circuit clear, formation joining. 15:47:50 DIR Hawk T2 [Hawk T2 C/S] turn right heading 110 degrees. 15:47:52 Hawk T2 DIR Right heading 110 degrees, [Hawk T2 C/S]. 15:47:54 DIR Hawk T2 [Hawk T2 C/S] descend to height 1200 ft. 15:47:56 Hawk T2 DIR Descend 1200 ft. 15:48:02 DIR Other ac1 [Other ac1 C/S] turn right heading 040 degrees, cockpit checks report complete. 15:48:04 Other ac1 DIR Right 040 degrees, Wilco [Other ac1 C/S]. 15:48:31 DIR Hawk T2 [Hawk T2 C/S] aerodrome 12 o clock 7 miles report visual. 15:48:35 Hawk T2 DIR Wilco [Hawk T2 C/S]. 15:48:38 Unknown DIR Unintelligible (2 transmissions at once) field in sight to Tower. 15:48:59 DIR Hawk T2 [Hawk T2 C/S] roger. RAF Valley APP: Believed to be Hawk T2 From To Speech Transcription Time Remarks APP Form Ldr [Form C/S] confirm Radar to Initial 15:47:05 Form Ldr APP [Form C/S] negative, happy to visual to Tower 15:47:07 3

37 From To Speech Transcription Time Remarks APP Form Ldr [Form C/S] own navigation, taking your 15:47:13 own terrain clearance, descent approved. Form Ldr APP Descent approved [Form C/S], [Form 15:47:16 C/S] Stud 2 go. APP Form Ldr [Form C/S] 15:47:19 RAF Valley ADC: From To Speech Transcription Time Remarks PAR ADC 6 Miles, [Other ac2 C/S] Straight-in PFL 15:46:59 Radar Clearance touch and go further. ADC PAR [Other ac2 C/S] touch and go further. 15:47:01 Radar Clearance ADC All Hawk, Straight-in PFL turning inbound 15:74:02 touch and go further. Form Ldr ADC [Form check-in], Valley Tower [Form 15:47:25 C/S] request join. ADC Form Ldr [Form C/S], Valley Tower, join RW 13 15:47:30 QFE 990 hectopascals circuit clear, instrument traffic Straight-in PFL 3 miles. Form Ldr ADC 13, 990, copy the instrument traffic, 15:47:40 [Form C/S]. PAR ADC 3 miles, [Other ac2 C/S] touch and go. 15:47:44 Radar Clearance ADC PAR [Other ac2 C/S] cleared touch and go, 15:47:46 Radar Clearance circuit clear, formation joining. ADC All Hawk 2 and half miles touch and go. 15:47:58 Helo ADC Valley Tower good afternoon, [Helo 15:48:39 C/S] ready for vertical departure, request cross 01 undershoot. ADC Helo [Helo C/S] Valley Tower, clear vertical 15:48:44 take-off surface wind , cross 01 undershoot. Helo ADC Clear vertical take-off, wind copied, 15:48:49 cross 01 undershoot, [Helo C/S]. Other ADC [Other Form check-in], Valley Tower 15:48:56 Form [Other Form C/S] ready for departure. ADC Other [Other Form C/S] Valley Tower, cleared 15:49:01 Form for take-off, surface wind Other ADC Cleared take-off, [Other Form C/S]. 15:49:06 Form ADC Form Ldr [Form C/S], unintelligible (2 15:49:11 transmissions at once) traffic believed to be you has traffic right 1 o clock 1 mile crossing right to left similar height. Unknown ADC Unintelligible, copied. 15:49:20 Form Ldr All [Form C/S] you tally left. 15:49:21 Form Ldr All Unintelligible, climb. 15:49:23 Form Ldr ADC Tower, [Form C/S] we ve just, uh been 15:49:28 close to unintelligible that traffic, it s now broken out. Form No2 All [Form C/S No 2] out, yo-yo. 15:49:35 Form Ldr Form No2 Copied. 15:49:38 SUP ADC Sup. 15:49:39 Intercom ADC SUP Come to tower immediately. 15:49:39 Intercom SUP ADC On my way. 15:49:40 Intercom 4

38 ] From To Speech Transcription Time Remarks Form Ldr ADC [Form C/S] 1 Flt, plus 1 and 3, now 15:49:41 positioning behind the Holyhead. ADC Form Ldr [Form C/S] 1 Flt. 15:49:46 Hawk T2 ADC Tower, [Hawk T2 C/S] with you, we ve come right, were clear of um [Form C/S]. 15:49:48 BM SAFETY POLICY AND ASSURANCE reports that this Airprox occurred at approximately 1549:23 on 18 Mar 13 between a formation of 3 Hawk T1s (Hawk T1 Formation) conducting a visual join and a Hawk T2 conducting a RtoI join; both elements were recovering to RW13 at RAF Valley. All heights/altitudes quoted are based upon SSR Mode C from the radar replay unless otherwise stated. Unfortunately, given the height and distance from the NATS radar heads of the occurrence, the Airprox was not recorded on radar; the Hawk T2 and Hawk T1 Formation dropped outside coverage at 1547:32 and 1548:12 respectively. Analysis DIR was manned by an experienced controller who described his workload as medium to low and reported that the task was not complex. In addition to the Hawk T2, he was providing an ATS to 2 additional Hawk T2s in the RTC; one conducting a SIPFL ahead of the incident Hawk T2 and one being vectored for a PAR behind the incident Hawk T2. APP was manned by a trainee and an instructor who described their workload as high to medium with moderate task complexity, albeit that the Hawk T1 Formation were the only ac to which APP were providing an ATS. ADC was manned by a trainee and an instructor, who described their workload as medium to low, with moderate task complexity. In addition to the Hawk T1 Formation, the ADC was providing an ATS to a departing Griffin helicopter and sequencing the departure of a formation of Hawk T2s against the Hawk T2 conducting a SIPFL. The incident sequence commenced at 1547:16 as the Hawk T1 Formation switched from APP frequency to TWR frequency. Subsequent to completing their DASOR, the instructor pilot leading the formation stated that his student, as PF, had effected the frequency change earlier than would be considered common. At this point, the Hawk T1 Formation was 4.5nm ESE of Valley, tracking WSW ly, indicating descent through 4600ft; Hawk T2 was 11.8nm WNW of Valley, tracking NNW ly, indicating descent through 3300ft. Figure 1 depicts the positions of the respective ac at this point; SSR 3A 3731 is the Hawk T2, SSR 3A 3732 is the Hawk T1 Formation. Figure 1: Positions of Hawk T1 Formation and Hawk T2 at 1547:16. 5

39 The Hawk T1 formation leader had intended to recover through a RtoI join, which would have required APP to hand the formation to DIR; however, as reported by the formation, a change in the recovery state allowed them to fly a visual join and thus they remained on the APP frequency until visual with the A/D. The ATC Order Book (ATCOB) Pt 3 Order 2 (Orders for the APP Controller) states that Once VMC, pilots are to be instructed to contact Tower. There is no geographical guidance to restrict where the pilot should be instructed to contact TWR. However, subsequent to completing the DASOR, DIR suggested that it was not considered good practice to transfer ac conducting a visual join at the point that the Hawk T1 Formation was transferred. Moreover, the ADC instructor highlighted that the Flying Order Book (FOB) General Orders Part 4 Para 2 states that Tower is to be contacted when approx 3 nm from Initial. Anecdotally, this was introduced as a result of a previous Airprox in the vicinity of the Initial Point (IP) and was designed to ensure that aircrew remained with APP or DIR as long as possible to facilitate the provision of TI and deconfliction by ATC. In this instance, by not specifying geographical guidance for the point of transfer from radar to TWR, the ATCOB does not appear to support the intent behind the instruction within the FOB. Supported by an analysis of the R/T and landline transcript, the unit determined that no liaison occurred between APP and DIR, nor SUP and DIR to advise DIR of the intentions of the Hawk T1 formation. This lack of liaison played an important role in the development of DIR s SA. At 1547:25, the Hawk T1 formation leader called TWR to request join, which was approved; the airfield details were passed and they were advised, circuit clear, instrument traffic straight-in P-F-L 3 miles [unrelated to the incident, SSR3A 3756 in Figure 1] which was acknowledged. At this point, the Hawk T1 Formation were 4nm SE of Valley, tracking WSW ly, indicating 4600ft, maintaining the SSR3A code previously assigned by APP; the Hawk T2 was 12.1nm WNW of Valley, tracking NNW ly, indicating descent through 3000ft. Subsequent to completing his DASOR, DIR related that his perception of the Hawk T1 Formation s maintenance of the SSR3A code assigned by APP was that the formation would shortly be handed to him to conduct a RtoI approach. DIR s expectation being that, if the formation had been transferred to TWR, then they would have been squawking the visual cct SSR3A code. For DIR, this view was reinforced by the fact that no liaison had been conducted by either the SUP or APP to confirm the intentions of the Hawk T1 Formation leader; in effect, liaison would have meant that the ac would have switched to TWR, no liaison would mean that the ac would be handed over when APP was ready. The ATCOB Part 3 Order 2 states that an ac conducting a visual recovery is to maintain its Squawk until in the Visual Circuit, where it will squawk However, there is a nuanced difference to this rule within the FOB General Orders Part 4 Para 3, which states that ac joining the visual circuit to land are to maintain their recovery squawk into the circuit. Ac joining to practice multiple visual circuits are to change their squawk to 3737 at the first opportunity. The FOB does not define whether the first opportunity to change the squawk should be taken inside or outside the visual cct. Whilst DIR was under a mis-apprehension that ac already transferred to TWR would be squawking SSR3A 3737, this error in knowledge was neither causal nor contributory to the Airprox. The key element was that the Hawk Formation leader s maintenance of the SSR3A code assigned by APP, coupled with the formation s position outside the visual cct and tracking away, reinforced DIR s belief that the formation would be handed over by APP. At 1547:32, the Hawk T2 passed outside NATS radar coverage and then, at 1547:50 the pilot was instructed by DIR to, turn right heading degrees, towards the IP. At this point, the Hawk T1 Formation was 2.8nm S of Valley, tracking WSW ly, indicating descent through 3600ft; through extrapolation of the radar data, the Hawk T2 was approximately 11.8nm WNW of Valley. At 1547:55 the Hawk T1 Formation, 2.9nm SSW of Valley, adopted a WNW ly track, indicating descent through 3300ft. At 1548:12, 2.6nm SW of Valley, the Hawk T1 Formation passed outside NATS radar coverage, indicating descent through 2700ft. Given that both the Hawk T1 Formation and the Hawk T2 had passed outside NATS radar coverage, the developing air picture was re-created based on extrapolation of the available radar data, the ADCs DASOR and conversation with both the ADC instructor and the DIR. The ADC reported that using the Hi-Brite equipment he observed a track lined up on the deadside of RW13 at 6

40 approximately 9 miles, indicating 1200ft, and another track approx 4 miles W of Valley, tracking NW also indicating 1200ft. At 1548:31, DIR advised the Hawk T2 pilot, aerodrome 12 o clock, 7 miles, report visual which was immediately acknowledged by the pilot, who then advised at 1548:38, field in sight to tower, which was acknowledged by DIR. Extrapolation of the available radar data suggests that DIR advised the Hawk T2 pilot of the location of the aerodrome at approximately 8.5nm from Valley. Although DIR could not recall the SSR Mode C of the Hawk T1 Formation, he could recall the incident geometry at the point at which he acknowledged the Hawk T2 transferring to TWR frequency. DIR stated that at this point, the Hawk T1 Formation were approximately 4 to 4.5nm WSW of Valley, on a NW ly track that would have seen them pass behind the Hawk T2 and that he did not consider there to be a requirement to pass TI, in accordance with the guidance laid out in CAP774. This perception was echoed by the ADC who stated that, when he observed the geometry initially on the Hi-Brite display, the Hawk T1 Formation were on a track that would pass behind the Hawk T2. At 1549:11, the ADC attempted to warn the Hawk T1 Formation about the proximity of the Hawk T2, advising, traffic believed to be you has traffic right 1 o clock, 1 mile, crossing right to left, similar height. The ADC subsequently stated that this warning was precipitated by seeing the track that had been approximately 4 miles W of Valley, tracking NW turn R approximately 90 towards the IP. Moreover, DIR related that this turn occurred after the Hawk T2 pilot had left the frequency and that, whilst he observed this turn on radar, insufficient time remained for him to pass a warning to the ADC. Subsequent analysis has determined that the ADC passed a warning based on an incorrect perception that the Hawk T1 formation was the radar contact that had been observed lined up on the deadside of RW13 at approx 9nm ; this perception was based on a number of factors, most of which were outside the ADC s control. The ADC did not receive a warn-in of the Hawk T2 from DIR in accordance with local orders. These require DIR to make a 2 minutes call to TWR with the ac callsign on the Radar Clearance Line when the ac is 10 nm from the airfield; TWR will respond with the cct state or a hold off instruction. Whilst DIR did attempt to comply with the rule, subsequent to completing the DASOR he related that the timing of the call coincided exactly with an un-related call by the PAR controller gaining a final clearance for the ac conducting a SIPFL, ahead of the Hawk T2. Having heard the ADC s response to the PAR controller, which included the visual cct state, DIR was content that the liaison had been effected; however, the ADC had not heard DIR s call. Moreover, even if DIR s liaison had been effective, given the content of the liaison as mandated by the ATCOB, it would not have assisted the ADC in positively identifying one or other of the incident ac; it would only have alerted them to the fact that there was more than one speaking unit conducting a visual join. This effect was accentuated by the fact that local orders do not require ac conducting visual recoveries to be warned-in to TWR. Finally, DF is not available on the Hi-Brite display and thus the ADC was unable to correlate an ac s transmission with a surveillance radar return. Consequently, the ADC s mental picture was constructed from the fact that the only callsign on frequency that was known to be conducting a visual join was the Hawk T1 Formation. The information on the Hi-Brite display that best fitted this mental picture was that the ac approaching the IP on deadside was the Hawk T1 Formation. This perception was reinforced by the fact that the Hawk T2 had been unable to establish 2-way R/T with the ADC due to other, un-related, visual cct traffic. The Hawk T1 Formation leader reported that, following the warning from TWR, no traffic was seen until [Formation C/S] 2 called look left. The Hawk T2 was then visually acquired by the reporter at approximately ½nm and converging and action was taken to break the conflict. The leader of the Hawk T1 Formation reported that he had discussed in cockpit about not turning in to initials until sure that the [Straight In] PFL traffic was clear. Approaching initials from the S, perpendicular to the RWY, none of [Hawk T1 Formation C/S] were visual with the [Straight In] PFL traffic. Subsequent conversation with the formation leader confirmed that the formation had adopted a N ly track towards the IP and that the formation leader s lookout was focussed towards the A/D. The Hawk T2 instructor reported that he observed the Hawk T1 Formation at range [approx 1.5nm] and directed the student pilot to look right for the traffic. When no action was taken the instructor 7

41 took control and manoeuvred the ac to the R to deconflict. The CPA occurred at approximately 1549:23; due to R/T congestion, the Hawk T2 pilot was unable to call TWR to request a join until 1549:48. Turning first to the ADC s role in the Airprox, based upon the available information, his perception of the identity of the Hawk T1 Formation was understandable. Moreover, use of the Hi-Brite display to enhance SA was laudable, as was the attempt to pass a traffic warning to the Hawk T1 Formation; unfortunately, due to the faulty mental model of the situation, the warning probably served to slightly delay the Hawk T1 Formation s visual acquisition of the Hawk T2. From DIR s perspective, and based upon his and the ADC s description of the building geometry prior to the transfer of the Hawk T2 pilot to TWR, his decision not to pass TI to the Hawk T2 was understandable. That said, given the location of the Hawk T1 Formation relative to the Hawk T2, and whilst cognisant that APP was a trainee and that DIR, as an experienced controller, may have been giving APP the opportunity to act for themselves, with hindsight a better course of action would have been for DIR to challenge APP on the intentions of the Hawk T1 Formation. It would then have become apparent that APP was no longer working the formation and have prompted DIR to pass TI to the Hawk T2. Moreover, whilst the timing of DIR s warning to TWR was unfortunate, the lapse in determining that the liaison had been effective directly contributed to the ADC s reduced SA. Finally, from APP s perspective, whilst the guidance within the ATCOB regarding the timing of the transfer of ac conducting visual joins to TWR is at variance both with the FOB and with what DIR considered to be good practice, the decision to release the Hawk T1 Formation to TWR was, again, in the circumstances understandable. Whilst BM SPA is cognisant that Valley is a busy unit, with the consequent requirement to minimise R/T and landline liaison, a better solution may have been for APP to have contacted TWR to point-out the Hawk T1 Formation, warning that it had left APP s frequency early. This Airprox resulted from a sequence of unconnected events, associated with a number of systemic issues, which breached the ATM related safety barriers and caused a conflict in the vicinity of the IP. The procedures within the FOB for ac recovering through the IP rely on see and avoid and, in this instance, that final safety barrier did not operate until a relatively late stage in the incident sequence. The crux of the matter, from an ATM perspective, was the lack of liaison that occurred between individuals within the ACR and between the ACR and TWR. Recommendations BM SPA has recommended to Stn Cdr RAF Valley that he consider: a. Initiating a review of the FOB and ATCOB to ensure that the documents fully complement each other; specifically that the differences between ATCOB Pt 3 Order 2 and FOB General Orders Part 4 Paras 2 and 3 are addressed. b. Directing a review of the requirements within the ATCOB to warn-in traffic to TWR that are conducting Radar-to-Visual and Visual recoveries. The ADC related that the SSR Code allocation for RAF Valley used to be sufficient to enable their controllers to differentiate between ac in receipt of an ATS from each control position; however, a recent change mandating the provision of a surveillance based ATS to Valley-based ac has meant that more ac are in receipt of an ATS at any one time. This has had a 2 nd order effect in that the controllers have been required to pool the unit s SSR 3A code allocation and thus have lost the situational awareness afforded by control position specific SSR codes. BM SPA has also recommended to Stn Cdr RAF Valley that he consider initiating a review of SSR Code allocation procedures; specifically those used to differentiate IFR and VFR traffic and the timing of SSR code changes with regards to ac conducting visual recoveries. Observations BM SPA is aware that many ATM units utilise the Radar Clearance Line (RCL) for some routine liaison calls, in addition to obtaining radar clearances. In this instance, the simultaneous use of the 8

42 RCL by PAR and DIR contributed to the incident in that the ADC did not receive notification on the Hawk T2. MMATM Chapter 24 Para 33, supporting RA 3024 (1), (2) and (3), makes reference to the PAR controller utilising the RCL to obtain radar clearances but does not specify the operational uses of the RCL. BM SPA has recommended to the MAA that they consider specifying the operational uses of the RCL. HQ AIR (TRG) comments that this incident has been thoroughly investigated and steps have been taken to resolve some discrepancies. It highlights the potential for errors and mistakes that prevent any system of safety defences being 100% effective. However, it also emphasises the importance of the final but generally most significant safety barrier, namely lookout, and it is reassuring to note that both parties saw each other in sufficient time to react. The report is a demonstration of an excellent reporting culture and the investigation and lessons learnt exemplify the benefits that this culture brings. PART B: SUMMARY OF THE BOARD'S DISCUSSIONS Information available included reports from the pilots of both ac, transcripts of the relevant RT frequencies, radar video recordings, reports from the air traffic controllers involved and reports from the appropriate ATC authorities. The Board first considered the two significant factors involved in this incident, which a military pilot Member noted were; the Hawk T1 formation leader switching to TWR earlier than would normally be expected and the ADC mis-id of joining traffic and subsequent incorrect TI call. However, whilst the situation was not ideal, the see and avoid principle had functioned correctly, no doubt with the assistance of TCAS in the Hawk T2 cockpit. Turning to controller aspects of the incident, the incident had highlighted a lack of effective teamwork across controller positions. Civilian controllers opined that organisational aspects effectively resulted in the SUP being a single point of failure within the system and that this would have been mitigated by improved teamwork between ATC positions. The military controller Adviser agreed that lack of effective liaison had played an important role in this incident and further noted that the change to the Flying Order Book mandating a TS for Valley traffic had removed ATC position-allocated squawk codes and hence the barrier of other controllers being able to detect pilots intentions. A military ATC Member stated that training reviews had resulted in an increased emphasis on team-working within RAF ATCO training. He also noted that APP had used channel intercom to ADC, as opposed to a dedicated phone line, and that this could result in a lack of formality in the exchange of information with a consequent erosion in the quality of information exchanged. The CAA SRG Advisor noted that aspects of this Airprox correlated with civilian Airprox in the cct and that the overriding consideration was to pass TI as appropriate to traffic at the time. Although the ADC had used his Hi-Brite display, in the absence of adequate information from APP or DIR about the ac joining, his SA was incorrect and the TI he provided was misleading. Since neither Hawk T2 nor Hawk T1 flight received useful information from ATC to assist with their responsibility to see and avoid other traffic, the Board concluded that, overall, ATC barriers had been ineffective. Conversely, and notwithstanding the late sighting of Hawk T2 by the T1 formation, aircrew barriers had been effective with the notable contribution of the Hawk T2 TCAS. The Board considered that the Hawk T2 instructor had seen the Hawk T1 formation with time to allow his student the opportunity to react. When this did not occur he took control and took effective action to avoid a collision. PART C: ASSESSMENT OF CAUSE AND RISK Cause: In the absence of TI, a conflict resolved by the Hawk T2 instructor. Degree of Risk: C. ERC Score: 2. 9

43 AIRPROX REPORT No Date/Time: 29 Mar Z (Good Friday) Position: 5256N 00238W (Tilstock Parachuting Site elev 301ft) Airspace: Shawbury AIAA (Class: G) Reporting Ac Reported Ac Type: Parachutist Rockwell RC114B Operator: NK Civ Pte Alt/FL: 1400ft 1500ft NK QFE (1002hPa) Weather: VMC NK VMC NK Visibility: 10km > 10km Reported Separation: Recorded Separation: 400ft V/0.5nm H NK NK REPORT FILED BY GA8 PILOT ON BEHALF OF PARACHUTIST AND D/Z CONTROLLER PART A: SUMMARY OF INFORMATION REPORTED TO UKAB THE PARACHUTIST reports taking avoiding action against a light ac at 1400ft, near the O/H of Tilstock parachuting site. The Parachutist assessed the risk of collision as Medium. [UKAB Note(1): Tilstock Parachuting Site was promulgated in the UK AIP ENR 5.5-6, valid at the time, as a Parachute Jumping site of 1.5nm radius, up to FL85, and normally active during daylight hours, Monday to Saturday and PH, during winter and 1hr earlier in summer.] THE GA8 PILOT reports conducting parachute dropping at Tilstock parachuting site. He was operating under VFR, in VMC, with an A/G service from Tilstock Radio [ MHz]. The white ac had navigation and strobe lights selected on, as was the SSR transponder, with Modes A, C and S. The D/Z controller had observed the RC114B fly through the Tilstock parachuting site O/H on a number of occasions in the half-hour prior to the Airprox; the GA8 pilot was also occasionally visual with the ac. Repeated efforts were made by the D/Z controller to contact the ac but with no response. At about 1301, the ac was seen to fly N-S along the central portion of the drop zone, with parachute canopies in the air above it. The D/Z controller instructed all parachuting activity to be suspended and pilots to land. Two of the parachutists in the air saw the subject RC114B, one of whom took avoiding action at about 1400ft, estimating a minimum separation of 400ft V and 0.5nm H. THE RC114B PILOT reports operating autonomously, testing a new avionics installation. He was operating under VFR, in VMC, listening out on Shawbury LARS [ MHz]. The white and blue ac had strobe lights selected on, as was the SSR transponder with Modes A, C and S. The ac was fitted with a TAS. Whilst level at altitude 1500ft, tracking the Shawbury VOR radial at 130kt, he received TAS information and saw a white, single-engine, ac in his L 11 o clock at more than 1nm range. He turned away to the R.

44 He assessed the risk of collision as None. PART B: SUMMARY OF THE BOARD'S DISCUSSIONS Information available included reports from the pilots of both ac and radar video recordings. Members first considered the actions of the RC114B pilot. It appeared from additional information submitted that the locally-based pilot was not aware of the operating hours of Tilstock parachuting site and believed that parachuting activity would be promulgated by NOTAM. Notwithstanding this error, GA Members noted that the incident could easily have been avoided by the pilot choosing to use a more appropriate position and a surveillance based ATS when exploring the capabilities of his new avionics fit. Given that radar replay showed the RC114B pilot flying O/H the D/Z at altitudes between 1700ft and 2200ft on six occasions in the half hour preceding the Airprox and that the pilot did not report seeing parachutists, the Board also opined that he was most likely not maintaining an effective lookout and that this was due to being distracted by manipulating the new equipment in the cockpit. Board Members reemphasised that mitigation against mid-air collision in class G airspace is achieved by effective lookout and opined that the RC114B pilot s lack of appropriate planning and ineffective lookout placed all the airspace users at risk. Board Members made several observations: Sortie planning should include the addition of a safety pilot if it is anticipated that a significant amount of time will be spent heads-in. Parachute D/Zs often promulgate extensive hours of operation and it is anecdotally reported that activity often does not occur to the same extent, hence engendering a sense of false-alert. The Board felt that it behoves all pilots to remain clear, especially upwind, of parachute sites during promulgated operating hours unless they can positively confirm that the site is inactive, noting that the absence of a response on the RT does not confirm that the site is inactive. The Board also observed that a parachuting D/Z such as Tilstock, with no civilian regulated or controlled airspace associated with it, does not have priority over other entitled airspace users. As such, the D/Z controller s decision to suspend operations was commended by the Board. There were no ATC barriers to prevent this occurrence and, although the see and avoid principle functioned as the only remaining barrier to a limited degree, it was fortuitous that the parachutist, who had right of way, saw the RC114B at a range where his limited ability to take avoiding action was not tested. Overall, effective and timely action was taken to prevent a collision. PART C: ASSESSMENT OF CAUSE AND RISK Cause: The RC114B pilot flew through a promulgated and active parachuting site and into conflict with a parachutist, who he did not see. Degree of Risk: C. ERC: 4.

45 AIRPROX REPORT No Date/Time: 6 Apr Z (Saturday) Position: N E (9.5nm SSW of Earls Colne) Airspace: London FIR (Class: G) Reporting Ac Type: Robin SU29 (A120T) Reported Ac Operator: Civ Club Civ Pte Alt/FL: 1900ft NR (QNH 1024hPa) Weather: VMC NR NR Visibility: >10km 10km Reported Separation: 0ft V / <50m H Recorded Separation: <0.3nm H NR Diagram based on radar data and pilot reports NM : Boreham 1235: Robin LTMA : : : LTMA Earls Colne 35: :11 36:07 36:47 35:47 (manoeuvring intermittent returns) SU29 PART A: SUMMARY OF INFORMATION REPORTED TO UKAB THE ROBIN PILOT reports that she was a student flying a white aircraft with blue markings on the wingtips and the lower fuselage, squawking Mode 3/A 7000 with Modes C and S switched on. She was flying VFR on a solo navigation exercise routing from Earls Colne via River Crouch, Hanningfield Reservoir, Chelmsford and was returning to Earls Colne, heading 050 at 90kts, cruising at 1900ft, she thought, on the QNH of 1024hPa. She was not in receipt of any ATS. Having confirmed her location, she was about to contact Earls Colne when she saw a red and white ac less than 1500m away descending vertically in front of her ac in what appeared to be an aerobatic dive. The Robin pilot expressed her surprised that she had not spotted the SU29 earlier, particularly as the Robin had a glass canopy and surmised that the other ac may have come out of the cloud base above her. She estimated the CPA as less than 50m H with the SU29 passing through her altitude. The Robin pilot initiated a right turn but was aware that the other pilot may not have seen her and may come out of the dive in any direction so she stopped the turn after approximately degrees. Subsequently she saw the SU29 exit the dive to her L and below her ac before it entered a sweeping level curve behind her ac and headed off in the direction of Andrewsfield. The Robin pilot circled over Birch to re-establish her bearings and then set course back to Earls Colne. The student s instructor supported her report and added that weather was SCT CU ft and that he estimated the cloud base was 2200ft AMSL. THE SU29 PILOT reports flying his red and white ac VFR out of North Weald on a local aerobatics and general handling exercise at 130kts in VMC with strobe lights operating; he had selected a mode 3/A code of 7000 and had Mode C switched on, he thought. [UKAB Note (1): The ac does not display an SSR code at any time during the radar replay and the pilot has been contacted so that he can check the serviceability of his transponder.] The pilot recalls operating on QNH and moving his practice area every 3 SU29 (Photo courtesy of Sukhoi) 1

46 or 4 manoeuvres in order to minimise the noise impact for individuals on the ground but that he was operating in the general area E of Stansted and N of Chelmsford. During the sortie, the pilot recalls seeing 3 or 4 aircraft, none of which he considered close enough to present a collision risk and could not be certain if any of them were the Reporting ac. UKAB NOTE (2): The Stansted weather at 1220Z was: METAR EGSS Z 06005KT 360V SCT044 09/M03 Q1024= UKAB NOTE (3): ANALYSIS OF THE STANSTED RADAR at 1235:11 shows the Robin 2.3nm SE of Boreham Heliport, tracking NE, Squawking Mode 3/A 7000 with Mode C indicating an altitude of 1500ft. At the same time the SU29 is displayed as a primary-only return 1.6nm E of the Robin tracking NW. At 1235:24 the ac are 1.1nm apart, the Robin continues on its track indicating 1400ft whilst the SU29 appears to manoeuvre between N and W. Between 1235:28 and 1235:47 the SU29 return appears to manoeuvre in a small area 0.5nm ahead of the Robin, which continues to track NE indicating 1300ft. At 1235:47 the primary return from the SU29 is lost at the measured CPA 0.3nm ahead of the Robin; it is likely that the actual CPA is not shown and it is likely to have been closer. Shortly afterwards the Robin is seen to turn left to track N and its Mode C indicates 1100ft. At 1236:07 the SU29 s primary return reappears 0.8nm SW of the Robin and continues to manoeuvre in the area without approaching the Robin any closer. PART B: SUMMARY OF THE BOARD S DISCUSSIONS Information available consisted of the reports from the pilots of both ac, a report from the student pilots instructor and radar video recordings. The Board noted that this was an encounter in Class G airspace. A Member noted that the SU29 pilot had seen 3 or 4 ac during his sortie but did not consider them to be a collision risk and theorised that the relative perceptions of the inexperienced student pilot and the experienced aerobatic pilot may have contributed to the difference in their reports. The Board felt that although the student pilot may not have been confident in her assessment of the separation distances, the radar evidence indicated that the CPA was closer than would be considered normal during aerobatics and that it was therefore likely that the SU29 pilot had not seen the Robin. On the subject of the SU29 s transponder, the Board agreed that it was entirely possible for a GA pilot receiving no ATS to fly an entire sortie, or possibly several, without knowing that their transponder was unserviceable but they were content that the SU29 pilot had been advised to have his transponder checked. One of the GA Members questioned why the SU29 pilot had selected a Mode 3/A code of 7000 rather than 7004 to indicate that he was carrying out aerobatics; other Members felt that the use of 7004 amongst the GA community was probably not well understood and therefore quite unusual. In assessing the Risk, the Board considered the measured and reported CPA distances and that it was unlikely that the SU29 pilot had seen the Robin. They agreed that the Robin pilot had taken appropriate actions given her experience level but that her avoiding action had probably not had a very large effect. While it was difficult to be certain how serious the collision risk had been, the safety of the ac had been compromised. The Board discussed the safety barriers which were relevant in this Airprox; they concluded that they were aircrew rules and procedures, visual sighting and aircrew action. The Board felt that the Rules and procedures had not produced an effect as no ATC service was sought and the Rules of the Air had not prevented the encounter; the Robin pilot saw the SU29 enabling her to increase separation after the CPA but she did not see it early enough for her avoiding action to have had more than a minor effect. In assessing the combined effect of the barriers remaining to prevent a collision the Board agreed that they were minimally effective resulting in a score of 20 on the Event Risk Classification Matrix. 2

47 PART C: ASSESSMENT OF CAUSE AND RISK Cause: Apparent non-sighting by the SU29 pilot. Degree of Risk: B. ERC Score: 20. 3

48 AIRPROX REPORT No Date/Time: 6 Apr Z (Saturday) Position: 5115N 00049W (3nm SW of Farnborough - elev 238ft) Airspace: Farnborough ATZ (Class: G) Reporting Ac Type: CE560XL PA28 Reported Ac Operator: Civ Comm Civ Trg Alt/FL: 1000ft 800ft NR NR (QFE) Weather: NR VMC CAVOK Visibility: NK 10km Reported Separation: Recorded Separation: 200ft V/0.5nm H 100ft V/500m H 100ftV / 0.1nm H CE560XL CPA 1728:43 100ft V 0.1nm H A19 A11 A : :56 A11 A06 Diagram based on radar data and pilot reports A07 A10 A :56 PA28 NM PART A: SUMMARY OF INFORMATION REPORTED TO UKAB THE CE560XL PILOT reports that he was 3nm from touchdown on Farnborough RW06, descending through 1000ft on the QNH at 160kt for a visual straight in approach, flying VFR in receipt of a DS from Farnborough APP, squawking Mode 3/A 1435, with Modes C and S selected. The ac is white with coloured stripes and was displaying HISLs, navigation lights and taxi lights. As the ac passed 3nm and just before the TCAS inhibition altitude of 900ft, the pilot was changing frequency to Farnborough TWR when he received a TCAS TA. The pilot of the CE560XL did not see the CPA but he heard the other pilot report that he was breaking R to avoid him and he then spotted the other ac slightly behind and to the R of his own ac. THE PA28 PILOT reports heading 330 in the visual cct, on base leg approaching finals, at 70kt and 800ft in a white ac with red and blue markings. His ac was squawking Mode 3/A 4021 with Modes S and C switched on. He was flying VFR in VMC with over 10km of visibility and CAVOK. The ac was being flown by a qualified pilot under supervision from an instructor in accordance with the Club s currency rules. The instructor reports they were flying a normal RH cct to RW06; the supervised pilot had established the ac on base leg in descent towards finals when the instructor reports that he saw the CE560XL approximately 500m away descending in landing configuration. The instructor took control and commenced a RH climbing avoiding action turn away from the CE560XL and informed TWR that he was, going around to avoid other aircraft following which he heard the CE560XL request clearance to land. The PA28 pilot felt that the TWR controller sounded surprised and recalls that the controller asked, confirm on finals? following which the CE560XL was issued landing clearance and he repositioned his PA28 to rejoin the cct on base leg. FARNBOROUGH APP was an instructor supervising an experienced controller who had returned from leave and they report vectoring a series of inbounds to RW06; due to gliding activity at Odiham, all inbounds were carrying out visual approaches. They report that they were aware the visual cct was active with the PA28 in it. They placed the CE560XL under a DS, which they limited from the R due to gliding at Lasham and Odiham. As the ac was vectored inbound APP reports spotting a primary contact at 3nm finals to RW06 and so called TWR who advised that he could see a hot air 1

49 balloon in that position. Farnborough APP report passing TI on the balloon and advised the CE560XL pilot that they would have to break off his approach if he was not visual with it. The pilot confirmed he was visual with the balloon and they instructed him to continue the approach with the balloon in sight. APP reports that they instructed the CE560XL pilot to contact TWR between 4.5 and 4nm. Shortly afterwards TWR called and asked if the CE560XL was coming straight in as he was not in 2-way RT contact with the pilot; APP reports that they then noticed that the PA28 had turned on to base leg at 2nm and so the instructor told TWR to, break off the PA28 whilst the UT tried to contact the CE560XL pilot. FARNBOROUGH TWR reports that he could see the CE560XL N of the extended C/L for RW06 indicating 2400ft but was not in RT contact with the pilot; he reports that he assumed the ac was continuing S and instructed the PA28 pilot to report finals. He then saw on the Air Traffic Monitor (ATM) and through the window that the CE560XL was turning towards finals and he immediately contacted APP to check if the CE560XL was coming straight in ; APP then instructed him to break off the PA28 and he reports that as he about to make the transmission the PA28 pilot informed him that he was taking avoiding action. TWR recalls that the CE560XL pilot then called on his frequency with less than 2nm to run and TWR issued a landing clearance following which the ac landed without further incident. He reports that the PA28 then extended downwind and then also landed without further incident. ATSI reports that the Airprox occurred on Saturday 6 April 2013 at 1728:38 UTC, 1.6nm SW of Farnborough on final approach for RW06 and within the Farnborough ATZ Class G airspace, between a CE560XL and a PA28. The Farnborough ATZ comprises a circle radius 2.5nm, centred on the midpoint of RW06/24 and extending to a height of 2000ft aal (elevation 238ft); except that part of the circle situated N of the M3 Motorway. The CE560XL was inbound IFR from Liverpool and was in receipt of a DS from Farnborough APP, with reduced traffic information due to gliding activity in the area. The CE560XL was cleared for a visual approach and had been transferred to Farnborough TWR on frequency MHz just prior to the Airprox. The PA28 was in the visual RH cct for RW06 and was in receipt of an Aerodrome Control Service from Farnborough TWR on frequency MHz. Farnborough APP was manned by an experienced controller acting as OJTI whilst training an experienced controller who was undergoing training following a return to work after a prolonged leave. The TWR controller was relatively new to the unit and had completed a VCR validation in June The Farnborough Manual of Air Traffic Services (MATS) Part 2, Page APR -13, Paragraph 2.12, states: Aircraft with an ORCAM squawk that are code-callsign converted do not need to be warned in to the VCR controller. However, there are occasions when an aircraft SHALL be warned in: When the inbound aircraft goes visual or makes a non-standard approach There was some ambiguity in the interpretation of this instruction and controllers accept that this would normally apply for ac making an ILS approach and requesting a visual or non-standard approach. However, it has become common practice when operating on RW06 in good weather conditions with glider activity, that when all IFR inbounds are making visual approaches, there was no requirement to warn-in the arrivals as it was assumed they will be completing a visual approach. However this procedure was not detailed in the Farnborough MATS Part 2. The visual cct is not very often active at Farnborough as the unit is not a training airfield and controllers may operate for a period of months without seeing any cct traffic. In addition there are 2

50 four radar positions and one TWR position and it is therefore common for dual valid controllers to spend a greater percentage of their operational time in radar. Because of the good weather the unit had been very busy with heavy LARS traffic throughout the day. The incident happened in the late afternoon and traffic levels had reduced and were considered to be light. The Approach and LARS(W) positions were not band-boxed. Throughout the day the unit had experienced multiple primary contacts believed to be gliders and this had resulted in traffic being tactically vectored onto a base leg for the visual approach to RW06. CAA ATSI had access to the RTF recording of Farnborough LARS(W), the area radar recording, written reports from the two pilots, the Farnborough LARS(W) controller report together with the Farnborough unit report. The workload and traffic levels of the LARS(W) controller were reported as low but with added complexity due to the gliding and balloon activity. The Farnborough weather is provided: METAR EGLF Z 06007KT 030V100 CAVOK 09/M05 Q1023= At 1716:10, the PA28 rejoined overhead for RH ccts on RW06 and shortly after TWR took over from the off-going controller. The PA28 continued in the cct and an inbound GLEX ac not involved in the Airprox was being vectored for a visual approach. At 1719:22, the PA28 requested a bad weather cct and shortly afterwards the GLEX at an altitude of 4200ft, was shown on a southerly track, crossing RW06 final approach at a range of 2.8nm. The GLEX was then vectored approximately 7nm S of the airfield followed by a R turn for a wide base leg to RW06. At 1723:40 APP passed traffic information on the PA28 in the GLEX ac s 1 o clock at a range of 3.5nm at 1000ft. At 1724:00, the PA28 was on short final and the GLEX reported the PA28 ahead in sight. The GLEX was then cleared for the visual approach turning final at not less than 3 miles DME. At 1724:16, the CE560XL contacted Farnborough APP maintaining an altitude of 6000ft. APP advised the CE560XL pilot to expect vectors for a visual approach to RW06 with current information W and QNH1023. The CE560XL was placed on a radar heading, descended to 3400ft and advised that on passing 5000ft, [when leaving CAS] to expect a DS with reduced traffic information due to gliding activity at Lasham and Odiham. At 1725:20, the CE560XL pilot was 15nm from touchdown and was asked to confirm that he would be able to accept a visual approach from that position. The CE560XL replied, Affirm and APP gave the CE560XL further descent to 2400ft. APP then observed a primary contact ahead of the CE560XL, 3nm from touchdown and at 1726:22 he telephoned the Tower to ask if TWR could see anything on a 3nm final. TWR confirmed that he could see a balloon. The balloon was not shown on the area radar recording. [At interview TWR estimated that the balloon was probably between 2000ft and 3000ft.] At 1726:33, APP transmitted, (CE560XL)c/s are you visual with a hot air balloon in your twelve o clock range of five miles. The CE560XL replied negative and APP responded, (CE560XL)c/s roger it s on three mile final approach and er if you re not visual with the hot air balloon I m going to need to break you off the approach and wait for it to move. The CE560XL pilot replied, Er okay?????????? (CE560XL)c/s and then, Er affirm we have it in sight (CE560XL)c/s. At this point APP had started a telephone call to the TWR but then advised TWR, it s OK he s got it in sight now and the call was terminated. 3

51 At 1726:36, the PA28 pilot reported downwind and TWR responded, (PA28)c/s Roger report final. The CE560XL is shown 7nm W of the airfield squawking 1445 with the PA28 downwind RH (Picture 1). (Picture 1 MRT at 1726:34) At 1727:13, APP asked the CE560XL pilot if he was happy to continue for the visual approach with the balloon in sight. The pilot replied, Continue approach with the balloon in sight. APP responded, And (CE560XL)c/s you are cleared the visual approach further descent in accordance with the standard noise abatement procedures to be established not less than three DME. This was acknowledged by the CE560XL pilot. The CE560XL was 5nm W of the airfield at 2700ft (Picture 2). (Picture 2 MRT at 1727:13) At 1727:34, the CE560XL was instructed to contact Farnborough Tower on 122.5MHz. This was acknowledged by the CE560XL pilot and this transmission ended at 1727:44. When questioned TWR believed that he saw the CE560XL on the ATM at 1900ft and had started to become concerned, but because the CE560XL had not called, he considered that APP probably intended to route the CE560XL S of the C/L to avoid the balloon. At 1728:13, APP showed the PA28 had turned onto base leg and the CE560XL was beginning to commence a L turn towards final approach (Picture 3). 4

52 (Picture 3 MRT at 1728:13) The CE560XL was indicating a groundspeed of 179 knots and the PA28 69 knots. TWR recognised that the CE560XL was turning towards final approach and at 1728:23, TWR telephoned APP (in an urgent tone), is the CE560XL coming in now (Picture 4). (Picture 4 MRT at 1728:23) APP responded, Er should be with you sorry-yes he s visual and TWR advised that the CE560XL had not called. Both controllers attempted to contact the CE560XL at 1728:36. The distance between the two ac was 0.3nm (Picture 5). 5

53 (Picture 5 MRT at 1728:36) At the end of the call (1728:38) the OJTI instructed TWR to break off the PA28. TWR responded, He s breaking off. The PA28 was shown commencing a R turn (CPA) with a minimum separation of 0.1nm at the same level (Picture 6). (Picture 6 MRT at 1728:38) At 1728:42, the CE560XL pilot called, (CE560XL) Farnborough Tower and the PA28 called, (PA28)c/s is avoiding. Radar showed two ac on parallel tracks with a minimum horizontal distance of 0.1nm (Picture 7). 6

54 (Picture 7 MRT at 1728:42) The CE560XL pilot reported, Er (CE560XL)c/s we re final Runway zero six. TWR replied, (CE560XL)c/s you are clear to land the wind zero five zero six knots there is traffic just turned right is a P A twenty-eight. The CE560XL pilot responded, Er we just saw it (CE560XL)c/s. At 1729:10, the PA28 pilot reported, (PA28)c/s apologies for that we were not aware of his approach. The CE560XL pilot advised, Er we were cleared to it. At interview TWR indicated that after he had taken over the watch, he noted that the previous inbound GLEX had been positioned to the S before joining base leg and had an expectation that the CE560XL would do the same. TWR believed that this opinion was reinforced when he thought APP was avoiding the balloon and again when the CE560XL hadn t called. TWR also indicated that, in his experience, the operating company would accept a delay in the interest of safety and he thought that they would probably be avoiding the balloon by routeing S of the C/L. TWR recalled that when the CE560XL was at 2400ft, APP confirmed that the pilot had sighted the balloon and TWR still believed the CE560XL was positioning to the S. Later when the CE560XL was at 1900ft he began to have concerns and monitored the CE560XL on the ATM. As the CE560XL started turning towards final he contacted APP to confirm their intentions and advise them that the ac had not called on the Tower frequency. TWR described how his experience of RW06 operations with an active cct was fairly limited and ccts are not a common feature at Farnborough. In hindsight, TWR believed that rather than giving the PA28 an early clearance to final when the CE560XL was within 15nm, he should perhaps have kept his options open by tactically holding the PA28 at the end of the downwind leg. TWR indicated that the PA28 had previously completed a bad weather cct and this may have led him to believe that his next cct would also be compact, which was not the case. At interview APP confirmed that the plan was to vector the CE560XL for a visual approach on L base. When questioned, APP confirmed that TWR was expected to monitor arriving ac on the ATM and APP does not generally provide a warning-in check. TWR would use his experience and judgement to manage the cct appropriately. There was some discussion regarding the variation in the two approaches made by the GLEX positioned S of the airfield and the CE560XL positioned for L base. APP indicated that this depended very much on the level of the ac and traffic situation, but it was not unknown for CPT inbounds to accept L base with a greater rate of descent. It was agreed that this procedure required an experienced understanding of the varied operations at Farnborough. APP believed that the additional RT and phone calls generated by the appearance of the balloon were contributory factors that delayed the transfer of the CE560XL to the Tower and this was 7

55 compounded when it took a minute for the crew to contact the Tower. [58 seconds from the end of one transmission to the beginning of the next]. When asked about the comparison between the arrival of the GLEX with traffic information about the PA28, compared to that of the CE560XL, APP indicated that when transferring the GLEX, it was apparent that he would be number 2. At the point the CE560XL was transferred to the Tower, the PA28 was observed downwind and there was an expectation that TWR would probably orbit the PA28 at the end of the downwind leg. There was some additional discussion regarding the CAA safety notice 2013/001- Integrating Traffic in the Vicinity of an Aerodrome. APP confirmed that due to the limited amount of cct activity, the unit were developing an Aerodrome control training module aimed at defensive controlling techniques. APP advised that in Class G airspace there is no requirement for airspace users to contact Farnborough. However the ATSU does have a very pro-active policy of communicating and interacting with local airspace users, aimed at encouraging them to participate either by contacting or communicating their intentions to Farnborough. The balloon operator had not done so on this occasion. As a result of this occurrence the ATSU issued an instruction to controllers (TOI 005/13) subsequently revised and replaced by TOI 008/13, which states: Controllers are reminded that existing unit procedures require any arrival other than an ILS to be notified to the VCR controller prior to commencing the approach at a suitable time, with the following exception: - During periods of Odiham glider activity, any ac inbound to Farnborough expected to complete a Visual approach shall be notified to the VCR controller at or approximately at 10nm from the airfield. VCR controllers should consider using defensive controlling techniques following receipt of such advice, with particular reference to other RW activity. This may include holding any cct traffic at a suitable location, or delaying outbound traffic. ANALYSIS There was some ambiguity in the guidance to controllers in the Farnborough MATS Part 2. The practice of not requiring APP to warn-in the arrivals when all IFR inbounds will be making a visual approach, led to the misunderstanding. The point at which ac request or elect to continue for the visual approach can vary, especially when in some situations the arrival might be vectored S of the airfield to lose height. Whenever there is doubt, the act of coordination or warning-in can clarify an intended or changed plan. Having just taken over the position, TWR observed the previous arrival position S of the airfield before turning back. This, combined with the distraction of gliders and the balloon on a 3nm final, very likely gave him an expectation bias that the CE560XL would follow the same arrival route. TWR s perception of the operating company led him also to believe that they would very likely take a short delay in order to avoid the balloon. This caused TWR to misunderstand the intentions of APP to position the CE560XL on L base. There had been no communication from the balloon company or pilot regarding their flight, which was shown intermittently on the Farnborough Radar display at 3nm on final approach. The balloon was not shown on the area MRT radar. The balloon disrupted and delayed the normal sequence of events, with consideration for a change of plan and cct combined with phone calls to the Tower and additional RT communications with the CE560XL crew. This delayed the normal transfer of the CE560XL to the Tower. At this point the CE560XL crew were probably concerned with preparation 8

56 for landing, positioning around the balloon and then contacting the Tower. There was consequently a delay of 58 seconds before the CE560XL contacted the Tower, which was just as the Airprox occurred. The altitude of the CE560XL as it approached L base should have signalled to TWR that the CE560XL was likely to be turning onto final. However, because the CE560XL had not called the Tower at the range expected for a L base join, this reinforced TWR s perception that the CE560XL was going to go through the C/L, probably to avoid the balloon. When the CE560XL turned towards final, the PA28 had already turned onto base leg and there was very little time to react. TWR was confused and rather than instinctively recognising the urgency of the situation and giving avoiding action or advice to the PA28, he contacted APP and asked if the CE560XL was coming straight in. At this point it was too late for the ATC to recover from the situation and the PA28 pilot executed an avoiding R turn. The unit procedures do not have a warning-in system for notifying or updating TWR regarding arrivals during the intermediate approach phase. Best practice at the unit requires TWR to monitor the ATM and to interpret the sequence and intentions of the APP operation. APP will coordinate with TWR when considered appropriate. Had TWR known that the CE560XL was joining on L base for the visual approach, he would have delayed the PA28. If there had been a doubt in the controller s mind he could have acted defensively by delaying the PA28 at the end of the downwind leg. The unit procedures rely on the experience of controllers and CAA ATSI consider that these procedures allowed for the possibility of ambiguity and misunderstanding. Farnborough do not operate a busy visual cct and there is little opportunity to develop defensive skills in the integration of arriving IFR and VFR cct traffic. In January 2013, the CAA promulgated Safety Notice 2013/001 regarding this subject. CONCLUSIONS The Airprox occurred when TWR misinterpreted the intentions of APP and cleared the PA28 to final approach, which brought the PA28 into conflict with the CE560XL that had also been cleared to final by APP. The following were considered to have been contributory factors: 1. The unit practice of not warning-in arrivals to the Tower in circumstances when all ac are making visual approaches, was not included in the ATSU MATS Part 2 and was at variance with the instruction which required controllers to warn-in ac on occasions when inbound ac make a visual or non-standard approach. 2. Without the appropriate coordination or warning-in, the unit practices and procedures relied upon TWR s experience and use of ATM, to interpret the intentions of APP in order to manage the arrival of the CE560XL into the visual cct. 3. The unexpected appearance of the balloon on final approach caused APP to consider an alternative plan that involved additional RTF and telephone communication. This, together with the increased workload on the flight-deck, resulted in the delayed transfer of the CE560XL to the Tower, which served to reinforce TWR s belief that the CE560XL would be vectored through the C/L to the S. CONCLUSIONS Notwithstanding the content of Farnborough TOI 2013/008, which expires on 30 June 2013, it is recommended that in light of this occurrence and CAA Safety Notice 2013/001, Farnborough ANSP 9

57 review and incorporate appropriate procedures into their MATS Part 2, for the integration of IFR arrivals into the visual cct pattern. PART B: SUMMARY OF THE BOARD S DISCUSSIONS The Chairman opened the discussion by praising APP for ensuring a seamless transition between the CE560XL s departure from CAS and the application of a DS. The discussions focussed on three areas: the steepness of the CE560XL s approach, the ATC procedures for notifying visual approaches to TWR and the actions of the TWR controller. A pilot Member familiar with the CE560XL family of ac advised the Board that the ac s steep descent was well within the capabilities of the ac, could be considered normal for an ac of this type and would not be uncommon at airports such as this one. It was agreed that the approach profile of this ac did not contribute to the occurrence sequence. Several ATC Members expressed surprise that the Unit procedures did not require APP to notify TWR of visual inbounds in these circumstance and commented that the inbound notification call was usually intended to alert TWR and allow him to plan more accurately for cct integration rather than simply to notify the type of approach. The discussion moved on the presence of the hot air balloon and its effect on the incident sequence. Members opined that the balloon did not contribute directly to the event but that the process of monitoring and passing TI on it and visually searching for it had been a minor distraction for APP, TWR and the CE560XL crew. Considering the actions of APP and TWR, ATCO Members agreed that the lack of a procedure for notifying visual inbound traffic to TWR, and the expectation that TWR would listen out and glean the required information had lead to TWR making incorrect assumptions about the expected flight path of the CE560XL; this procedural factor had been fundamental to the genesis of this encounter. The lack of notification about the CE560XL s arrival from APP and TWR s assumption that the CE560XL would follow the track of the previous inbound contributed to TWR s erroneous SA and resulted in the close proximity of the CE560XL and the PA28. Taken together, the Board concluded that the Cause of the Airprox was that ATC did not integrate the ac safely into the visual cct. Turning to the resolution of the conflict, ATC Members felt that when TWR realised there was a conflict he should have broken the PA28 off without delay rather than calling APP for further information on the CE560XL. In the event ATC did not contribute to the resolution of the conflict. For their part, the CE560XL crew had some information from their TCAS and a TA, but they did not see the PA28 until after hearing the PA28 pilot call going around and the CPA. Fortunately the PA28 pilot had checked the final approach before turning in from base leg and took appropriate action when he realised that the other ac was proceeding straight to finals. The Board noted that the activation of visual only approaches to this RW was an infrequent event and felt this made it even more important that a co-ordination procedure for TWR and APP was put in place. They discussed the recommendations made in the ATSI report and sought the opinion of the SRG Advisor who, whilst unable to speak for any Regional Inspectors, felt that this occurrence had similarities with some previous events and the recommendations were commensurate with the guidance in CAA Safety Notice 2013/001 regarding the coordination and integration of visual cct traffic with inbound ac. The Board considered that the recommendations made by ATSI were appropriate and did not require any additions. In assessing the Risk, the Board considered the CPA distance of 100ft V and 0.1nm H and that the CE560XL crew had not seen the PA28 in time to take any effective action. There was some discussion about whether this was a B or a C Risk and opinion was closely divided. It was decided by a vote that although the distances were quite close, in the context of a cct environment and as the PA28 pilot had seen the other ac in time to take fully effective avoiding action, albeit quite late, that the event was a C Risk. The Board discussed the safety barriers which were relevant in this Airprox and which had been effective in preventing a collision. The Board concluded that none of the ATC safety barriers had 10

58 been effective and the Airprox was caused when ATC did not integrate the ac safely into the visual cct. Both aircrews had a responsibility to see and avoid other aircraft but this barrier had only a limited effect as the CE560XL crew did not see the PA28 before the CPA. Overall the Board assessed the barriers to have had a limited effect resulting in a score of 21 on the ERC matrix. PART C: ASSESSMENT OF CAUSE AND RISK: Cause: ATC did not achieve safe integration of visual cct traffic. Degree of Risk: C ERC Score: 21 11

59 AIRPROX REPORT No Date/Time: 20 Apr Z Position: 5201N 00012W (3 5nm E Henlow - elev 170ft) Airspace: LFIR (Class: G) Reporting Ac (Saturday) Reported Ac Type: Vigilant T1 SportCruiser Operator: HQ Air (Trg) Civ Pte Alt/FL: 1500ft 1200ft QFE (1027hPa) QNH Weather: VMC CLOC VMC CLOC Visibility: >10km 10km Reported Separation: 50ft V/300m H Recorded Separation: 0 1nm H 200ft V/0 5nm H SportCruiser Henlow ~3nm A20 48:42 A16 Luton CTA ft LTMA 4500ft+ 48:22 A18 48:46 NMC 48:42 NMC NMC 1448:02 Graveley ~3 5nm Radar derived Levels show altitudes as Axx LON QNH 1031hPa 0 1 Vigilant 48:22 NMC 48:46 A16 NM LTMA 4500ft+ Luton CTA ft LTMA 3500ft+ PART A: SUMMARY OF INFORMATION REPORTED TO UKAB THE VIGILANT T1 PILOT reports flying a dual training sortie from Henlow, VFR and in receipt of an A/G service from Henlow Radio on 121 1MHz, squawking 7000 with NMC. The visibility was 10km in VMC and the ac was coloured white with hi-vis wing markings. Having completed the exercise and in preparation to rejoin the RH cct downwind for RW09R she descended using S-turns to the E of Henlow just W of the A1. Despite carrying out full lookout scans before and during the turns and seeing no traffic, when she exited heading 270 at 60kt and 1500ft QFE 1027hPa descending and checked her blind-spot she noticed a low-wing ac close behind (300m) and about 50ft above. The other ac appeared to be tracking the A1 S bound and it may have had to change track to the L to avoid her ac. She took no avoiding action owing to her late sighting and any potential confliction had passed. At the time she was the only ac which had reported rejoining on frequency. She assessed the risk as low. After landing she checked with the Duty Instructor with respect to transiting ac but no flights had been logged. At the time of the Airprox she had been a sufficient distance from the airfield where a call from transiting ac would not be required under good airmanship. Civilian ac operating from Henlow were unlikely to be flying in this area as they were operating from RW09L with a LH cct to the N of Henlow; the club had reported no incidents. Members were reminded of the possibility of ac hand-railing line features such as the A1 road. THE SPORTCRUISER PILOT reports en-route from Coventry to a private site 6nm SSE Henlow and in receipt of a BS, he thought, from Luton Radar on Mhz, squawking an assigned code with Modes S and C. The visibility was 10km in VMC and the ac was coloured white/red with strobe lights switched on. He had commenced a gentle descent heading S at 100kt in preparation for landing with about 4nm to run. Descending through 1200ft QNH, he thought, ATC called him with an alert to pop-up traffic below and ahead crossing L to R. He immediately saw another ac, a Grob Motorglider, 0 5nm ahead and ft below so he turned L to avoid passing directly O/H it. Having done this his co-pilot in the RH seat saw the other ac pass down their RHS by 0 5nm and 200ft below. He assessed the risk as low. He opined that given the proximity of Henlow to the Luton CTR it would be useful if Henlow traffic monitored the Luton frequency. ATSI reports that the Airprox occurred at 1448:46 UTC, 3 5nm to the E of Henlow Airfield, within Class G airspace, between a Vigilant T1 and a SportCruiser. 1

60 Henlow does not have an ATZ but is promulgated in the AIP as having aerial sporting and recreational activity. ENR Page (4 Apr 13) promulgates parachute activity within a circle of 1 5nm and vertical limit of 3500ft. ENR Page (4 Apr 13) promulgates glider activity within a circle of 2nm with no specified vertical limit. The Vigilant was returning to Henlow VFR, after the completion of a local flight and was in communication with Henlow Radio (A/G) on frequency 121 1MHz. The SportCruiser was operating on a VFR flight from Coventry inbound to Graveley, which lies 6 3nm NE of Luton airport within the Luton CTR, and was in receipt of a TS from Luton Radar on frequency MHz. The LTC Luton INT (Radar) controller s workload was assessed as medium, with a number of zone transit ac and IFR inbound ac being vectored for the ILS RW08. CAA ATSI had access to RT recordings for LTC Luton Radar, together with area radar recording and written reports from the 2 pilots concerned. The Luton METARS are provided: EGGW Z 09006KT 040V150 CAVOK 13/M03 Q1031= and EGGW Z 13005KT 080V200 CAVOK 13/M03 Q1031= At 1439:20 the SportCruiser flight contacted Luton Radar and, once 2-way communication was established, reported, (SportCruiser c/s) SportCruiser Coventry to Graveley we re erm just er west of Saint Neots at this time four thousand er three hundred feet one zero three one requesting er traffic service and zone penetration for Graveley. The SportCruiser pilot was instructed to squawk 4671 and shortly afterwards was identified by Luton Radar, (SportCruiser c/s) you are identified er twenty miles north-northeast of Luton on a Traffic Service with Luton QNH one zero three one. This was acknowledged correctly. At 1442:41, the SportCruiser was shown 9 4nm N of Henlow and was given a clearance to enter the Luton CTR, (SportCruiser c/s) you re cleared to enter the Luton zone er on your own navigation towards Graveley when ready not above two thousand four hundred feet VFR. The pilot replied Clear to enter the zone not above two four zero zero feet one zero three one (SportCruiser c/s). At 1445:02, the Luton Radar controller passed TI to the SportCruiser flight regarding another ac which was crossing 2 1nm ahead. At 1446:44, the SportCruiser was 5nm NE of Henlow tracking S and the Vigilant was shown 6nm S of the SportCruiser s position tracking E. At 1447:03, the label of the Vigilant below CAS merges with an inbound ac, which was within CAS at 5000ft and being vectored downwind LH for RW08. Shortly afterwards at 1447:28 the Vigilant was shown to have turned onto a N ly track. The distance between the 2 ac was 2 5nm. At 1448:00 the Luton Radar controller advised, (SportCruiser c/s) there s pop up traffic just left of your twelve o clock may be low level no height information. The pilot responded, (SportCruiser c/s) looking er visual. The SportCruiser was at an altitude of 2000ft and Vigilant was shown to have commenced a R turn (Picture 1). 2

61 (Picture 1 -MRT radar at 1448:02) At 1448:42, the Vigilant had completed a RH orbit and was shown tracking W. between the 2 ac was 0 1nm (Picture 2). The distance (Picture 2 -MRT radar at 1448:42) Shortly afterwards at 1448:46, the SportCruiser at 1600ft was shown to have made a slight L turn passing 0 1nm behind the Vigilant (Picture 3). 3

62 (Picture 3 -MRT radar at 1448:46) At 1449:05, the SportCruiser pilot reported on final for Graveley and reported that he would change squawk to 7000 when on the ground. This was acknowledged by the Luton Radar controller. The Luton radar controller was not aware that an Airprox report had been made and consequently no report was received from the controller. The SportCruiser was in receipt of a TS from Luton Radar. The Luton Radar controller passed TI on pop-up traffic, which resulted in the SportCruiser becoming visual with the other traffic. CAP744, Chapter 3, Page 1, Paragraph 1, states: A Traffic Service is a surveillance based ATS, where in addition to the provisions of a Basic Service, the controller provides specific surveillance-derived traffic information to assist the pilot in avoiding other traffic. Controllers may provide headings and/or levels for the purposes of positioning and/or sequencing; however, the controller is not required to achieve deconfliction minima, and the avoidance of other traffic is ultimately the pilot s responsibility. Paragraph 5 states: The controller shall pass traffic information on relevant traffic, and shall update the traffic information if it continues to constitute a definite hazard, or if requested by the pilot. However, high controller workload and RTF loading may reduce the ability of the controller to pass traffic information, and the timeliness of such information. CAP774, Chapter 1, Page1, Paragraph 2, states: Within Class F and G airspace, regardless of the service being provided, pilots are ultimately responsible for collision avoidance and terrain clearance, and they should consider service provision to be constrained by the unpredictable nature of this environment. The SportCruiser flight, in receipt of a TS, was passed TI by the Luton Radar controller and reported the Vigilant in sight. The Airprox occurred when the SportCruiser and Vigilant passed in close proximity within Class G uncontrolled airspace. HQ AIR (TRG) comments that the Airprox was reported as a result of an effective non-sighting by the Vigilant pilot, who was concerned by the proximity of the other ac. Lookout is recognised as the primary mitigation for Vigilant operations and is taught, practiced and assessed accordingly; however, it is never 100% effective. Vigilant operators, particularly when in the immediate vicinity of 4

63 home base, will routinely have their unit s AG frequency on their single radio so could not monitor other frequencies. By way of additional mitigation, work is ongoing to fit PowerFLARM to the RAF s Vigilant fleet during 2013, which should in future provide a degree of improved awareness of transponding traffic and other FLARM/ADS-B equipped ac. PART B: SUMMARY OF THE BOARD'S DISCUSSIONS Information available included reports from the pilots of both ac, transcripts of the relevant RT frequencies, radar video recordings, reports from the air traffic controllers involved and reports from the appropriate ATC authorities. As this encounter occurred in Class G airspace, both pilots were responsible for maintaining their own separation from other traffic through see and avoid. The SportCruiser pilot had enhanced his SA by receiving a TS from Luton Radar and the controller passed TI on the Vigilant when the confliction became apparent. The SportCruiser pilot saw the Vigilant and initiated a L turn to pass behind it. However, the Vigilant instructor did not see the SportCruiser approaching from her R, which had right of way, only visually acquiring it as it passed behind and above. Although the pilot would have had little opportunity to see the SportCruiser whilst belly-up during the first part of her R turn through 270 from N, through E to W towards Henlow, thereafter Members believed that there should have been enough time for her to clear her flightpath, prior to the CPA, and that this effective non-sighting by the Vigilant crew had caused the Airprox. That said, the SportCruiser pilot s sighting and action taken left the Board in no doubt that any risk of collision had been quickly and effectively removed. In assessing the effectiveness of the safety barriers remaining, the Board agreed that the ATC barriers had been effective, the SportCruiser pilot s SA was improved from the TI given. However, with the incident only being observed by one of the crews prior to the CPA, the see and avoid safety barrier had had limited effectiveness, the Board assigning an ERC score of 4. PART C: ASSESSMENT OF CAUSE AND RISK Cause: Effectively a non-sighting by the Vigilant crew. Degree of Risk: C. ERC Score: 4. 5

64 AIRPROX REPORT No Date/Time: Position: 25 Apr Z 5156N 00324W (1nm W Liverpool) Airspace: Liverpool CTR (Class: D) Reporting Ac Reported Ac Type: A319 PA38 Operator: CAT Civ Club Alt/FL: 1000ft 1400ft QNH (1019hPa) QNH NR Weather: IMC VMC NR Visibility: NR 10km Reported Separation: 0ft V/0.5nm H Recorded Separation: 0ft V/1.2nm H 200ft V/1.0nm H 2-3nm H PART A: SUMMARY OF INFORMATION REPORTED TO UKAB THE A319 PILOT reports conducting the initial climb after T/O from Liverpool John Lennon A/D. He was operating under IFR in IMC through cloud layers with a RCS from Liverpool TWR [ MHz], he thought. The ac had beacon and landing lights selected on but the strobe lights were U/S. The SSR transponder was selected on with Modes A, C and S and the ac was fitted with TCAS II. Liverpool TWR advised him of light ac traffic approaching the visual cct from the W, not above altitude 1500ft. Almost immediately afterwards, whilst IMC at a position 1nm W of the RW09 threshold, heading 280 at 160kt and climbing through altitude 850ft, he received a TCAS TA. He became visual with traffic in his R 1 o clock shortly afterwards, whilst between cloud layers, on a reciprocal heading, as he climbed through its level at altitude 1000ft and at a range of about 0.5nm. He advised Liverpool TWR of the Airprox after reaching cruise level and was informed that the conflicting traffic had been cleared to Garston Docks and to remain VMC and within sight of the A319. The A319 pilot did not believe this would have been achievable under the meteorological conditions prevailing at the time. He assessed the risk of collision as Medium. THE PA38 PILOT reports conducting a trial flying lesson, seated in the RH seat with the customer in the LH seat. He was operating under VFR in VMC in contact with Liverpool APP [ MHz]. The white and red ac had navigation, landing and strobe lights selected on, as was the SSR transponder with Modes A and C. The ac was not fitted with an ACAS. He had planned to route via the River Mersey to the Seaforth VRP but after departure from RW27 could see that the weather to the N was not suitable. He assessed the weather to the S as CAVOK and requested a zone departure via the Chester VRP. He was re-cleared to the Garston Docks VRP and advised to report when visual with a departing A319, which was routeing to the Wallasey VOR. On completion of his turn on to track for the Garston Docks, heading 120 at 85kt and altitude 1400ft, he reported visual with the A319, some 2-3nm to the SW, and was re-cleared behind it to the Chester VRP. He assessed the risk of collision as None. 1

65 THE LIVERPOOL ATC INVESTIGATOR reports that the PA38 pilot intended to leave Liverpool CAS to the N under VFR, having departed from RW27, and was initially routeing to the N. However, due to showers in the vicinity, the pilot requested to change routeing to the S towards Chester. The controller cleared him to Garston Docks, an intermediate VRP within the CTR, which is approximately 1nm N of the climb out path. This is not a normal route for VFR ac when Liverpool operates from RW27, but the controller specified this VRP as a clearance limit and had given TI on the departing A319. The PA38 pilot advised the controller that he had the A319 in sight. The PA38 pilot maintained visual separation from the departing A319 iaw VFR in Class D airspace. However, this was not sufficient to avoid a TCAS TA being activated, which caused some concern to the A319 crew. No avoiding action was required, and the A319 pilot was able to continue the climb without further incident. ATC acknowledged that the use of Garston Docks VRP under these circumstances was inadvisable. He reported the Liverpool weather as follows: METAR EGGP Z 27014KT 240V BKN030 12/08 Q1019= METAR EGGP Z 29015KT SHRA SCT007 BKN015 10/09 Q1019= [UKAB Note (1): A graph of altitude and separation range was constructed, as follows: ATSI reports that this Airprox occurred at 1232:38, 2.5nm to the W of Liverpool A/D, on the edge of the Liverpool ATZ and within the Liverpool CTR Class D CAS, between an Airbus A319 and a Piper PA Tomahawk (PA38). The Liverpool ATZ comprises a circle radius 2.5nm, centred on the midpoint of RW09/27 and extending to a height of 2000ft aal (elevation 81ft). Background The A319 was departing under IFR and was in receipt of an ACS from Liverpool Tower [ MHz] subject to a radar release from RAD. The PA38 was departing on a local flight to the N but due to poor weather had been instructed to hold at Garston Docks which lies 1.62nm to the NW of the RW09 threshold and 1nm N of the extended C/L. The PA38 was operating on a VFR clearance and in receipt of a RCS from Liverpool RAD. ] 2

66 CAA ATSI had access to RTF recordings for Liverpool TWR and RAD, area radar recording, written reports from the two pilots concerned and a report from the ATSU. Factual History The PA38 departed from RW27 and contacted Liverpool RAD at 1229:23, reporting on track for VRP Seaforth which lies to the NW of Liverpool at the mouth of the River Mersey. At 1229:07, the RAD contacted TWR and released the A319 for departure climbing straight ahead to 4000ft. At 1230:50, the TWR cleared the A319 for take-off RW27. Meanwhile, due to low cloud in the area, the PA38 pilot requested a reversal of routeing to leave the CTR at Chester (S of the A/D). The RAD instructed the PA38 pilot to remain at Garston Docks, which was acknowledged by the PA38 pilot. At 1231:09, RAD advised the TWR that the PA38 pilot had requested to reverse course and route direct to Chester. The TWR reported the A319 would be airborne at At 1231:30 the RAD passed TI to the PA38 pilot, reporting that an A319 was departing from RW27 to the W and to expect onward clearance to Chester once the A319 had passed abeam. The RAD instructed the PA38 pilot to remain N of Garston docks, which was acknowledged correctly. At 1232:08, as the A319 approached 1000ft, the TWR advised the A319 pilot that he may see traffic on TCAS in [his] R 1 o clock at a range of 3nm, a Tomahawk at not above 1500ft remaining in that location. The A319 pilot acknowledged with, Copied. Radar showed the range between the two aircraft as 2.3nm (see Figure 1 below). Figure 1: Composite Radar Picture at 1232:08 At 1232:21, the RAD asked the PA38 pilot if he was visual with the departing A319 to which the pilot confirmed that he was. At this point the TWR contacted RAD to report that there was very low cloud and rain in the vicinity of the two ac. The RAD confirmed that the PA38 pilot had sighted the A319 but would be held N of Garston Docks. At 1232:38, the PA38 pilot reported that he was in visual contact with the A319. The RAD cleared the PA38 pilot to route behind the departing A319 and cautioned him that the recommended Wake Turbulence separation distance was 5nm, which was acknowledged (see Figure 2 below). 3

67 Figure 2: Composite Radar Picture at 1232:38 At 1232:55, the TWR instructed the A319 pilot to contact Scottish Control [ MHz]. At 1238:10, the PA38 pilot reported leaving the CTR at Chester and a BS was agreed. At 1245:01, the A319 pilot contacted the TWR and requested details on the conflicting traffic, advising that on departure the PA38 had come within 0.5nm of the A319 in IMC at 1000ft. The A319 pilot advised that he would contact ATC if he intended to file a report. Analysis The PA38 pilot had planned to leave the Liverpool VFR via Seaforth VRP but became concerned about the low cloud and requested an alternative clearance. The RAD instructed the PA38 pilot to route initially to Garston Docks. For VFR flights within Class D airspace the Manual of Air Traffic Services (MATS) Part 1, Section 3, Chapter 4, Page 1, Paragraph 3.4 states: When issuing instructions to VFR flights, controllers should be aware of the over-riding requirements for the pilot to remain in VMC, to avoid obstacles and to remain within the privileges of his licence. This may result in the pilot requesting an alternative clearance, particularly in marginal weather conditions. The A319 was already rolling when the TWR was made aware of the change of routeing for the PA38, and the TWR passed TI to the A319 pilot at the earliest opportunity, once he was safely airborne. The RAD was content that there was sufficient separation between the A319, climbing straight ahead, and the PA38, operating VFR and holding 1nm N of the C/L. This belief would have been reinforced when the PA38 pilot reported the A319 in sight and confirmed he was holding N of Garston Docks. The A319 pilot was not aware of the PA38 pilot s change of routeing prior to departure and the late TI in the low cloud and rain most likely caused him some concern. MATS Part 1, Section 1, Chapter 2, Paragraph 2 defines the minimum services that are to be provided in each airspace classification. The relevant part of the table is reproduced below: 4

68 The TWR was not able to pass timely TI to the A319 pilot and there would have been little opportunity for him to assess the situation or request avoiding action had he considered it appropriate. The PA38 pilot had reported the A319 in sight and both the TWR and RAD were content that safety would not be compromised. Due to the limited time available and the weather conditions, the A319 pilot was advised about the situation at the earliest opportunity. This occurred just after the A319 became airborne and the A319 pilot was understandably concerned about the relative position of the other traffic in what he considered to be IMC conditions. Conclusions The PA38 pilot elected to reverse direction due to low cloud NW of the A/D which resulted in him being held at Garston Docks. The PA38 pilot, operating under VFR, was passed appropriate TI and reported visual with the A319. The late change of plan, which occurred as the A319 was rolling, resulted in the late passing of TI to the A319 pilot, as soon as he was safely airborne. Whilst the PA38 pilot and the two controllers were content that safety was not compromised, the A319 pilot became concerned about the relative position of the PA38 in IMC conditions immediately after T/O. PART B: SUMMARY OF THE BOARD'S DISCUSSIONS Information available included reports from the pilots of both ac, transcripts of the relevant RT frequencies, radar photographs/video recordings, reports from the air traffic controllers involved and reports from the appropriate ATC and operating authorities. Board Members agreed that the A319 pilot had been startled by the proximity of the PA38, to which the TI and TCAS had drawn his attention. This was aggravated by his perception of the weather conditions and he had, quite correctly, decided to file an Airprox. The PA38 pilot had been faced with deteriorating weather to the N and had changed his routeing intentions, which he communicated to ATC. He was issued with a re-clearance and, whilst the Liverpool ATC report concluded that the use of Garston Docks VRP under these circumstances was inadvisable, Board Members were of the opinion that both pilots and ATC were operating correctly and that the incident was precipitated by the timing of TI to the A319 pilot, which unavoidably occurred after T/O and at a high workload phase of the flight. The Board noted that the PA38 pilot did not hold to the N of Garston Docks, as cleared, but opined that his position did not materially alter the situation. It was noted that separation minima do not apply between VFR and IFR traffic in Class D, deconfliction being achieved by the VFR pilot being visual with the IFR traffic and/or the IFR pilot requesting traffic avoidance, after the passing of TI to both flights. Both pilots were entitled airspace users, operating normally within the requirements of Class D airspace under VFR and IFR. The CAA SRG Advisor noted that this was a good example of Class D airspace being used as designed. The Board unanimously agreed that the ATC and pilot safety barriers functioned correctly and that normal procedures, safety standards and parameters applied. 5

69 PART C: ASSESSMENT OF CAUSE AND RISK Cause: The A319 crew was concerned by the presence of the PA38 to the north. Degree of Risk: E. ERC Score: 1. 6

70 AIRPROX REPORT No Date/Time: 28 Apr Z (Sunday) Position: 51286N 00943W (1.5nm S of Kenley Aerodrome) Airspace: (Class: G) (Class: G) Reporting Ac Reported Ac Type: Viking Glider GA7 Cougar Operator: HQ Air (Trg) Civ Pte Alt/FL: 1700ft ft (QFE 1001hPa) (QNH 1018hPa) Weather: VMC CAVOK VMC CAVOK Visibility: 40km 10nm Reported Separation: 10ft V/50ft H Recorded Separation: NR V/NR H <0.2nm H (before loss of radar contact) Viking Glider 1040:01 CPA 1040:56 0.2nm H NM :23 GA PART A: SUMMARY OF INFORMATION REPORTED TO UKAB THE VIKING PILOT reports flying a GH and instructor training sortie just to the S of Kenley in a white glider in contact with Kenley Radio. The Captain reports that he spotted the blue and white GA7 Cougar in their 9 o clock, slightly below them as the student instructor was conducting a nose high steep stall 8-10 secs before the CPA. He took control of the glider, turned L and descended with 80 L bank whilst attempting to maintain visual contact with the GA7 and present its pilot with a plan view of the glider. He was able to identify the other ac, read its registration and estimated the CPA to be 10ftV and 50ft H. He reports that the other ac did not appear to manoeuvre. He assessed the risk of collision as High. THE GA7 PILOT reports flying VFR from Rochester to Fairoaks via Sevenoaks Lakes and Ockham in his white ac squawking Mode 3/A 7000 with Mode S selected on. The pilot set his altimeter to 1018hPa at Rochester, cruised at 2000ft and selected Biggin APP frequency. He reports that Biggin APP was busy so he kept a listening watch. He recalls hearing traffic heading for Sevenoaks at 2000ft so he descended to 1800ft and set course for Ockham. He reports that his route took him to the S of Kenley and that he did not see any gliders. The pilot included his map in his report: 1

71 UKAB Note (1): ANALYSIS OF THE RADAR shows the GA7 tracking W abeam Sevenoaks squawking Mode 3/A 7000 and as it approaches Hurley Lodge the squawk changes to 7047, Biggin Hill Airport Conspicuity. At 1040:01 the radar recording shows the GA7 3.3nm SE of Kenley tracking 275 whilst the Viking is 0.9nm SSE of Kenley; the ac are 2.6nm apart. The Viking appears to manoeuvre L and then R without covering much track distance and then disappears from the radar recording at 1040:23, 1.3nm SSE of Kenley, in the GA7 s R 1 o clock position at a range of 1.9nm. The GA7 continues to maintain its track and at 1040:44 the Viking reappears in the GA7 s R 1 o clock position at a range of 0.5nm. The Viking paints 3 more returns which give the impression of a manoeuvre to the R through the GA7 s 12 o clock position and then disappears again 0.2nm ahead of the GA7. This point is the closest measurable CPA but the relative speeds of the ac mean that it is likely that the CPA was much closer as reported by the Viking pilot who was able to identify the GA7 s registration. PART B: SUMMARY OF THE BOARD S DISCUSSIONS Information available consisted of the reports from the pilots of both ac and radar video recordings. The GA and Gliding Board Members lead the discussion on this event; they were surprised that the GA7 pilot did not see any gliders as it is a busy gliding area at weekends. The Board debated the route chosen by the GA7 pilot and the manoeuvring area selected by the Viking crew and concluded that given the congested nature of the airspace they had both made reasonable choices. In Class G airspace both pilots had a responsibility to see and avoid the other ac; the Viking glider had right of way. Part of the Cause, therefore, was that the GA7 pilot did not see the Viking at any stage; for his part the Viking Captain had seen the GA7, but it was a late sighting at 8-10 seconds before the CPA, and the other part of the Cause. Turning to the Risk, the Board considered the measured and reported CPA distances. Although the Viking pilot s estimate of 10ftV/50ftH could not be confirmed by recorded radar, the Board was in no doubt that it was a close encounter. However, the Board concluded that once the Viking pilot had seen the other ac he had taken a sensible course of action that had a positive effect on the outcome and prevented a collision. Nevertheless the Board agreed unanimously that safety had been compromised and that this was a B Risk Airprox. The Board discussed the safety barriers which were relevant in this Airprox and concluded that they were aircrew rules and procedures, visual sighting and aircrew action. Neither ac was in receipt of an ATC service and, as the GA7 pilot had not seen the glider, he could not apply the Rules of the Air. Whilst the Viking pilot saw the GA7 and took appropriate action, it had been so late that this barrier was not fully effective. The Board concluded that the remaining safety barriers had been minimally effective resulting in a score of 20 on the ERC matrix. PART C: ASSESSMENT OF CAUSE AND RISK Cause: A non-sighting by the GA7 pilot and a late sighting by the Viking pilot. Degree of Risk: B ERC Score: 20 2

72 AIRPROX REPORT No Date/Time: 30 Apr Z Position: 5710N 00357W (Aviemore) Airspace: Scot FIR (Class: G) Reporting Ac Reported Ac Type: Tornado GR4 DG-808C Operator: HQ Air (Ops) Civ Pte Alt/FL: 250ft agl 3100ft RPS (1016hPa) QNH NR Weather: VMC CLBC VMC CLBC Visibility: 10km 50km Reported Separation: 100ft V/0nm H Recorded Separation: NK 300ft V/0nm H PART A: SUMMARY OF INFORMATION REPORTED TO UKAB THE TORNADO GR4 PILOT reports transiting at low-level as number 2 of a pair. The formation was operating autonomously under VFR in VMC, communicating on an inter-formation frequency and listening out on the low level common frequency [ MHz]. The grey camouflaged ac had the SSR transponder selected on with Modes A, C and S. The ac was not fitted with an ACAS. Whilst transiting along the W side of the River Spey valley, heading 045 at 420kt and 250ft AGL and keeping clear of the Highland Wildlife Park to the W, he saw a glider at an estimated range of 500m, slightly above and to the L, which had been hidden behind the canopy arch. He bunted and passed 100ft below it. The pilot noted that he had seen the glider at the very last moment. He assessed the risk of collision as Very High. THE DG-808C PILOT reports conducting a long flight with a reasonably high workload due to deteriorating soaring conditions. He was operating autonomously under VFR in VMC, listening out on the glider common frequency [ MHz]. The white self-launch glider was not fitted with lighting or an ACAS. The SSR transponder was selected off due to the high power drain over the extended flight and his location in class G airspace. Whilst in level flight, heading 065 at 70kt, he saw a Tornado very late in his R 3 o clock at a range of 800m, passing R to L. He achieved a limited pull up due to his low energy and passed an estimated 300ft directly over the top of the Tornado. He assessed the risk of collision as High. [UKAB Note(1): The DG-808C is a single seat, high performance, self-launch motor glider with a choice of wing-spans and a MAUW of 600kg on the 18m wing-span version. The ac has a V NE of 146kt, a typical V S of 37kt and a glide ratio of 1:50. 1

73 LOSSIEMOUTH OCCURRENCE SAFETY INVESTIGATION reports that a pair of Tornado ac was flying a routine formation training mission, conducting a low level route in LFA 14 on Tuesday 30th April Summary On the day in question, a formation of 2 x GR4 Tornado ac took off from RAF Lossiemouth at 1055 to conduct low level, air-to-air refuelling and air-to-ground weaponry exercises. During the planning phase the crews briefed a warning NOTAM in the Feshiebridge area, which their track was planned to cross; however, as this was a warning and not an 'avoid', the crews elected not to change the plan but to maintain a good look-out. The glider pilot planned a navigation exercise from Portmoak A/D in Kinross, returning to Portmoak via Dufftown, Maylodge, Loch Laggan and Rhynie. He took off just before 1000 on a flight that was to last 6hr 42min. The ac was a self-launch glider equipped with a stowable engine but following initial launch, the engine remained retracted for the entirety of the flight. The glider pilot s altitude was dictated by the prevailing soaring conditions which, at the time of the incident, had deteriorated. The Tornado sortie proceeded as planned with the formation remaining on the planned route. As the formation returned to base they were flying NE bound, in LFA 14 at 250ft agl, generally following the course of the River Spey. At approximately 1413 the pilot of the number 2 ac saw a glider at an estimated range of 500m, slightly above and slightly to the L, tracking E. The pilot bunted to avoid a mid-air collision. The glider pilot pulled up, gaining about 50ft in altitude as the Tornado passed under and adjacent to him coming within an estimated 100ft vertically. No other crew member in the formation saw the glider. Observations The incident occurred in Class G airspace in a notified LFA with consistently high levels of fast-jet operations. The crews were well briefed before the sortie, including the warning NOTAM at Feshiebridge, which was cancelled during the sortie, 45min before the incident. It is possible the Tornado's canopy arch obscured the glider until the last moment; however, it also presented as a small white ac on a white background. No specific flight safety recommendations can be made; both pilots were entitled airspace users in class G airspace where the principle of see and avoid is used to prevent collision. However, military crews must remain aware that light aircraft, including gliders, can and will operate in class G, from low level below radar coverage, to higher altitudes. ] 2

74 Recommendation The Tornado flight recording capability runs out after 3hr and the incident occurred 3hr 20min into the sortie, so no visual record of the event exists. It is recommended that the Tornado fleet software is upgraded such that sorties are visually recorded in their entirety, regardless of sortie length. HQ AIR (OPS) comments that given the relative speeds in this instance it is unlikely that the glider was hidden by the canopy arch but it may well have been obscured by other elements of the Head- Up Display. Given that the geometry appears to have been very close to a collision course, the glider would have presented an almost stationary visual contact. PART B: SUMMARY OF THE BOARD'S DISCUSSIONS Information available included reports from the pilots of both ac and a report from the appropriate operating authority. Board Members questioned whether this incident was unusual in that the glider was encountered at relatively low level, in a position and at a time that may not normally be anticipated, the majority of glider flying occurring at the weekend. However, Glider pilot Members pointed out that glider flying can take place throughout the year on every day of the week when there are suitable conditions and that it was entirely feasible to encounter a glider on a ridge line, relatively close to the ground, as pilots take advantage of orographic lift. In this case the glider pilot was 4hr 16min in to a 6hr 42min sortie, returning to Portmoak and covering some 360km in total. The Military pilot Member opined that the Tornado crews were probably concentrating their lookout towards the promulgated glider site at Feshiebridge and it was fortunate that the number 2 pilot saw the glider, albeit at the very last moment. The glider also represented a small target with low contrast against the sky. Members opined that the glider pilot saw the Tornado at about the same time and that both pilots took effective avoiding action. Both pilots were operating in class G airspace and had equal responsibility to see and avoid ; the glider pilot had right of way. In this scenario, Members opined that the only relevant safety barriers were visual lookout and pilot action; although both pilots eventually saw the other ac, the geometry of the encounter and the close miss distance persuaded Members that the barriers had been only minimally effective resulting in safety margins much reduced below normal and an ERC score of 20. PART C: ASSESSMENT OF CAUSE AND RISK Cause: A conflict in Class G airspace. Degree of Risk: B. ERC Score: 20. 3

75 AIRPROX REPORT No Date/Time: 01 May Z Position: 5405N 00039W 5nm SE Malton Airspace: Vale of York AIAA (Class: G) Reporting Ac Reported Ac Type: Typhoon FGR4 Tucano T1 Operator: HQ Air Ops HQ Air Trg Alt/FL: 9000ft FL80 RPS (1020hPa) Weather: VMC NR VMC CLAH Visibility: 40km 30km Reported Separation: 700ft V/0nm H Recorded Separation: NK V/0.2nm H 1000ft V/0nm H PART A: SUMMARY OF INFORMATION REPORTED TO UKAB THE TYPHOON PILOT reports leading a 3-ship formation, conducting visual Air Combat Manoeuvering (ACM) in the Vale of York AIAA. They were operating under VFR in VMC with a BS from LATCC(Mil), he thought. The ac was predominantly grey camouflaged, with a green, red and yellow painted fuselage spine and tail fin. Navigation lights and HISLs were selected on, as was the SSR transponder with Modes A and C. The ac was not fitted with an ACAS. After terminating the last training serial, and as the formation was rejoining into close formation, he saw a Tucano ac approximately level and 1nm directly ahead. He was heading S at 300kt and about 10000ft and assessed the Tucano to be tracking in a W ly direction. He called the formation to climb and gave TI to the rest of the formation who all quickly became visual with the Tucano. He was the closest member of the formation and assessed he passed 700ft directly above it. At no point prior to the incident had the formation received any information calls from London regarding this traffic. The Airprox was relayed to LATCC(Mil) via telephone on landing. He perceived the severity of the occurrence as High. THE TUCANO PILOT reports conducting an instructor check sortie in the Vale of York AIAA. She was sitting in the front, with the instructor student in the back, operating without an ATS under VFR in VMC. The black and yellow ac had landing lights, navigation lights and strobe lights selected on, as was the SSR transponder with Modes A and C. The ac was fitted with TCAS I. As she was setting up for a stall, she became aware of several TCAS contacts several miles to the NW of her position. She started to turn S, away from the contacts, when she became visual with 2 Typhoons flying in trail at approximately her level, which seemed to be conducting ACM. Due to their proximity and their continued manoeuvring, she elected to abandon the stall set up and manoeuvred to maintain visual contact. Shortly afterwards, turning through W at 100kt, the leading Typhoon pilot appeared to gain visual with her ac, positively changed heading and height, and passed about 1000ft above. She perceived the severity of the occurrence as Low. [UKAB Note(1): The LATCC(Mil) R/T transcript is reproduced below. The Typhoon formation was initially composed of 4 ac divided in pairs into 2 Flights, each with an individual C/S (labelled Flight1 1