Operational Performance and Capacity Assessment for Perth Airport

Transcription

1 Operational Performance and Capacity Assessment for Perth Airport Version July 2012 Prepared by: NATS Consultancy Page 1

2 The recipient of this material relies upon its content at their own risk, and it should be noted that the accuracy of the output modelling is directly linked to the accuracy of the supplied input data. Save where expressly agreed otherwise in writing and so far as is permitted by law, NATS disclaims all liability arising out of the use of this material by the recipient or any third party. Page 2

3 This page is intentionally blank Page 3

4 1 Document Control Version /07/2012 Operational Performance and Airport Page 4

5 2 Table of Contents Operational Performance and Airport 1 Document Control Table of Contents Executive Summary Detailed Summary Introduction Background Current Operational Performance Study Runway Operating Modes and Those To Be Assessed Use of Data and Observations Document Structure Operational Performance Metrics Definition of Terms Comparison Airports Traffic Schedule and Mix Arrivals Departures Observations ATC Observations of Airfield Operations Ground Infrastructure and Operations Data and Reporting Key Recommendations for Target Capacity Assessment Reduced Arrival-Arrival Separations Reduced Arrival-Departure-Arrival Separations Arrival-Departure Balancing Additional Exit Taxiways Current Capacity Index Introduction to HERMES Modelling Simulation Validation Current Capacity Index Assessment Current Capacity Index Summary Target Capacity Index Target Capacity Assumptions Target Capacity Index Assessment for Runways 21 and Target Capacity Assessment for Runway Target Capacity Index Summary Glossary Of Terms References Appendix A (Aerodrome Chart including Terminal WA) Appendix B (Daily Delay Profiles) Page 5

6 This page is intentionally blank Page 6

7 3 Executive Summary Operational Performance and Airport During February 2012 NATS Services visited Perth Airport to investigate opportunities to improve airport capacity as part of Airservices Australia s collaborative Airport Capacity Enhancement (ACE) programme. The NATS study provides the ACE programme with a comprehensive summary of the operational performance of the airfield together with an assessment of a capacity index based on the observed performance. Observations of the airfield have been made by NATS Air Traffic Control (ATC) experts and a number of recommendations identified to shape and prioritise future ACE initiatives. The potential benefits of these recommendations have been assessed and a target capacity index for Perth Airport established. The key success of the study has been the development of a baseline against which ACE initiatives and future capacity studies may be measured. In doing so, future decision making may now be informed by a quantitative assessment of operational benefits. The key conclusions of the study are described below and explored in detail in this report: Based on the observed performance of the airfield the current capacity index was determined to be 590 ATMs per day for Runway 21/24 operations and 524 ATMs per day for Runway 03 mixed-mode operations. The reaction time for pilots to respond to ATC runway line-up instructions was observed to be markedly higher than the benchmark set at similar high performing international airports. The time taken for aircraft to line-up for departure on Runway 21 was observed to be around 20% higher than the benchmarks set at similar high performing international airports. The runway occupancy of aircraft landing on Runway 03 was identified to be a considerable limiting factor to potential capacity enhancements in this runway mode. The positive application of ATC speed control during the approach phase of flight was observed to have the opportunity deliver to a higher capacity operation. There is an opportunity for Airservices Australia and Perth Airport to work together to establish a robust concept of operations, that incorporates both airfield and airspace, to support future infrastructure change programmes. Based on the recommended changes to the operation of the airfield the target capacity index was assessed at 696 ATMs per day for Runway 21/24 operations and 622 ATMs per day for Runway 03 mixed-mode operations. Page 7

8 3.1 Detailed Summary Operational Performance and Airport During the site visit key operational performance statistics were gathered both through direct observation and via electronic data capture. The NATS team also studied the ATC environment, procedures, and airport infrastructure with a view to identifying options to increase airport capacity. The NATS study is broken down into four stages: 1. Data collection and direct observations 2. Identification of the key operational performance statistics 3. Determination of the capacity index for the current operation 4. Determination of the near-term target capacity index following the release of any identified latent capacity Operations at Perth Airport are complex due to the intersecting runways and the corresponding taxiway and terminal infrastructure; however a number of recommendations have been identified which NATS believes can be implemented to deliver substantive benefits to the Perth Airport operation and enhance the customer experience. In assessing the capacity index under current operational performance an agreed delay criteria of no more than 10 minutes average delay for arrivals and departures in any given half-hour period was applied. This led to the following key results: The current capacity index for Runways 21/24 was estimated to deliver 590 ATMs per day; The current capacity index for Runway 03 was estimated to deliver 524 ATMs per day; The maximum achievable service rate using Runways 21/24 was estimated to be 42 movements. This was observed in the model in 3 separate hours; (UTC+8) where there is an arrival departure bias close to 50:50 and and where there is a bias towards departures; The maximum achievable service rate using Runway 03 was estimated to be 40 movements; During periods of high arrival demand, the maximum number of arrivals that can be handled in any given hour is 25 (based on the performance observed during the NATS visit); In the current capacity index profile for Runways 21/24 more than 30 ATMs are consistently handled each hour from (UTC+8), with 40 or more ATMs handled in six different hours across the day. In order to further enhance capacity, the NATS team has identified a number of key issues that are recommended to be addressed in support of capacity enhancement activities: Arrival runway occupancy time for Runway 03 was observed to be suboptimal for all aircraft types and could become capacity limiting in periods of high arrival demand; Page 8

9 Operational Performance and Airport Line-up times for Runway 21 were observed to be around 20% higher than similar benchmark airports; however plans exist to address this through the construction of a new holding point closer to the runway; Reduced arrival spacing was identified as a key opportunity to increase capacity and reduce delay in all runway configurations; Opportunities were identified to reduce average pilot reaction time through engagement with turboprop operators to deliver a more resilient high-intensity operation; Taxiway infrastructure was identified as having the potential to limit future airfield capacity, particularly upon completion of Terminal WA; It was noted that in general high quality electronic data is available for use in analysis; however significant benefits could be gained through the application of more automated data fusion and storage methods. This could facilitate more rapid analysis of airfield performance for the purposes of further optimising airfield procedures; The NATS Team also noted that there is not currently a joined-up approach to airfield master planning and future airspace development activities. It is recommended that a robust combined concept of operations, that incorporates both airfield and airspace, is developed before embarking on any future infrastructure change projects. The effect of implementing reduced arrival spacing and reduced runway occupancy has been simulated and the following key conclusions developed: The target capacity index for Runways 21/24 was estimated to deliver 696 ATMs per day; The preliminary target capacity index for Runway 03 (without reduced arrival runway occupancy) was estimated to deliver 578 ATMs per day; The target capacity index for Runway 03 was estimated to deliver 622 ATMs per day (this could be raised to 626 with improvements in pilot reaction time); The peak simulated achievable service rate for Runway 21/24 is 55 movements in an hour ( , UTC+8); The peak simulated achievable service rate for Runway 03 is 45 movements in a single hour ( , UTC+8) with reduced arot and 42 for the same hour without reduced arot; It should be noted that no consideration has been given to the capability of the airspace, terminal, or taxiway infrastructure to meet the levels of demand simulated here. Furthermore, the proportion of night-time operations has remained consistent with the current operation. Table 1 compares the average observed traffic profile with the current capacity index and the target capacity index for Runways 21/24. Table 2 compares the average observed traffic profile with the current capacity index and the target capacity index for Runway 03. These capacity indices represent the maximum traffic demand profile that could be contained with the 10 minute average delay criteria based on current procedures and following the implementation of NATS recommendations respectively. Page 9

10 Hour (UTC+8) Observed Current Capacity Index Target Capacity Index Arrivals Departures Combined Arrivals Departures Combined Arrivals Departures Combined Total Table 1. Comparison between the average hourly observed traffic profile, the current capacity index profile and the target capacity index profile for Runways 21/24. Page 10

11 Hour (UTC+8) Operational Performance and Airport Observed Current Capacity Index Preliminary Target Capacity Index Target Capacity Index Arrivals Departures Combined Arrivals Departures Combined Arrivals Departures Combined Arrivals Departures Combined Total Table 2. Comparison between the average hourly observed traffic profile, the current capacity index profile, the preliminary target capacity index and the target capacity index profile for Runway 03. Page 11

12 4 Introduction 4.1 Background Operational Performance and Airport Airservices Australia, in consultation with airline representatives and a number of airport operators, has initiated the Airport Capacity Enhancement (ACE) program to help address issues of delay and congestion at busy Australian Airports. This ACE programme is closely based upon a proven European process, which NATS has effectively managed at its busiest airports. Airservices contracted NATS to help support the Australian ACE programme, specifically to support Airservices Australia s ACE Near Term Benefits Strategy by providing an operational performance assessment of Perth Airport, identify potential quick wins and thereby assessing what potential latent airport capacity could be released in the near/short term. Perth Airport currently handles approximately 500 daily movements (busy day) which if sustained could potentially total approximately 180,000 annual movements. Forecast growth for the airport anticipates that total passenger movements could double by 2029 (from 2008 levels). Current master plans are being developed for additional taxiway and runway infrastructure. Until this additional runway and taxiway capacity is made available the existing demand and short term anticipated growth needs to be handled by the existing airfield system. It is therefore key to ensure that all aspects of airfield capacity are understood and any latent capacity released in order to support this increased demand Project Scope The scope agreed for the NATS study addresses the first three activities identified in the Airservices Australia s ACE Near Term Benefits Strategy as being critical to the release of latent airport capacity. In particular, these activities focus on the following: Undertaking a baseline Air Traffic Management (ATM) operational performance study at Melbourne, Perth, Brisbane and Sydney Airports; Analysing and reporting on current operational performance at these airports; Identifying achievable short to medium term capacity and performance targets. It is anticipated that the engagement of NATS to execute this will hasten the delivery of the Australian ACE Programme benefits by leveraging their considerable experience and expertise as an Air Navigation Service Provider (ANSP) and provider of consultancy services of relevance to this study. The study for Perth Airport will focus on delivering the following - 1. Current/Benchmark Operational Performance Analysis 2. Operational Performance Target Identification 4.2 Current Operational Performance Study In late February 2012 a team of Capacity Analysts and ATC Experts visited Perth Airport with the intention of collecting data from both electronic sources and Page 12

13 through manual observations. The operations at Perth Airport were observed over a period of 7 days, with flight data being collected over the same period. In total, 1,595 arrivals and 1,611 departures were recorded operating on Runways 03/21 and 06/24. NATS directly observed 954 flights (362 arrivals and 592 departures), 29% of the 3,206 total runway movements. The majority of these observations were made for the Runway 21/24 operations. These recorded and observed movements form the core data used in the analysis within this study. A summary of the study days is shown in Table 3. Study Days Total Number Of Arrivals Total Number of Departures 20/02/ /02/ /02/ /02/ /02/ /02/ /02/ Table 3. Summary of the Perth study period. During the visit, the NATS team viewed the current operation at the ATC facilities and established communication links with operational staff to begin the analysis of that operation. In addition they held a series of meetings with key stakeholders and subject matter experts including Airline representatives, Airservices Australia and Perth Airport. Subsequent to the visit further information has helped to ensure the accuracy of this report. 4.3 Runway Operating Modes Perth Airport operates a range of runway modes as follows; Single runway operations on Runways 03, 06, 21 or 24 Weather conditions dictate the use of a single runway for both arrivals and departures. The provision of departure opportunities leads to a reduction in the arrival rate, particularly during periods of low cloud & reduced visibility. Runway 03 for arriving traffic and Runways 03 & 06 for departing traffic, provides for maximum departure rate by using Runways 03 & 06 for departures & 03 for arrivals. Runway 21 for departing traffic and Runways 21 & 24 for arriving traffic, provides for maximum arrival rate with dependent use of Runways 21 and 24 for arrivals. Departures are processed from Runway 21. Runway 03 for arriving traffic and Runway 06 for departing traffic; apart from the occasional long haul departure off of Runway 03, arriving aircraft are unencumbered by departing aircraft, resulting in an improved arrival rate. The 2 modes to be assessed for this project are: a) Runway 21 for departures and Runways 21 and 24 for arrivals b) Single Runway 03 mixed mode The current agreed arrival rates for these runway modes are shown in Table 4. Page 13

14 Runway Conditions Flow Sequence Rate Time AVG ARR Rate 21/24 VMC /24 IMC-A /24 IMC-B VMC IMC-A IMC-B Table 4. Agreed arrival rates for different runway modes (Reference 1). The Time AVG values are supplied in minutes of separation between pairs of arriving flights (as described in the PH TCU Local Instructions, Reference 1) and the ARR Rate values are supplied as the corresponding number of arrivals in a one hour period. Note therefore that under typical arrival departure arrival operations a rate of 24 would be expected to deliver a total of 48 runway movements 24 arrivals and 24 departures. 4.4 Use of Data and Observations Electronic data has been used extensively throughout the NATS study to understand the operation of the airport and surrounding airspace, as well as to derive operational performance statistics. Such data includes NOC KPI reports, schedule data, and EUROCAT radar data. Table 5 shows a breakdown of the number of flights in each runway mode for each of the study days, where all the sources of electronic data were available. Since an explicit record of the runway mode in operation was not available the runway modes have been estimated based upon periods of consistent delivery of arriving and departing aircraft to the runways as detailed in the individual modes. Runway Mode 03 Mixed 03A/03D/06D 06 Mixed 21 Mixed 21A/24A/21D Date Grand Total 20/02/ /02/ /02/ /02/ /02/ /02/ /02/2012 Combined Arrivals Departures Combined Arrivals Departures Combined Arrivals 5 5 Departures 6 6 Combined Arrivals Departures Combined Arrivals Departures Table 5. Summary of runway modes in operation on the Perth study days. Whilst this electronic data has formed the basis of most of the analysis in the study, a number of key performance parameters could not be assessed without direct observation. In these cases the parameters are estimated from the flights directly observed during the NATS visit. During this time, 83% of flights were observed during Runway 21 and 24 modes of operation and 17% were observed during Runway 03 and 06 modes of operation. Page 14

15 4.5 Document Structure Operational Performance and Airport The remaining sections of this report detail the NATS findings and recommendations that have arisen during the course of the site visit and subsequent investigations. Section 5 provides the Operational Performance Summary and key metrics. Section 6 provides NATS observations and ATC findings. Section 7 summarises the recommendations made. Section 8 covers an analysis of current capacity index based on the observed operational performance. Section 9 provides the target capacity assessment. Page 15

16 5 Operational Performance Metrics 5.1 Definition of Terms Airport metrics and particularly airport related delays can be defined in many different ways; this section provides the NATS definition of the terms used in this project these are the same as those that are routinely used in the UK for all NATS runway capacity assessments Arrivals Arrival Schedule Demand The hourly (or half hourly) arrival runway demand based on the planned gate arrival time as notified in the daily schedule. Arrival Utilisation The number of aircraft with a cross threshold time (collected from observations and runway log data). This is the time at which an arriving aircraft passes over the start of the runway - typically reported by hour. Cordon Time The time at which an arriving aircraft crosses the circular cordon which is established at a specific range around the airport. Aircraft are expected to absorb any delay associated with the runway inside this cordon. In the case of Brisbane, this was agreed as 90nm. Undelayed Flying Time The time it takes an aircraft which is not delayed to fly from the cordon to runway threshold. Threshold Demand Time The time at which an aircraft is expected to cross the runway threshold if it experiences no delay. This is calculated by adding the undelayed flying time to the cordon time. Arrival-Arrival (AA) Spacing The difference between the times at which two consecutive queued arriving aircraft are observed to cross the runway threshold (In the case of Brisbane this does not account for the displaced threshold on Runway 01). Arrival-Departure-Arrival (ADA) Spacing The difference between the times at which two arriving aircraft are observed to cross the runway threshold where there is a single departure interleaved between the two (In the case of Brisbane this does not account for the displaced threshold on Runway 01). Arrival Runway Occupancy (arot) The time interval between an arriving aircraft crossing the runway threshold and exiting the runway (In the case of Brisbane this does not account for the displaced threshold on Runway 01). Page 16

17 Arrival Delay Operational Performance and Airport The difference in time between the actual cross-threshold time and the threshold demand time. During this study average arrival delay is used as a measure of performance and is defined as the arithmetic mean across all movements in a defined time period (i.e. the sum of arrival delay during a period of time divided by the number of arrival movements in that time). Average arrival delay is typically reported for half-hour, hour, standard operating day (17 or 18 hours), or 24-hour periods Departures Departure Schedule Demand The hourly (or half hourly) departure runway demand based on the planned gate departure time as notified in the daily schedule. Departure Utilisation The number of aircraft with an airborne time (collected from observations and runway log data) typically reported by hour. Departure Line-up Time The time an aircraft takes to manoeuvre from the holding point to become lined up on the runway. Pilot Reaction Time (PRT) The time between an aircraft being cleared for take-off and starting its take-off roll. Departure Runway Occupancy Time (drot) The time between an aircraft being cleared for take-off and becoming airborne. Departure-Departure (DD) Separations The time interval between a departure immediately followed by another departure on the same runway (i.e. airborne time to airborne time for queued traffic). Departure Delay The length of time an aircraft has to wait at the holding point for access to the runway. This may be either directly observed or estimated based on the time the aircraft leaves the stand and the time the aircraft takesoff. During this study average departure delay is used as a measure of performance and is defined as the arithmetic mean across all movements in a defined time period (i.e. the sum of departure delay during a period of time divided by the number of departure movements in that time). Average departure delay is typically reported for half-hour, hour, standard operating day (17 or 18 hours), or 24-hour periods. Page 17

18 5.1.3 Capacity Capacity Index The runway demand profile that provides the maximum runway utilisation over a given period of time without exceeding a specified delay criteria or delivering an imbalance in the overall number of arrivals and departures during that period of time. The capacity index is expressed as the number of arrival and departure movements in each hour over the given period of time. Current Capacity Index The simulated capacity index obtained using the observed operational performance parameters. Target Capacity Index Baseline The simulated capacity index obtained using the proposed changes to the operation. For the purposes of assessing the current and target capacity indices, a baseline simulation has been developed using the observed operational performance and traffic demand. The baseline simulation aims to replicate the performance observed during the NATS study period and is not intended to provide a benchmark for current performance. 5.2 Comparison Airports For the purpose of providing a comparative context of key airport performance metrics Airports A and B are defined here and used throughout this document. Note it is always very difficult to compare the operation of different airports as each has its own specific infrastructure and procedural functions. Additionally Perth Airport has a challenging mix of aircraft types that is not fully reflected in the comparison airports. Airports A and B should be considered as a reference guide only Airport A Airport A has a high intensity single runway operation that handles up to 800 ATMs during a 17 hour period. The wake turbulence mix is predominantly medium (~80%) and heavy (~10%). Declared capacity is profiled throughout the day and when the demand is present Airport A regularly delivers in excess of 50 ATMs per hour Airport B Airport B has a busy single runway operation that handles up to 580 ATMs during a 17 hour period. The wake turbulence mix is predominantly medium (~90%). The schedule follows a series of cyclical peaks throughout the day and declared capacity is profiled. During peak periods Airport B can deliver in excess of 40 ATMs per hour. Page 18

19 5.3 Traffic Schedule and Mix Typical Schedule Profile The daily average schedule demand for the 7 study days was 444 movements. The daily average utilisation for the 7 study days was 454 movements. The hourly average demand and utilisation profile is shown in Figure 1. Whilst there is some offset between the scheduled gate demand and the actual runway utilisation, it should be noted that the maximum average achieved hourly throughput that was observed was 37 movements. This occurred during the busy (UTC+8) hour which is departure biased and is broadly consistent with the average DD separation of 89 seconds (discussed later in Section 5.5.4). However, this falls someway short of the maximum scheduled demand of 42 movements which occurs in the same hour. Figure 1. Hourly average schedule demand and utilisation Wake Turbulence and Carrier Breakdown Figure 2 shows that medium wake turbulence category aircraft account for the largest proportion of traffic at Perth Airport with 77% of the movements falling into this category. Page 19

20 Figure 2. Aircraft wake turbulence split at Perth Airport. Qantas (QFA) services account for approximately 29% of movements at Perth; the wake turbulence mix for their operations is shown in Figure 3. Figure 3. Aircraft wake turbulence split for Qantas (QFA). Virgin Australia (VOZ) services account for approximately 11% of movements at Perth Airport; the wake turbulence mix for their operations is shown in Figure 4. Page 20

21 Figure 4. Aircraft wake turbulence split for Virgin Australia (VOZ). Skywest Airlines (OZW) services account for approximately 10% of movements at Perth Airport. 100% of the aircraft operated by Skywest Airlines at Perth Airport are medium wake turbulence category aircraft. Figure 5 below shows the Wake Turbulence split across the different airline operators at Perth Airport. Figure 5. Aircraft wake turbulence split by airline operator. Page 21

22 5.4 Arrivals The key runway performance statistics for arriving aircraft at Perth are highlighted in Table 6 alongside those for Airports A and B for comparison. Perth Airport A Airport B Average arot, secs Average AA Separation, secs Average ADA Separation, secs Average ADDA Separation, secs Average Arrival Delay, Mins Table 6. Key arrival performance statistics with comparison airports. The wake turbulence profile for arrivals based on the average observed runway utilisation is shown in Figure 6. Figure 6. Average number of hourly arrival movements by wake turbulence category Achieved Arrival Rate The daily average arrival schedule demand for the 7 study days was 227 movements. The daily average arrival utilisation for the 7 study days was 228 movements. Figure 7 shows the peak arrival hour was during (UTC+8) with an average of 21 movements. Page 22

23 Figure 7. Hourly average arrival schedule demand and utilisation Arrival Runway Occupancy Times (arot) Average Arrival Runway Occupancy (arot) performance statistics are calculated from observed data. Arrival data was observed during Runway 03, Runway 21 and Runway 24 operations Runway 03 Table 7 shows that the average arrival runway occupancy for Runway 03 is 68.8 seconds. Runway Exit Approximate Distance From Threshold, m Average arot, secs Utilisation K % J % Rwy 06/ % D % W % Total 68.8 Table 7. Average arrival runway occupancy time & exit distance from threshold for Runway 03. Figure 8 shows that the arot distribution for Runway 03 broadly resembles a normal distribution around the overall mean arot of 68.8 seconds. Page 23

24 Figure 8. Arrival runway occupancy distribution by exit (Runway 03). Figure 9 shows the wake turbulence usage of runway exits for the two most commonly used exits. Figure 9. Wake turbulence proportion by runway exit (Runway 03). Figure 10 and Figure 11 show the arot split by operator for the most commonly used exits D1 and J2. Note that these charts are based on low sample sizes, but are shown for information. Page 24

25 Dominant operators VOZ and QFA have arot values below average for exit D1. Skywest Airlines average arot is above average for this exit. Figure 10. Average arrival runway occupancy by operator for exit D1 (Runway 03) Qantas and Virgin Australia also have below average arot values for aircraft vacating Runway 03 at exit J2. A total of 7 Skywest Airlines aircraft were observed using this exit, with the average arot value coming out above the overall average for the exit. Figure 11. Average arrival runway occupancy by operator for exit J2 (Runway 03). Figure 12 and Figure 13 show the average arot times for each exit, split by aircraft type. As with Figure 10 and Figure 11, counts are low. Page 25

26 The average arot values for aircraft types using D1 range from 65 seconds to 83 seconds. Note that the largest arot is for the B461, of which there was only one observed flight vacating at D1. The most common aircraft type is the B738, which has a below average arot value. Figure 12. Average arrival runway occupancy by aircraft for exit D1 (Runway 03) The most common aircraft type to vacate Runway 03 at J2 is the F100, which as shown in Figure 13 has an above average arot value. Figure 13. Average Arrival Runway Occupancy by Aircraft for Exit J2 (Runway 03) Page 26

27 Runway 21 Operational Performance and Airport Table 8 shows that the average arrival runway occupancy time for Runway 21 is 53.0 seconds. Runway Exit Approximate Distance From Threshold, m Average arot, secs Utilisation D % Rwy 06/ % J % K % L % Total 53.0 Table 8. Average arrival runway occupancy time & exit distance from threshold for Runway 21. Figure 14 shows the distribution of arot values on Runway 21 largely reflects a normal distribution around the overall mean arot value of 53.0 seconds. The second peak that occurs when arot values exceed 65 seconds is due to the higher proportion of heavy aircraft vacating at K and L1 as shown in Figure 15. Figure 14. Arrival runway occupancy distribution by exit (Runway 21). It is evident from Figure 15 that as distance from the runway threshold increases there is a higher proportion of heavy wake turbulence aircraft vacating the runway. At exit L1, 55% of aircraft are in the heavy wake turbulence category. Page 27

28 Figure 15. Wake turbulence proportion by runway exit (Runway 21). Figure 16 shows the arot split by operator for the most utilised exit J2. The main user of this exit was Skywest Airlines, whose average arot value is approximately 2.5 seconds above average. The average arot of Virgin Australia is also slightly above average, whilst that of Qantas is below average. Note that counts for JST, KNX and RAM are low. Figure 16. Average arrival runway occupancy by operator for exit J2 (Runway 21). Figure 17 shows the average arot values for aircraft vacating at K. 75% of aircraft vacating here were operated by Qantas. All other operators had a low sample size. Page 28

29 Figure 17. Average arrival runway occupancy by operator for Exit K (Runway 21). Figure 18 shows average arot values for aircraft vacating at L1. Note that the majority of the figures represent low counts particularly Thai Airlines, China Southern and Tiger Airways, each of which were only observed once. Figure 18. Average arrival runway occupancy by operator for exit L1 (Runway 21). Figure 19, Figure 20 and Figure 21 show the average arot values for different aircraft types for each of the 3 main exits; J2, k, L1. The counts for each individual aircraft type are comparatively low for most types, but are provided for information. Page 29

30 Figure 19 shows that average arot values for aircraft types using J2 range from 36 seconds to 54 seconds. Note - low counts were recorded for the following aircraft types: A320, B190, B461, B712, BE20, DH8A and SW4. Figure 19. Average arrival runway occupancy time by aircraft type for exit J2 (Runway 21). Figure 20 shows a wide spread of arot values for exit K on Runway 21. There was only one observed turboprop utilising this exit; the DH8D with the largest arot. Figure 20. Average arrival runway occupancy time by aircraft type for Exit K (Runway 21). Page 30

31 Figure 21 shows that average arot values for exit L1 range from 59 seconds to 81 seconds, with the most common aircraft types being the A333 and B738 whose arot values are close to the average of 72.5 seconds. Figure 21. Average arrival runway occupancy time by aircraft type for exit L1 (Runway 21) Runway 24 Table 9 shows that the average arrival runway occupancy for Runway 24 is 53.3 seconds. Note that during the observations, one example of a heavy aircraft was observed landing and having to backtrack this single time which was recorded as 162 seconds is not included below. Runway Exit Distance From Threshold, m Average arot, secs Utilisation C % Rwy 03/ % A % J % J % V % Total 53.3 Table 9. Average arrival runway occupancy time & exit distance from threshold (Runway 24). Figure 22 shows the distribution of arot values on Runway 24 largely reflects a normal distribution around the mean arot value of 53.3 seconds. Page 31

32 Figure 22. Arrival runway occupancy distribution by exit (Runway 24). Figure 23 shows the breakdown of wake turbulence types using each of the most utilised exits A and J1. Figure 23. Wake turbulence proportion by runway exit (Runway 24). Figure 24 and Figure 25 show the average arot values for aircraft using exiting Runway 24 at the main exits A and J1, split by operator. Counts for most operators are comparatively low, with the exception of Qantas, Jetstar Airways, Virgin Australia and Skywest Airlines. Page 32

33 Figure 24 shows the arot split by operator for exit A. Dominant operators Virgin Australia and Qantas have arot values below average for exit A. Figure 24. Average arrival runway occupancy by operator for exit A (Runway 24). Figure 25 shows average arot values for aircraft vacating at J1. Dominant Operators Virgin Australia and Skywest Airlines are both above average at this exit. Jetstar and Qantas Airlines are both below average for this exit. Figure 25. Average arrival runway occupancy by operator for exit J1 (Runway 24). Figure 26 and Figure 27 show the average arot values for each aircraft type vacating Runway 24 at the main exits A and J1, split by aircraft type. Note that Page 33

34 the counts for individual aircraft types are low in most cases, but are provided for information. Figure 26 shows that average arot values for aircraft types using exit A range from 43 seconds to 66 seconds. Figure 26. Average arrival runway occupancy by aircraft type for exit A (Runway 24). Figure 27 shows that average arot values for aircraft types using exit J1 range from 43 to 69 seconds. Figure 27. Average arrival runway occupancy by aircraft type for exit J1 (Runway 24). Page 34

35 5.4.3 Arrival-Arrival (AA) Separations The distribution of Arrival-Arrival (AA) separations for Perth Airport, when operating on runways 03 is shown in Figure 28. The distribution for operations on runways 21 and 24 is shown in Figure 29. When operating the runway mode 21A/24A/21D four types of AA separations are possible, based on the runways used by the leading and following aircraft. Figure 28. Arrival-arrival separation distribution (Runway 03). Figure 29. Arrival-arrival separation distribution (Runways 21 and 24). Page 35

36 From analysis of the distribution and considering the published flow rates; values above 275 seconds will be excluded from statistics used for configuring the simulations. Based on this analysis, the average AA separation observed at Perth is 149 seconds. AA Separations accounted for 48% of observed arrival separations Arrival Departure Arrivals (ADA) Separations The distribution of ADA separations for Perth Airport, when operating on Runways 03 and 06, is shown in Figure 30. This covers the two runway modes 03A/03D and 03A/03D/06D. In these modes there are 2 possible ADA separations, depending on whether the departure operates off of Runway 03 or Runway 06. The distribution of ADA separations when operating on Runways 21 and 24 is shown in Figure 31. Here there are 4 possible ADA separations as shown in the graph. Values above 275 seconds have again been excluded so as to ensure the sample contains only queued arrivals. After these considerations the average ADA Separation is 179 seconds. ADA Separations account for 40% of the observed arrival separations Proportion, % A-03D-03A 03A-06D-03A Figure 30. Arrival-departure-arrival separation distribution (Runways 03 and 06). Page 36

37 Figure 31. Arrival-departure-arrival separation distribution (Runways 21 and 24) Arrival-Departure-Departure-Arrival (ADDA) Separations In addition, 40 instances of arrival-departure-departure-arrival (ADDA) spacing were observed, accounting for around 12% of all observed arrival spacing. Of these only 3 were observed on Runways 03 and 06. The distribution for Runway 21 and 24 operations is shown in Figure 32. Page 37

38 Figure 32. Arrival-departure-departure-arrival spacing distribution (Runways 21/24). Analysis of the distribution and consideration of the calculated ADA and DD separations suggest that ADDA separations over 425 seconds should be excluded to ensure that operations are pressured. With this consideration the average ADDA separation at Perth is 290 seconds. A summary of the observed arrival spacing at Perth Airport is shown in Table 10. Method Of Spacing Average Observed Spacing (Seconds) Proportion of Observed Operations (%) Arrival-Arrival Arrival-Departure-Arrival Arrival-Departure-Departure-Arrival Table 10. Summary of arrival spacing observations Arrival Undelayed Flying Times The undelayed flying time of arriving aircraft is the time difference between the aircraft passing a nominal cordon point set up around the airfield and the cross runway threshold time. At Perth Airport a cordon distance of 200nm was established in order to capture runway delay absorbed by arriving aircraft whilst airborne. The Perth cordon is shown schematically in Figure 33. For the purpose of analysing the performance of the airfield, the cordon around the airfield is divided into 12 sectors. This allows the flying time to be split by different cordon entry points. For example when Perth Airport is operating on Runway 03, an aircraft arriving from the North would be expected to have a longer flying time than an aircraft arriving from the South East. Standard flying Page 38

39 times for each sector have been calculated separately for jet and turboprop aircraft. Figure 33. Perth Airport area showing 200nm cordon and key arrival routes. Note that over 500 turboprop movements were first observed in the radar data inside the 200nm cordon and therefore could not be used within undelayed flying time or delay calculations. The distribution of flying times observed within the 200nm cordon is shown in Figure 34. Page 39

40 Figure 34. Arrival undelayed flying time distribution. Table 11 shows the undelayed flying times used for each sector and runway configuration. Jet/Prop Jet Prop Runway 03 Undelayed Flying Sector Count Time (Mins) Runway 06 Runway 21 Runway 24 Undelayed Undelayed Undelayed Count Flying Flying Flying Count Count Time Time Time (Mins) (Mins) (Mins) * * * * * * * * * * * Table 11 Standard undelayed flying times for Perth Airport, split by runway and by entry sector. (*estimated undelayed flying time due to low traffic count size). Page 40

41 Analysing the distribution of flying times highlights where there may be airborne congestion due to the runway capacity. The undelayed flying time is the time at which the cumulative flying time function is observed to rise steeply. The undelayed flying time is added to the time at which the aircraft passed the 200nm cordon boundary. This produces an estimated cross runway threshold time for the aircraft. Arrival delay is then calculated as the difference between the aircraft s actual threshold time and when the aircraft was estimated to cross the runway threshold Calculated Arrival Delay Figure 35 shows the average arrival delay (blue line) together with the hourly average arrival utilisation and average arrival schedule demand. The average arrival delay was 3.67 minutes (across the 13-hour period , UTC+8), with a peak hourly average delay of 9.5 minutes during the 1900 (UTC+8) hour. During the observation period, the peak arrival delay recorded for an individual flight was 34 minutes. Figure 35. Average scheduled arrival demand, actual utilisation and calculated delay. It is to be noted that in the current and target capacity index assessments (Sections 8 and 9), departure delay is considered over an 18 hour period (UTC+8). For comparison, the average observed departure delay over the equivalent 18-hour period was 3.29 minutes. Delay profiles for each study day are consolidated for reference in Appendix B. Page 41

42 5.5 Departures A summary of Perth Airport departure statistics and a comparison with reference Airport A and Airport B is shown in Table 12. Perth Airport A Airport B Average Line Up Time, secs Average drot, secs Average PRT, secs Average DD Separation, secs Average Departure Delay, Mins Table 12. Perth departure statistics. The wake turbulence profile for departures based on the average observed runway utilisation is shown in Figure 36. Figure 36. Average number of hourly departure movements by wake turbulence category Achieved Departure Rate The daily average departure schedule demand for the 7 study days was 217 movements. The daily average departure utilisation for the 7 study days was 230 movements. Page 42

43 Figure 37. Hourly average departure schedule demand and utilisation. Figure 37 shows that demand is high in the hour, with average scheduled demand being 40 movements. Utilisation is also at its highest during this hour, reaching an average of approximately 34 departures Departure Runway Occupancy & Line Up Time Table 13 summarises the observed line up times and departure runway occupancy times, split first by runway and then by holding point. The notation A-V denotes aircraft that have entered runway 06 at A, then backtracked to V for departure. This is due to restrictions on aircraft size on taxiways in the vicinity of taxiway V. Runway Holding Point Average Line Up Time, secs Average drot, secs Utilisation M % L % 03 L % K % Rwy 03 Overall A-V % V % Rwy 06 Overall W % B % D % J % Rwy 21 Overall Table 13. Average departure runway occupancy time & line up time split by runway, then by holding point Page 43

44 The distribution of these drot values is shown by entry point in Figure 38. Figure 38. Departure runway occupancy time distribution (all runways). Similarly, the distribution of line up times is shown in Figure 39. Figure 39. Departure line up time distribution (all runways). Page 44

45 5.5.3 Pilot Reaction Time Operational Performance and Airport Average Pilot Reaction Time (PRT) at Perth is 11.7 seconds. This is shown, together with the average figures for each operator in Figure 40 and for each aircraft type in Figure 41. Figure 40. Average pilot reaction time split by operator. Figure 41. Average pilot reaction time split by aircraft type. Page 45

46 Of the dominant operators, Skywest Airlines, Virgin Australia and Qantas all have average pilot reaction times lower than the overall Perth average. Note the following operators and aircraft types were observed infrequently ANZ, AWQ, LKF, MAS, PBN, RON, SIA, GAO, TFX and TGW; B772, C404, C680, D328, PA31, PAY3, B462, C550, H25B, B461, B733 and B73Y Departure-Departure (DD) Separations The distribution of Departure-Departure separations when operating on Runway 03 only or Runway 03 and Runway 06 is shown in Figure 42. The mode 03A/03D/06D allows for departures off both Runway 03 and Runway 06, so there are 2 types of possible DD separations in this mode. Figure 43 shows the distribution of DD separations on Runway 21. A cap of 150 seconds has been placed on DD separations so that only queued departure pairs are included in the Perth Airport capacity index simulation. As a result the average DD Separation at Perth is 89 seconds. Figure 42. Departure-departure separation distribution (Runways 03 and 06). Page 46

47 Figure 43. Departure-departure separation distribution (Runway 21) Calculated Departure Delay The average departure delay is shown together with the average departure utilisation and average departure schedule demand in Figure 44. The average departure delay observed during the study period was 1.86 minutes (across the 18-hour period , UTC+8), with a peak hourly average delay of 6.5 minutes during the 1500 (UTC+8) hour. During the observation period the highest departure delay recorded for an individual flight was 18.5 minutes. Page 47

48 Figure 44. Average departure scheduled demand, actual utilisation, and calculated delay. Departure delay has been calculated for all flights however the methods vary depending on if the flight has been directly observed. For observed departing flights, a range of key timing occurrences are recorded up until the aircraft becomes airborne. These include join holding point time, line up time, clear take-off time and airborne time. The calculated departure delay is the duration that the aircraft is observed to spend at or in queue at the runway holding point. Departure delay for flights that have not been directly observed is calculated by using an estimated undelayed airborne time. The estimated undelayed airborne time is calculated by adding a standard taxi time to an aircraft s gate departure time. The standard taxi time is calculated from observed data and consists of aircraft manoeuvring time to reaching the runway holding point, line up time and departure runway occupancy time. The calculated departure delay is the difference between the estimated undelayed airborne time and the aircraft s actual airborne time and is an estimate of the time an aircraft spent at the holding point. The delay profiles for each day observed are consolidated for reference in Appendix B. Page 48

49 6 Observations Operational Performance and Airport 6.1 ATC Observations of Airfield Operations Runway Occupancy arot Runway 03 Average observed arot (at 68.8 seconds) on Runway 03 is sub-optimal for all aircraft types. Three main factors appear to influence arot on Runway 03: a) No RETs off the runway; most exits are set close to 90 degrees, with the exception of Juliet 2, Papa and November which are angled off the runway but not at a typical RET design angle; b) Crew s desire to vacate at an exit close to their parking position in order to minimise taxi time; c) ATC s requirement to de-conflict arriving aircraft with aircraft taxiing on Alpha for departure on Runway 03, or those waiting to depart from Runway 06. arot Runway 21 Average observed arot (53.0 seconds) is close to a good arot for all aircraft, however, the arot for aircraft vacating at Kilo (57.0 seconds) and Lima (72.5 seconds) is still sub-optimal. The completion of taxiways Victor, Charlie 6 and Alpha 6, currently under construction, will potentially improve arot for medium traffic vacating the runway, but it is expected that the new exit will be set at a sub-optimal distance for some heavy jets. arot Runway 24 Average observed arot (53.3 seconds) is close to a good arot for most aircraft, with the exception of those exiting at Victor and heavy aircraft requiring a backtrack after landing. There is anecdotal evidence to suggest that aircraft brake to vacate at Alpha to minimise taxi time to their parking gate; consequently they are slow to cross Runway 03/21 during their landing roll which could impact on ADA times for cross runway operations (arrivals on 24, departures on 21). Braking to vacate at Juliet or Victor would likely improve the ADA time but could potentially increase congestion near the 900s stands area. The impact of this could be mitigated by ATC passing the desired exit point to inbound aircraft in order to tactically manage the balance between minimal ADA times and taxiway congestion. Landing Distance to Runway Exit Feedback from flight crew indicates that promulgated landing distances to each exit would aid in planning their exit point prior to landing, based on aircraft weight and weather conditions, with a potential improvement in arot. RETs The inclusion of RETs on 03/21 would improve arot in certain instances but careful positioning would be required for those vacating onto Alpha to minimise taxiway congestion and complexity. While optimal distances are relatively straightforward to define, a separate activity would be needed to assess the Page 49

50 different options considering the complexity of the existing taxiway infrastructure and operation. In event of additional RETs being put in place on Runway 03/21, and in addition to providing aircraft operators at Perth with ROT performance information, there would be benefit in publishing an agreed preferred Runway Exit Taxiway (by type e.g. Medium Jet, Heavy Jet etc.) together with information on distance of the RET (from threshold to breakaway point of lead off paint line) and exit design speed in Perth Airport Section of AIP, together with any other High Intensity Runway Operation (HIRO) information (see UK AIP, AD 2-EGKK-1-10, h. Minimum Runway Occupancy Time ). LUTs Runway 21 LUTs from Bravo (55.9 seconds) and Whiskey (50.1 seconds) are greater than typical benchmark values including those of example Airport A (45 seconds) and Airport B (48 seconds). It is understood that as part of a future planned extension to taxiway Alpha, a new holding point will be constructed closer to the runway than existing Bravo & Whiskey; this should have the advantage of a reduced LUT for aircraft departing from this area. PRTs The results for PRT are sub-optimal and demonstrate a significant spread across aircraft operators and types. Responses from the Pilot forum suggested that the increased PRT for turboprop operators were due to additional take-off checks required, compared to jet operators, following the receipt of a take-off clearance. Figure 40 and Figure 41 provide some evidence to corroborate this perception. It is recommended that the turboprop community be engaged and encourage to amend their SOPs in order to minimise PRT and improve ATC confidence Arrival Spacing Arrival spacing is the main factor in runway capacity of any airport, as the ADC Controller only has the option of inter-leaving one or more departures between successive arrivals. If this arrival spacing is too small then departure movements will be missed (instead of ADA, the sequence becomes AA). Similarly, movements will be lost if spacing is too large, but not large enough for an additional departure (instead of ADDA, the sequence becomes ADA). Discussions with TWR and TCU controllers, together with analysis of radar data indicate inconsistencies in the delivery of final approach spacing, both in miles between aircraft and their relative speeds. Consistent delivery of arrival gaps is key to maximising runway throughput and would ensure a degree of confidence in the TWR controllers that each and every gap may be utilised. By analysing the distribution of observed arrival spacing (during queued traffic), it can be seen that: Mean ADA spacing was observed to be 179 seconds; Mean AA spacing was observed to be 149 seconds. Current flow rates for different runway configurations and weather conditions are detailed in PH TCU and PH TWR Coordination and Standard Operating Procedures, Version 18: 02 Feb 2012, LOA_315. Within this LOA there are two Page 50

51 tables that are used independently to determine the arrival rate and final approach spacing, these are as follows: a) Page 15: Arriving Aircraft Spacing. This table details the final approach spacing that the TCU is required to deliver based on runway in use and time of day. The spacing is set at either 4 or 5nm depending on these conditions. The LOA allows for a reduction in these spacing s in certain circumstances, but not for the purposes of increasing arrival rate during periods of low departure demand. b) Page 23: Acceptance Rates. This table details the acceptance rates at the Feeder Fix that the TCU Flow Controller uses depending on different runway configurations and weather conditions. Both of the above tables are used independently and do not complement each other nor provide comparable results. For example, during Runway 21 operations in IMC-A conditions during a weekday, the Arriving Aircraft Spacing table requires that approach deliver 5nm spacing between subsequent arrivals. However, the Acceptance Rates table details an acceptance rate of 22/hour. The rates and spacings detailed in the above tables are unnecessarily restrictive and do not cater for differing departure/arrival demand bias or the presence of strong headwinds. There is no evidence of a consistent dialogue between the tower and approach supervisors (or equivalent), or between ADC and APC when determining final approach spacing. The introduction of the Arriving Aircraft Spacing table has been beneficial in ensuring that approach provides the tower with departure gaps, but it also removes the onus on controllers and supervisors to tactically review departure/arrival demand. Due to the location of the Feeder Fixes it is acknowledged that a short notice change to the final approach spacing and acceptance rates is difficult to achieve. The minimum rates detailed in the LOA could further be reduced to achieve minimum radar or wake turbulence spacing as required. The difference in VMC and IMC rates is driven by local experience and the perception that aircraft performance differs significantly in these conditions. Discussion with local operators of jet traffic indicate that intermediate and final approach speeds do not necessarily differ during these different weather conditions, pointing to the potential issues lying with non-jet operators. The use of visual approaches was also observed on a regular basis, although a consistent answer for their frequent use was not forthcoming. A visual approach does require less HMI input and therefore a reduced workload for the approach controller, however, it also makes consistent final approach spacing harder to achieve, as the aircraft is no longer required to fly at a particular speed. The following recommendations are made: a) Both tables are combined in order to provide a consistent approach in determining arrival rates and spacing; b) Acceptance rates are determined by minimum required final approach spacing and not by cloud and visibility conditions (except where low visibilities conditions apply); c) A minimum final approach spacing is promulgated (e.g. 3nm) for all runways, which should be the aspiration of all controllers to achieve when there is little or no departure demand; Page 51

52 d) Tower and the TCU constantly review the balance of departure/arrival demand using locally available information such as COBTs (2130 to 0030 UTC) and CTMS (METRON when operational) and to ensure that acceptance rates are optimised in sufficient time prior to aircraft arriving at the feeder fix; e) An on-going analysis of arrival spacing is recommended to determine success of any changes carried out and to provide operational feedback to ATC/Supervisors Positive ATC Speed Control Regime Following discussions with TCU and TWR controllers, together with observations and data analysis, it is apparent that a consistent ATC speed regime is not in place at Perth Airport. Furthermore, anecdotal reports suggest that in some cases when ATC speed control is applied, subsequent speed reductions required by the flight crew are initiated without reference to ATC. With the intention to optimise CTMS arrival rates to ensure minimum arrival spacing (subject to ADC assessment), it will be necessary for Approach Controllers to apply a more positive and standardised final approach speed control regime. This positive, ATC driven, speed control will assist in optimising spacing as the level of headwind on approach changes or as visibility deteriorates. Engagement with the flying community will be a key requirement to ensure that the majority of operators are in agreement as to the speed regime to be adopted. Those operators that are unable to comply with a new standardised requirement, such as those flying light turbo-props, should advise ATC accordingly of the speeds that are appropriate to their specific aircraft type. This information may then be used by approach control to tactically increase spacing behind such aircraft as required. Speed at the Feeder Fix varies considerably, driven by the crew s requirement to arrive at the feeder fix at a certain time. Where an aircraft is attempting to lose time and avoid holding, that aircraft may cross the feeder fix at a relatively low speed compared to following aircraft that may be using high speeds to gain time. This disparity in speeds often requires the approach controller to instruct the first aircraft to speed-up, a situation that is not desirable by flight crews. In moving to this more refined speed control regime, it is recommended that: a) Standard speeds used by ATC and the approximate distances from touchdown at which pilots should expect, though not fly until directed by ATC, should be published in the Perth Airport specific pages in AIP (see UK AIP, for LGW/EGKK AD 2-EGKK-1-15, 5. Detailed Procedures); b) ATC training/simulations be carried out in order to enable controller s to practice this speed optimisation technique to deliver specific target spacing at touchdown; c) A review of the approach controller s area of responsibility is undertaken to ensure minimum distraction from the primary task of delivering accurate spacing and speed control by other aircraft operating in and around the CTR; Page 52

53 d) A minimum speed at the Feeder Fix is promulgated to prevent catch-up by following aircraft. Analysis of final approach speed compliance is recommended to review application of speed control and subsequent compliance by specific type and operator to allow appropriate follow up to maintain performance and reduce variation. This could be achieved by performing an analysis of airborne radar data Auto-Release At present all departures are subject to a formal release by the TCU departures controller. The requirement by the TCU for a release is driven by the traffic complexity and constraints placed on approach by the proximity of RAAF Pearce and Jandakot. Whilst few delays were observed in obtaining a release, the process does place additional workload on ADC and departures. As Perth moves towards HIRO, all opportunities to reduce workload and streamline procedures should be investigated. With this in mind, consideration for establishing an auto-release procedure for all SID departures should be a long-term goal and will require the interaction of Jandakot and Pearce on Perth traffic to be considered and deconflicted where possible. In the short-term it is recommended that work continue to develop an auto-release procedure for Runway 03/06 northbound departures when Pearce is not active TWR TCU Supervisor Coordination In addition to those comments made in Section above, more direct coordination between the TWR Line Manager and the Flow Controller/TCU Supervisor should focus on arrival strategy such as favouring a wave of ADDDAs during the first wave of morning departures. a) In addition a daily informal, but outcome focussed, PDR (Plan Do Review) meeting/discussion between line managers may prove useful in planning first wave based upon weather conditions etc., reviewing the success of this plan after first wave and adjusting as required at least once per shift (but as frequently as deemed appropriate with experience); b) TWR Line Manager coordination with the AOC could assist in better planning ability of SMC to integrate towed movements and reduce the impact of aircraft held on stand as part of the new Departure Management Procedure TCU Flow Controller The Flow Controller in the TCU is a key member of the ATC network management team whose responsibilities include an oversight of the arrival and departure flow within the Perth TMA. However, observations show that the majority of the Flow Controller s time is taken up performing manual calculations to determine CTMS arrival rates. These calculations are time consuming and can lead to a significant increase in workload during runway changes. Moreover, the manual process of determining Feeder Fix arrival times can result in errors that ultimately affect capacity. The introduction of MAESTRO in the TCU would have a significant impact on the workload of the Flow Controller as the calculation of arrival times would be Page 53

54 performed automatically. Changes to arrival rates could be performed rapidly in response to tactical arrival/demand management. As well as eliminating a potential source of errors, MAESTRO would enable the Flow Controller to effectively manage the tactical flow within the TMA and liaise effectively with the TWR. This should ultimately lead to improved capacity at Perth airport together with a standardised operating methodology in line with other major Australian Airports Tower Approach Liaison It is evident that the tower and approach control functions have become focussed on their respective operations, striving to deliver either a maximum arrival or departure rate without considering the flow of traffic throughout the ATC network as a whole. It is recommended that a comprehensive liaison visit regime be initiated between tower and approach to facilitate an understanding of each other s roles and requirements. Not only would this be beneficial in understanding capacity constraints affecting HIRO but could also improve controller awareness and likely actions during emergency situations. Additionally, liaison visits would foster a team mentality where both disciplines are working together towards the common goal of higher capacity at Perth airport SMC Workload During the busy first-wave departure period it was observed that SMC experienced a significant increase in workload. High workload limits SMC s ability to adequately plan an optimal departure sequence, often resulting in a fire-fighting approach to air traffic, rather than a proactive one. Although a level of start-up metering was employed in the form of a Taxi Slot Time system, bunching of calls was observed as well as additional requests made by crews looking to move their slot forwards. On the 8 th March 2012 a new departure management procedure was introduced which provides all operators with a calculated off-blocks time, COBT. Operators will be required to call for pushback/taxi clearance within a tolerance of COBT plus 10 minutes. Note the NATS site visit occurred before the introduction of this revised process. The impact of the new procedure on SMC s workload and ground congestion is yet to be determined, however, it is recommended its effectiveness be reviewed 3 months after implementation. Should SMC s workload still continue to be high, the implementation of a Surface Movement Planner position (SMP) should be considered. The role of the SMP would be to pass clearances and control startup requests to meter the traffic and ensure that SMC is able to provide an effective service. Page 54

55 6.1.9 Non-Jet SIDs The use of some SIDs by both jets and non-jet is sub-optimal and if possible the introduction of early turns for non-jet traffic would be advantageous in reducing DD delay. The effective use of non-jet SIDs can be seen on Runway 21. However, local airspace and environmental restrictions currently impede their use on other runways. A review of the long-term introduction of additional non-jet SIDs is recommended to ensure maximum available departure capacity in line with projected traffic growth at Perth Optimal Runway Configurations Choice of runway operating configurations is determined by both wind and environmental reasons. The optimal runway configuration for a predominantly departure bias is 03A/03D/06D; choice of departing runway is based on departure route and aircraft type (large aircraft normally departing Runway 03). The most efficient mode for an arrival bias is 21A/24A/21D; choice of arriving runway is based on arrival route and aircraft type, with large arriving aircraft typically using Runway 21. A more efficient use of this south westerly configuration during periods of balanced departure and arrival demand would be to increase the number of arrivals using Runway 24 as follows: Runway 24 all arrivals, except those aircraft requiring Runway 21 for performance reasons; all arrivals spaced at minimum radar or wake turbulence spacing. Runway 21 all departures (the availability of a nonjet SID is of significant benefit from this runway); It is understood that a change in the frequency of such arrivals is likely to result in a response from the local noise community; however, such responses must be balanced and accounted for in relation to the beneficial impact of improved capacity at Perth LAHSO Operations Procedures for LAHSO have previously been developed and implemented at Perth but its use was not a success and it has since stopped. Reasons for this are varied, but feedback from the ATC and flying communities indicate that lack of consistent application across all airline operators and aircraft performance issues were factors. It is recommended that LAHSO operations are re-considered for the following modes of operation: a) Balanced departure/arrival demand: Runway 03 LAHSO & heavy DEPs, Runway 06 DEPs; b) Runway 24 LAHSO for turboprops, Runway 24/21 normal ARRs, 21 DEPs Jandakot and RAAF Pearce Airports The impact of Jandakot and RAAF Pearce airports on the capacity of Perth in certain weather conditions and runway configurations can be significant. Whilst Jandakot serves an important role in handling GA flights that might otherwise Page 55

56 use Perth, its effect on departing and arriving aircraft delays can be problematic. Whilst it is beyond the remit of this report to provide a comprehensive solution with regards to Jandakot and Pearce, a future capacity model for Perth must account for its impact. IFR traffic using Jandakot restricts IFR traffic using Perth due to the requirement for ATC to establish radar separation between both aircraft. In order to minimise this impact, several options should be considered: a) Establish a deemed separation for IFR departures from Jandakot from eastbound SIDs from Runway 21 at Perth; b) Develop a SID from Jandakot that leaves controlled airspace to the southwest and is deemed separated from Runway 21 SIDs at Perth; c) Restrict IFR training at Jandakot to weekends when traffic demand at Perth is significantly reduced. Military airspace surrounding RAAF Pearce places significant restrictions to traffic departing Perth to the North, limiting the potential to split jet and non-jet traffic after departure and therefore incurring capacity constraints. During peak departure demand, and prior to Pearce becoming active at 08:00 local, an agreement has been reached with the RAAF that enables northbound departures from Runway 03 to fly through the military airspace. Whilst this procedure provides shortened tracks for departures, it also results in increased RTF as ADC must first cancel the SID previously issued by SMC then instruct the aircraft to fly a radar heading after departure. In the short term, it is recommended that all northbound departures prior to 0800 local be issued with a Perth 4 SID by SMC, thereby reducing subsequent workload for ADC. As a long-term goal, it is recommended that Airservices Australia investigate a solution similar to that already underway in the UK: the Flexible use of Airspace concept (FUA). The basis of this concept is that airspace should be considered as one continuum and not designated as either civil or military. The airspace would be used flexibly on a day-to-day basis with any segregation only of a temporary nature Wake Turbulence Requirements The HEAVY HEAVY distance based departure wake turbulence separation of 4nm in trail (as detailed in MATS ) is not required and should be deleted as significant experience of Heavy-Heavy departing pairs in UK has resulted in there being no requirement for Wake Turbulence separation between two heavy departures (see UK CAA CAP493 MATS Part 1, Section 1, Chapter 3 Page 12, Para 9.6) Terminal WA The construction of Terminal WA, due for completion in 2013, will have a significant impact on surface traffic flow as major operators relocate from the aprons on the western side of the airfield. There is little evidence of a plan between all stakeholders that joins up all activities (ground and air) to ensure that capacity (and potentially safety) are not compromised when the new terminal becomes operational. It is suggested that a concept of operations is developed at the earliest opportunity in conjunction with ATC, Perth Airport and airline operators. Page 56

57 Standardisation Operational Performance and Airport A standardised approach across all disciplines is key to ensuring the success of HIRO. Predictability and consistency will help ensure that no opportunities are missed and that airport capacity is maximised. It is recommended that measures are taken to achieve standardisation which include: a) ATC, pilot and airport operators training for new procedures; b) ATC competency checks and appropriate training where required, to reduce inconsistency across controllers and disciplines; c) Joint Airport, Pilot and ATC forums where HIRO topics are discussed and reviewed, including runway occupancy metrics; 6.2 Ground Infrastructure and Operations The initial focus of the ACE program is on maximising runway throughput. Notwithstanding this, the programme acknowledges that airport efficiency can also be constrained by taxiways and aprons/gates. This section provides a summary of the NATS observations relating to the taxiway and stand/gate allocation functions. The key ground infrastructure for efficient HIRO requires adequate runway exit and line up points, a taxiway system to allow a constant stream of departure traffic to feed the departure runway and enough taxi infrastructure and stand capacity to ensure that arriving traffic does not impact the operation Taxiway Layout and Surface Throughput J, A & Runway 06/24 Intersection The Juliet, Alpha & Runway 06/24 intersection is recognised as one of the primary choke points on the airfield that impacts surface throughput, particularly during Runway 03 operations. All aircraft taxiing from the northwest stands to depart Runway 03 must pass through this intersection, potentially coming into direct conflict with aircraft landing Runway 03 and vacating at either Kilo or Juliet 2, as well as those aircraft using Runway 06. The loss of this intersection would have a significant impact on the capacity at Perth, requiring aircraft to be processed through the eastern side of the airfield; requiring all aircraft to cross Runway 03 at Delta or Bravo/Whiskey. The construction of a new taxiway between the Runway 06 threshold and taxiway Alpha is already underway which will provide an additional taxi route as well as providing some mitigation against the loss of this intersection. Due to the complexity of the existing infrastructure there is no other obvious quick win solution that could enhance the taxiing operation in this area without significant change. Runway Crossings The layout of the airfield is such that aircraft frequently have to cross an active runway after landing or prior to departure. The impact on runway throughput is discussed below, however, delays to aircraft waiting to cross an active runway also impacts gate-to-gate performance. Consideration should be given to utilising runways based on parking position rather than airborne routes. Page 57

58 Additionally, several delays were observed with towed aircraft waiting to cross Runway 03/21 from the eastern to the western side of the airfield, particularly during the busy morning period. Tugs towing aircraft are given a lower priority by ATC compared to other aircraft movements and therefore incur more ground delay whilst traversing the airfield, this is exacerbated by the limited number of taxiways available to SMC and resultant congestion during the first wave of departures. The delay caused to towed movements being repositioned from remote stands can consequently affect the first wave departure performance of the airlines involved. While NATS understands that some work has already been undertaken in assessing the impact of towed operations, we were not able to ascertain the full extent of the activity completed so far. Therefore it is recommended that ATC and the airlines determine the impact and extent of these delays and explore the following options (assuming they have not already been assessed): a) Increasing the priority given to towed movements; b) Delaying towed movements to the end of the morning peak departure period; c) Planning stand-to-stand towing routes to minimise impact on the flow of normal taxi routes. Parallel Taxiways The absence of dedicated parallel taxiways impacts gate-to-gate performance, particularly on Runways 03/06, as outbound aircraft from the terminal 2 & 3 stands push-back onto a main taxi route to the Victor holding point for Runway 06. This taxi route can often become congested with several aircraft holding on the taxiway for departure, preventing inbound aircraft from reaching their stand. The limited number of available taxi routes also hinders towed aircraft during the first wave of departures; particularly as such movements are often in the opposite direction to the general flow of aircraft. With the existing taxiway infrastructure there is no obvious solution to this issue. Taxiway Width Restrictions Width restrictions on primary taxi routes limits flexibility for SMC when taxiing large aircraft, resulting in choke points and taxi delay. Such restrictions were evident on Victor and Hotel 3 and a section of Juliet due to WIP. The width restriction on Victor also results in large aircraft, such as B767, being required to either backtrack Runway 06 for departure (or Runway 24 after landing) or taxi to Runway 03 for departure; both resulting in additional taxi time and fuel burn for such aircraft as well as potential impact to airfield capacity. Taxiway Papa & C3 Intersection At present, there are no taxiway paint markings to enable aircraft to turn left from C3 onto Papa, or vice-versa; this is likely due to the absence of a taxiway filet to enable such turns to take place. With the growth of traffic on the eastern side of the airfield, it is recommended that the construction of a taxiway filet, or addition of paint markings, be considered prior to the completion of Terminal WA this would provide more flexible use of taxiways for crossing traffic. Ahead of Terminal WA s opening it is not known if this would provide any tangible capacity benefit. Page 58

59 Helicopter Aiming Point (HAP) Operational Performance and Airport At present there is no dedicated HAP, the intersection of Romeo and Whiskey being used instead. This potentially causes the following problems: delay for traffic using the aprons serviced by taxiway Romeo and wake turbulence issues due to the aiming point being sited closer than 760m to Runway 03/21. Extensive use of the current aiming point was observed on one day when the frequency of fire-fighting helicopter movements were significantly increased, however, this occurred at a time when there was little demand for the apron serviced by Romeo. As traffic levels grow, it is recommended that consideration be given to the siting of a new HAP that does not impede the flow of traffic and is sited more than 760m from the runways Taxiway Layout and Runway Throughput J, A & Runway 06/24 Intersection The impact of the J, A & 06/24 intersection on surface throughput is discussed in the previous section. It is evident, however, that the complex nature of this intersection also has the potential to impact runway capacity. ADC will often instruct aircraft to vacate Runway 03 at Delta rather than Juliet 2 in order to avoid a confliction on the taxiway with opposite direction traffic, this resulted in an observed increase in arot times. Holding Points Bravo & Whiskey The current positioning of the Bravo and Whiskey holding points are such that they are located a significant distance from the runway edge, resulting in increased departure LUTs onto Runway 21. An extension to taxiway Alpha is planned that will also include the siting of a new holding point closer to the runway. Whilst this will improve LUTs, it could also increase ADC RTF (as discussed in section below). Runway Crossings The layout of the airfield is such that aircraft frequently have to cross an active runway after landing or prior to departure. Such crossings potentially impact the operation as follows: 1. Potential loss of an arrival or departure gap whilst an aircraft crosses the runway after landing or prior to departure, with a resultant impact on capacity; 2. Taxi delay whilst coordination is effected between ADC and SMC prior to the runway crossing; 3. Safety any runway movement carries an inherent safety risk. It is recommended that runway crossings be minimised by allocating the arrival or departure runway based on parking position rather than arrival or departure route. This methodology will become particularly significant when Terminal WA becomes operational in 2013, with the consequential shift of some airline operators from the west to the east side of Runway 03/21. It is further recommended that all runway-crossing movements be retained on the ADC frequency. Whilst eliminating any delay in ADC/SMC coordination, this change would also align crossing procedures with international best practice, which documents that all vehicular and aircraft crossings of active runways Page 59

60 should be carried out on the ADC RTF frequency to aid pilot /driver situational awareness. RETs The lack of dedicated RETs invariably results in increased arrival runway occupancy times and is evident in the data presented in section 5.4. Increased arot ultimately results in increased arrival spacing when operating in a single runway configuration with a consequential decrease in runway capacity, this is particularly evident for Runway 03 that has larger overall arot. Setting optimal locations for RETs based on traffic mix is relatively straightforward however positioning them into existing infrastructure as complex as that at Perth Airport is much more difficult and needs careful consideration. From the NATS observations and site assessment it is clear that there is no ideal location for new RET infrastructure without consequential changes to the existing taxiway infrastructure. Single Runway Entry Points Single full-length entry points onto the runways limits SMC s ability to deliver an optimum departure order for ADC. Intersections are available for departure but are often limited to 1 or 2 aircraft before the taxiway behind them becomes blocked. Additionally, the majority of intersections are positioned more than 150m from the full-length holding point, resulting in potential wake turbulence considerations. It is recommended that additional runway entry points be considered in the Master Plan to provide additional flexibility to ATC in determining the optimal departure sequence as well as providing an element of redundancy in the system should a main holding point become unserviceable Pushback/Taxi Procedures/Holding Bay Capacity & Impact on Congestion & RTF Parallel Taxiways The absence of dedicated parallel taxiways often results in congestion and a consequential increase in workload for SMC. Aircraft on the 900 stands and those on Terminals 2 & 3 stands mostly push back onto the taxiways behind (Hotel, Juliet and Victor). This becomes a significant problem when operating Runway 06 for departures, as these are the principle taxiways used for aircraft routing to holding point Victor. Additionally, aircraft held remotely on these stands are required to be towed to other main apron stands, often during the morning peak departure time and against the flow of aircraft taxiing to Victor. The above issues result in delays to aircraft taxiing to Victor, aircraft inbound to stands blocked by a departure queue at Victor and delays in repositioning aircraft from remote stands that are part of the first wave of departures; reducing on time performance. It also results in a significant increase in workload for SMC whose cognitive process is taken up by resolving conflictions and managing delay rather than providing the ideal departure order at the holding points. It is recommended that an investigation is performed on the first wave departure operation to assess if any enhancements can be made to help minimise such issues. Careful planning of the anticipated departure order may Page 60

61 be of benefit although NATS understands how difficult this can be particularly when runway modes may change. Holding Bay Capacity There are no dedicated holding bays at Perth for departing traffic, this combined with the lack of parallel taxiways means that there is little option for ADC to change the departure sequence once a queue of aircraft has formed at a holding point. Holding Points Bravo & Whiskey Whilst the siting of a new holding point between Bravo/Whiskey should reduce LUTs, it should be noted that the likely operating procedure by ATC will be for SMC to instruct aircraft to taxi to either Bravo or Whiskey, requiring ADC to make an additional transmission to instruct the desired aircraft to taxi from those holding points to the new holding point located closer to the runway. The requirement to make an additional transmission is not ideal, particularly when moving towards a HIRO environment, when the desire should be to minimise all RTF, subsequently reducing ATC workload and potentially improving cognitive capacity. In addition to the recommendation discussed above in Single Runway Entry Points it is further recommended that consideration be given to the construction of an additional entry point to the runway, whilst mitigating against any potential increase in ADC RTF, it would also provide a degree of taxiway redundancy should the full length holding point become unserviceable. RTF Currently, all departing traffic informs ADC when they are ready for departure, causing three potential issues for ADC: a) The transmission from the pilot could occur at a critical time for ADC when they were about to issue a landing or take-off clearance; b) The transmission is sometimes made prior to the aircraft leaving the apron and before ADC has received the departure flight progress strip, leading to confusion; c) Increased RTF congestion. In the short-term, it is recommended that crews report ready for departure when at, or close to, the holding point and only when they are number one or two for departure. In the long-term it is recommended that departures Monitor ADC, thereby reducing RTF congestion on the ADC frequency. During low intensity operations, combined with reduced taxi times, an additional transmission may be required by ADC to confirm if the aircraft is ready. However, the additional workload incurred by ADC in this situation is outweighed by the benefit of reduced RTF loading during HIRO Recommendations for Optimum Surface Traffic Flows Perth ATC currently demonstrates effective use of taxiways and routings to maximise surface throughput. Two future options should be considered: Page 61

62 a) Taxiway Alpha be utilised for turbo-prop departures on Runway 06 following the completion of taxiway Juliet WIP. This would split the majority of jet and non-jet traffic allowing ADC to finesse the departure order and minimise overall departure delay. b) Victor, Runway 06 threshold, and the new taxiway linking Victor and Alpha be used for outbounds during periods of high arrival demand on Runway 03 operations enabling Juliet 2 to be used consistently for vacating aircraft, thereby minimising arots Apron / Gate Management The current process for gate allocation at Perth Airport is split between the airport and the main airlines, Qantas managing the majority of stands that fall outside the remit of the airport stand planning function. The allocation process is strategically managed around the 2 seasons of operation running March to October and October to March and tactically managed via regular schedule updates from each airline. Such updates were reported to be available at various times with refinement occurring -1 month, -1 week down to day before operations and update messages on the day of operation. Through discussions with the airport and airline representatives NATS understands that the planning functions have access to good situational awareness via various electronic and web based data feeds. However there is currently no Eurocat (airborne radar) or ASMGCS (ground radar) situational awareness display available to them. Visibility of these displays (in the case of ASMGCS once it is available) could provide benefit to the stand planners in their tactical allocation / reallocation of gates and better heads up to predict issues ahead of time. Additionally this data can support the planning units in any data recording requirements that they have. For example stand utilisation data was supplied by Perth Airport, this was cross checked against NOC supplied runway log data there were a small number of inconsistencies noted. The ATC tower has access to the T1 and T3 stand plan however the display is currently positioned behind the controller positions. It is understood that there is a project in progress to add the T2 data to this display. The combination of this additional data together with a more appropriate position for the tower display would be of potential benefit to the ATC operation. It is recommended that until this is implemented Qantas inform the tower (the supervisor or coordinator) when an inbound aircraft is likely to be delayed accessing their stand. For stands where SMC can identify that an arrival flight will need to hold - either through direct visual observation or from situational awareness via strips that the gate is still occupied better and more consistent coordination of the likely length of time before the stand is available would allow SMC to manage the traffic more efficiently. This information can be very difficult to predict but further investigation may be of benefit in order to determine if improved coordination is possible. A first step to achieve this could be through the establishment of a working group or closer liaison between the planning functions and the tower to ensure each area understands the issues and can collaborate in their solution. Stand plan and towing plan performance could be part of external daily PDR discussions between the ATC Line Manager and AOC supervisor. Page 62

63 NATS was informed of a recent activity to reduce the need for domestic towing by reconfiguration of stands via changes to paint lines. Such activities can result in reduced need to tow, associated reduction in taxiway congestion, delay and potential capacity increase. Undertaking a regular review of such developments will ensure that any such potential quick wins are realised. It was noted during discussions with Perth Airport that Skippers and closed charters operating out of the north western stand areas do not appear in schedules and typically operate to their own agenda. Additionally because of the nature of the passenger loading for these flights i.e. they are not able to taxi out when another aircraft is loading passengers it is typical for more than one flight to require taxi out from the apron area at the same time. While in the tower NATS observed some bunching as a result of this. It is not clear if there is a gate and operations management strategy for such movements and as such it is recommended that further investigation is made to ascertain if improvements in the operation are possible for example to deliver a more smoothed demand Other Observations - Vehicles Reports from ATC indicate that calls are routinely made by vehicles requesting to transit the taxiways, increasing workload for SMC and causing RTF congestion. It is recommended that ATC and Perth Airport review the vehicles that use the manoeuvring area and develop a list of companies that may freerange without reference to ATC. 6.3 Data and Reporting This section summarises the NATS team views on the availability of key data and reports and their potential use for on-going performance monitoring and capacity index assessments at Perth Airport. At the time of the NATS site visits, Metron Traffic Flow (MTF) was not live and therefore the comments below are based on CTMS, they do not account for any changes either to the operation or the resulting data that may be available as a consequence of MTF NOC National Operations Centre The Airservices National Operations Centre has been established to provide centralised Air Traffic Flow Management (ATFM) within the Melbourne & Brisbane FIRs. Strategic planning is carried out by the NOC and pre tactical planning is carried out by the NOC in conjunction with ATC and aircraft operators. The NOC is responsible for preparing: Information & data for the Daily Operations & Safety Report (DOSR) Delay Analysis Report which presents a detailed picture of the level of delays into BNE, MEL & SYD Airports Responses to data requests The NOC has access to Airservices system information (Eurocat, Aerobahn & Metron) and processes airline schedule information & data. The NOC is the sole source for airport performance reporting & monitoring. There are similarities between the Airservices NOC and NATS Operational Analysis (OA) department in respect of the services they provide to operational Page 63

64 units, however within NATS the OA department has additional service level agreements in place to provide both routine and ad-hoc detailed analysis and simulation modelling assessments. While not in direct scope of this project it is worth noting that as the NOC has a similar remit to OA in terms of data, there may be potential benefits to the wider Airservices operation if the NOC were to provide a similar set of more detailed analysis and simulation services Schedules Schedule data is the foundation of the daily plan of aircraft movements and is the reference data from which delay values and on time performance can be calculated and assessed. Additionally for non-schedule coordinated airports such as Perth the daily schedule can be a potential source of key data to help support the airfield planning and tactical management. A simple analysis of the daily schedule can provide insight into potential issues that may arise; typically schedules are planned against 5 minutes on/off block times, thus a simple analysis can provide a visualisation of the schedule demand each 5 minutes. Peaks are easily identified as are periods of sustained demand. There is potential benefit to be gained through greater visibility of the daily schedule, particularly when considered with other key airport information such as delay levels. A review of the current or previous day s operation, recorded delays and any operational issues (possibly as part of a daily PDR meeting) may provide greater insight into root causes of increased delays and therefore support the airfield operation. Figure 45 and Figure 46 show the Perth schedule data for 15th February 2012; each line represents the number of scheduled arrival and departure movements (based on their on/off block time) for each 5 minute period of the day. Peaks in the arrival schedule occur at numerous times throughout the day. A value of 6 arrival movements being scheduled between 0940 and 0945, 1140 and 1145, and between 1820 and Similarly departure peaks are seen throughout the day with the busiest 5 minute intervals occurring between 0630 and 0635 and 1430 and 1435, when 12 departure movements are scheduled in each period. By way of an example; in order to manage 12 departures in a 5 minute period from a single runway without delay would require no arrival demand and DD separations between each departure of 25 seconds. Conversely if all 12 departures called ready at the start of the 5 minute period then assuming a separation of 60 seconds between departures would require a total of 11 minutes to depart all the traffic. The average departure delay would be 6 minutes, the peak value 11 minutes and the total delay 66 minutes. It is easy to see that as consecutive 5 minute scheduling periods are filled, delays can accrue at an accelerating rate. Page 64

65 Figure 45. Example demonstrating arrival schedule bunching (note intervals without scheduled demand are not shown). Figure 46. Example demonstrating departure schedule bunching (note intervals without scheduled demand are not shown). During such bunched periods of demand it is normal to observe increased levels of delay and congestion, with corresponding increases in the level of ATC workload and traffic management complexity. This is particularly relevant at Perth Airport considering the high bunching of demand especially for departures in the early morning period. NATS understands that Perth Airport is currently considering a form of schedule coordination. Any such coordination activity that aims to smooth localised peaks will be of potential benefit to the airfield efficiency and hence support capacity enhancement. Page 65

66 6.3.3 KPI data Operational Performance and Airport During the NATS visit the NOC supplied data referred to as a KPI report this included data such as actual runway and stand times for arrivals and departures. The data also provided information on the aircraft types and operators which facilitated improved segregation of the data. This data was used to augment the data directly observed and helped facilitate the departure delay calculations; however the potential value of this data was compromised due to the limited accuracy of the data stored (e.g. times to the nearest minute) and also the limited coverage of stand departure times. Discussions with NOC indicated that more accurate data may currently be recorded and it would be beneficial to any future analysis of the Perth operation if this data is made available Eurocat Airborne Surveillance Data Airborne surveillance (radar) data was supplied to NATS in a CSV format; this included both 4D (latitude, longitude, height and time) track information and a flight plan reference which allows a detailed analysis of actual tracks flown. This data was sourced from the Brisbane and Melbourne centres, with the Brisbane centre covering the northern Australian area including Brisbane Airport and surrounding airspace and the Melbourne centre covering Melbourne, Perth and Sydney Airports and airspace. This data was used in the NATS analysis to determine pre-arrival holding and sequencing delay; this being calculated on an aircraft by aircraft basis as the difference between the actual time to fly from a defined cordon to runway threshold and a calculated standard (un-delayed) time. While the calculation of this delay is a key input to the operational performance assessment, further analysis of this data can yield other key operational performance metrics such as: CDA / CCD compliance monitoring; Environmental metrics - fuel burn and CO 2 emissions; Noise monitoring; ATC Sector loadings and complexity; Achieved runway separations; Approach speed profiles. NATS has developed tools to analyse such radar track data to compute and routinely monitor such key metrics. The most relevant to capacity monitoring and enhancement is that of runway separations. It is recommended that Airservices develop a way of routinely monitoring achieved runway separations. This information could be linked to runway modes and operational separation parameters (similar to those automatically logged in the MAESTRO system at Brisbane, Melbourne and Sydney) and weather data to allow performance monitoring of all key runway separation parameters such as AA, ADA, ADDA, DD etc. This would provide Airservices with a quick and effective way of tracking the most important parameters in relation to airfield capacity. Additionally this sort of analysis and reporting can be used to track aircraft performance and particularly aircraft speeds on the various approach legs including the final approach speed to allow a detailed understanding of typical Page 66

67 airline/aircraft type speed profiles to be developed. This information could support a national programme of agreeing approach speeds; this being a key driver of runway capacity as described earlier in this report ASMGCS Ground Surveillance Data ASMGCS is not currently available at Perth Airport however it is anticipated to be available in the near future. ASMGCS data can provide a wealth of key airfield performance related information that can be of huge benefit in capacity and performance monitoring and improvement programmes. While not currently available at Perth, during NATS visit to Melbourne and Sydney Airports we were able to obtain sample raw ASMGCS track data and test it within NATS tools. We were able to confirm that the data present at those airports and anticipated to be available at Perth (once the ASMGCS system is operational) is ideally suited in supporting detailed analysis of airfield operational parameters and ultimately performance and efficiency monitoring. NATS has developed tools to automatically analyse and report on key airfield parameters by utilising ASMGCS and supporting data such as schedule or flight plan information. By performing an analysis of the ground track data the following events can be identified for departure flights Pushback time Taxi out time Join runway holding point time Line up time Wheels roll time Wheels up time And for arrivals flights Cross runway concrete time Over threshold time Exit runway time And for taxiing and runway crossing traffic Enter / exit taxiway block times Runway hold time for cross Runway exit time for cross Exit point used Exit speed profile On stand time From these parameters the following key performance metrics can be calculated: Arrival runway occupancy time Line up time for departure Departure runway occupancy time Runway holding point delay time Runway crossing delay Runway crossing duration Taxi in / out times Runway separations AA, ADA, ADDA, DD etc This information can then be analysed and assessments can be made; these can include comparisons across carriers and aircraft types as well as correlations of Page 67

68 separation parameters to local operating conditions and runway modes. From this analysis further capacity modelling or target performance setting can be undertaken. Additionally the regular supply of this data can be of huge benefit to the airport users and stakeholders; within the UK regular meetings are held with the pilot community where this analysis is presented and discussed Stand Data Perth Airport supplied a report which included for each arriving and departing aircraft, the on/off chocks time and terminal used. The times were rounded to the nearest minute. For the majority of flights, the time correlated well with the equivalent time from the NOC KPI report however there were some differences. The stand data was used to correlate aircraft to terminals for the purpose of estimating join runway holding times and subsequent departure delay analysis. This data could be of potential further benefit either in looking at runway demand/separations or stand demand/usage. However since coverage was incomplete and times were recorded with a low level of precision, further investigation would be needed before this data is used for any detailed analysis MAESTRO Logs As MAESTRO is not currently in place at Perth Airport, no MAESTRO or equivalent log information was available. NATS understands that there is a discussion paper currently being considered regarding the implementation of MAESTRO at Perth Airport. A brief analysis of sample MAESTRO logs was carried out while the NATS team were on site at Melbourne Airport. The log data appears to contain the MAESTRO delay information for each flight as it progresses from the first time it appears in the MAESTRO system to the time it lands. There is potential therefore for this log to provide a wealth of information relating to the MAESTRO delays for flights (assuming MAESTRO is implemented at Perth). NATS suggests that if it is implemented then as part of its implementation further investigation is made into the availability and potential benefit of analysing this data. For example an analysis could be performed of MAESTRO arrival rates and separations, MAESTRO logged delays and achieved runway separations (using radar or ASMGCS data when available). By gaining a detailed understanding of the relationship between the configured MAESTRO rates, the resulting delays to aircraft, the achieved runway separations and by considering airline schedules and weather, etc. there is potential for refinement of MAESTRO rates and hence capacity enhancement Multiple Data Sources The above sections outline the data sources identified during the NATS site visit that are of most use in the support of airfield operational performance monitoring and capacity index review. While each data element is of use individually, the most benefit is derived when the individual data sources are combined into a coherent single data stream. Page 68

69 The OA department in NATS has developed processes and tools to perform this data fusion. Additionally OA holds the responsibility to ensure that any proposed changes to central systems include requirements for data logging to allow ongoing operational performance monitoring and analysis to take place. NATS is now in a position that would enable these experiences to be shared and work with Airservices Australia and Perth Airport to help deploy a similar automated data collection and fusion system for Perth Airport. This would facilitate rapid assessment of airfield performance for the purposes of further optimising airfield procedures. Page 69

70 7 Key Recommendations for Target Capacity Assessment 7.1 Reduced Arrival-Arrival Separations In Section the opportunity to increase runway capacity and reduce delays through reduced arrival spacing was discussed. For the purposes of simulating and analysing the target capacity scenario, it is recommended that arrivalarrival separations are reduced to ICAO wake turbulence minima at threshold. These are shown in Table 14. Table 14. ICAO minimum threshold spacing between wake turbulence constrained arrivals. 7.2 Reduced Arrival-Departure-Arrival Separations In Section the opportunity to increase runway capacity and reduce delays through reduced arrival spacing was discussed. For the purposes of simulating and analysing the target capacity scenario, it is recommended that arrivaldeparture-arrival separations are reduced to 150 seconds. 7.3 Arrival-Departure Balancing In high-intensity operations, it is common to adapt the arrival flow rate to balance arrival and departure demand for the runway. In conjunction with the recommendations for simulation outlined in Sections 7.1 and 7.2, the selection of appropriate arrival spacing should be considered with the balance of arrival and departure demand. 7.4 Additional Exit Taxiways In Section it was observed that certain runway configurations and exit points exhibit large arot times. The lack of suitable runway exit infrastructure may result in an increased arot particularly for heavy aircraft, this may become a limiting factor in reducing arrival separations. As a result, it is recommended that this effect is tested during the simulations. Page 70

71 8 Current Capacity Index 8.1 Introduction to HERMES Modelling HERMES is a stochastic fast time model which simulates aircraft movements on and around one or many runways. HERMES estimates the achievable runway throughput and delays that may be experienced for arrival and departure aircraft based on a specific traffic movement profile and operational statistics. Randomisation of the operational statistics is applied in the model over a large number of iterations to reflect the perturbations in performance typically experienced in an air traffic environment. Such perturbations are achieved through random selection of data based on the observed distribution of each input parameter. The simulation model is configured based on the current operational performance, using observed statistics and traffic samples for each wake turbulence category as shown in Section 5. Several aspects of the simulation model are validated to ensure that current operational performance is replicated appropriately; in particular confirming that the simulated delay profile can be attributed to observed changes in the airfield operation. This includes monitoring the magnitude of delays from the simulation against the observed delays, reviewing the delay profile throughout the day, and the response of the model to both varying demand and changes in operational parameters. The process applied for developing and validating the simulations of the runways at Perth Airport follows the method developed and applied over many years at capacity constrained airports across the UK. Using the validated model, the current capacity index is considered by growing traffic within the simulation up to a defined level of acceptable delay. The effects of implementing the recommendations from Section 7 are considered in Section 9 to assess the target capacity index. 8.2 Simulation Validation The simulation for each runway mode has been configured to reflect the operation of Perth Airport as observed during the period 21 st -28 th Feb Specific traffic movement profiles have been used for each observation day which have been generated from the Perth KPI reports supplied by the NOC in February and March Such profiles consider factors such as haulage category, wake turbulence classification, and engine type (jet/turboprop). Table 15 shows the average observed hourly traffic profile from across the observation days. Page 71

72 Hour (UTC+8) Observed Arrivals Departures Combined Total Table 15. Average observed hourly traffic profile. During the observation period, the peak arrivals hourly rate was 21 movements, seen at 1100, 1800 and 1900 (UTC+8). The peak departure hourly rate was 34 movements at 0600 (UTC+8). The peak hourly movement rate was also at 0600 (UTC+8), with 37 movements comprised of 34 departures and 3 arrivals. The operational statistics used to configure the model are taken from Section 5 and primarily include arrival/departure spacing, arrival/departure runway occupancy times, departure line-up times, and pilot reaction times Runway 21 (Arrivals/Departures) and Runway 24 (Arrivals only) Model Validation The simulation has been configured for Runway 21 (arrivals/departures) and Runway 24 (arrivals only), which was in operation during the majority of the NATS visit. This configuration has been implemented using the description of operations agreed in the Perth Assumptions (Reference 6). Of key importance are the following assumptions: Page 72

73 Operational Performance and Airport Arrival operations are dependent on each other hence any following arrival will not be within 4nm of the runway intersection when the leading landing aircraft crosses the threshold of the adjacent runway. Arrivals from the East land on Runway 24 with arrivals from the North landing on Runway 21, except any heavy category aircraft from the East, which also land on Runway 21. Separations between simultaneous departures vary depending on aircraft type (jet/turboprop) and direction: o Jet followed by Jet (J-J): Same track 3nm, different track 1800m; o Jet followed by Prop (J-P): 1800m; o Prop followed by Jet (P-J): Same track 5nm, different track 1800m; o Prop followed by Prop (P-P): Same track 3nm, different track 1800m. A comparison of the simulated average arrival delays for the Runway 21/24 configuration and the observed average arrival delays for all modes of operation is shown in Figure 47. It is clear that the simulated delays closely match the observed delays, both in terms of the magnitude and the shape of the profile throughout the day. Figure 47. Simulated and observed arrival delay profiles for Runway 21/24 operation. The simulated average departure delays for the Runway 21/24 configuration are shown in Figure 48 compared to the observed average departure delays for all modes of operation. The simulated delays are a close match for the observed delays, both in terms of the magnitude of delay in the morning peak period and the shape of the delay profile throughout the day. Page 73

74 Figure 48. Simulated and observed departure delay profiles for Runway 21/24 operation. During the middle part of the day ( , UTC+8) when there is low departure demand (Table 15), the observed delays are influenced by bunching of departure aircraft on specific days. Fewer direct observations of the operation were undertaken during this period of the day during the NATS visit and so the departure delay line contains mainly calculated delays as described in Section In addition, the impact of both variable levels of PRT and aircraft at departure holding points not being ready to depart is likely to have affected the estimated delays. As a result, the levels of simulated departure delays are appropriate and provide the requisite level of confidence, following the pattern that would be expected when compared to the observed delays. Both of the blue lines in Figure 47 and Figure 48 show a strong pattern of correlation when compared to the observed delays, so will be carried forward to the current capacity index assessment in Section Page 74

75 8.2.2 Runway 03 Mixed-Mode Model Validation Runway 03 mixed-mode operations were only observed for short periods of time during the NATS visit and therefore limited data was available to directly validate the Runway 03 simulation. For a consistent demand profile, it would be expected that a single runway operating in a mixed-mode configuration would result in higher average delays when compared to a multiple runway operation. This is reflected in the rates displayed in Table 4. Since the arrival demand is only serviced by Runway 03 this also affects departures, as the higher arrival demand can restrict the size of the gaps between arrivals for departing aircraft. The simulated average arrival delays for Runway 03 are shown in Figure 49, compared to the observed average arrival delays for all modes of operation. It is clear that the peak average arrival delay simulated for Runway 03 is higher than the observed delay (and that simulated for the Runway 21/24 configuration in Section 8.2.1) in both the morning and afternoon peak periods. Figure 49. Simulated and observed arrival delay profiles for Runway 03 mixed-mode. Note that the simulated arrival delay exceeds 10 minutes during both the 1800 and 1900 hours (UTC+8). The simulated average departure delays for Runway 03 are shown in Figure 50, compared to the observed average departure delays for all modes of operation. As with arrivals, the peak average departure delay simulated for Runway 03 is higher than the observed delay in both the morning and afternoon peak periods. Page 75

76 Figure 50. Simulated and observed departure delay profiles for Runway 03 mixed-mode. Since the simulated delay profiles for both arrivals and departures follow the pattern that would be expected, and in lieu of any observation data to directly validate the simulation, this evidence is deemed sufficient to provide confidence that the simulation is accurately reflecting the mixed-mode performance of Runway 03 and that an appropriate level of delay has been simulated for this runway configuration. The data provided in Section and Section demonstrated that the delay profiles obtained for both the Runway 21/24 and Runway 03 simulations produce a magnitude and profile of delay that corresponds well to the observed/expected delay. This provides the requisite level of confidence to take these simulations forward for use in the capacity index calculations. Page 76

77 8.3 Current Capacity Index Assessment For the purposes of determining the capacity index for Perth Airport within current operating practices, the maximum allowable average delay in any given half-hour period was agreed to be 10 minutes. Having configured the simulation models using current operational performance and traffic demand (Section 8.2), the current capacity index has been assessed by increasing the traffic demand with consideration to Perth Airport s forecast traffic growth * and the impact upon arrival and departure delay relative to a defined level of delay Forecast Traffic Growth The current traffic profile has been grown across the day to estimate the overall number of ATMs the current runway infrastructure can accommodate under the observed operational performance statistics. To determine the current capacity index, traffic is iteratively added to the average observed traffic profile shown in Table 15 until the 10 minute delay criteria is reached. Analysis of the forecast traffic growth at Perth Airport showed that the proportions of aircraft types observed on the NATS visit are likely to remain consistent over the next eight years, as seen in Figure 51. As a result, the current proportions of aircraft types and wake categories will be retained in the modelled simulation for the current capacity index. Figure 51. Forecasted proportions of movements by wake turbulence category for * Provided by Perth Airport on 22 nd Feb Page 77

78 Traffic has been grown proportionally within each hour based on the current arrival/departure balance in each hour, thus respecting the current peak periods of demand for both arrivals and departures. In assessing the current capacity index, traffic has been added first in these peak hours, then in the shoulder periods around these peaks, and finally any remaining natural gaps have been filled. Within each hour the traffic profile has been considered in 15-minute intervals. Schedule co-ordination has not been considered and so traffic is added to the observed profile with the aim of retaining the observed balance of arrivals and departures, wake turbulence categories, and jet/turboprop demand within each hour Runway 21/24 Current Capacity Index In general, it has been seen that the observed average arrival delay was higher across the day than departure delay (as shown in Sections and 5.5.5), despite the overall number of arrival and departure ATMs being similar across the day (Table 15). In Section the current arrival spacing was identified as being a key limit to future capacity. Table 16 outlines the maximum achievable arrival service rates under the average arrival spacing observed during the NATS site visit. Method Of Spacing Average Observed Spacing (Seconds) Proportion of Observed Operations (%) Table 16. Maximum arrival service rates with current observed arrival spacing. Maximum Number of Arrivals per Hour Arrival-Arrival Arrival-Departure-Arrival Arrival-Departure-Departure-Arrival As an example, an hour dominated by arrivals (and thus AA separations) is limited to 25 movements. For an hour with a more equal balance, running ADA separations throughout, it would be possible for 41 movements in the hour (made up of 21 arrivals and 20 departures). This suggests that the arrival delays will be the limiting factor to any future capacity growth, since the delays are directly driven by the separations applied between arrival aircraft. In addition, with consideration to the observed hourly traffic profile (Table 15), it is clear that the balance between arrivals and departures for many periods of the day is not at or near 50:50. During the day there are clear periods of higher arrival demand (i.e and , UTC+8) and periods of higher departure demand (i.e and , UTC+8). As a result, the demand profile across the day is regularly either heavily arrival or departure biased. Respecting this balance will inevitably also restrict capacity growth, particularly in the periods of higher arrival demand, due to the restrictions caused by the arrival spacing (Table 16). Through progressively adding more traffic into the simulation, the runway capacity profile that could be contained within the agreed average 10 minute delay criteria for Runways 21/24 is shown in Table 17. The peak arrival utilisation is 25 movements and the peak departure utilisation is 39 movements. Peak capacity of 42 movements occurs in 3 hours of the day. The peak hours of 42 movements all occur in hours where the balance between arrival and departure movements is either close to a 50:50 balance, or where Page 78

79 departures are dominant. During the (UTC+8) hour there are 39 departures, whilst the and hours (UTC+8) contain 22 and 26 departures respectively. This demonstrates that the total number of movements possible within a particular hour within the current operational environment is restricted by the number of arrivals within that hour. Hour (UTC+8) Grown Arrival ATMs Grown Departure ATMs Table 17. Current capacity index profile for Runways 21/24. Grown Capacity Index ATMs Total The total number of movements across the day has to 590 movements. The breakdown of where additional movements have been accommodated is shown in Table 18. Page 79

80 Hour (UTC+8) Arrivals (%) Grown Arrival Movements Additional Arrival Movements Hour (AEST) Departures (%) Grown Departure Movements Additional Departure Movements Total 294 Total 296 Table 18. Placement of additional movements in the current capacity index profile for Runways 21/24. A total of 69 arrival movements were added to the observed average daily traffic profile across the whole day. Of these, only 27 (39%) were added during the arrival biased hours and (UTC+8). During periods where observed arrival traffic was already high (above or around 20 ATMs) there was limited potential for additional arrival movements; and hours (UTC+8). A total of 67 departure movements have been added to the observed average daily traffic profile across the whole day. Of these, 42 (63%) were added during the departure biased hours and (UTC+8). For departures, in contrast to arrivals, it was possible to grow traffic where observed departure traffic was already high, for example during and (UTC+8). By respecting the current balance between arrivals and departures, the total number of departures has been limited by the total number of arrivals possible across the day. As a result, it is again evident that the arrival delays are the limiting factor to the growth of the current capacity index. This is also clear when reviewing the simulated arrival and departure delays under this grown traffic profile, which are shown in Figure 52 and Figure 53 respectively. On both figures, the blue line represents the baseline simulated average arrival delay and the orange line is the simulated current capacity index. Page 80

81 Figure 52. Simulated arrival delay for the current capacity index profile for Runways 21/24. Figure 53. Simulated departure delay for the current capacity index profile for Runways 21/24. Average arrival delay was simulated to be 5.40 minutes over the 18 hour period (UTC+8), with the peak average arrival delay of minutes occurring during the (UTC+8) period. Average simulated departure delay is 2.31 minutes over the 18 hour period (UTC+8), with a peak average departure delay of 9.25 minutes occurring during the (UTC+8) hour. Page 81

82 Whilst the 10 minute average delay criteria is to a small degree exceeded for arrivals, during these peak hours the average delay for all ATMs (arrivals and departures) still remains below 10 minutes Runway 03 Current Capacity Index In section it was discussed how a single runway operating in a mixed-mode configuration would result in higher average delays when compared to a multiple runway operation. It would be expected that this would also limit the traffic growth for the current capacity index, against the same delay criterion. The current runway capacity index profile that could be contained within the agreed average 10 minute delay criteria for Runway 03 is shown in Table 19. The peak arrival demand is 22 movements and the peak departure demand is 37 movements. The peak demand is 40 movements, which occurs at 0600 (UTC+8), during the peak departure hour. This is two movements fewer than was seen during this hour for Runways 21/24. In addition, for Runways 21/24 this peak was matched during hours ( and UTC+8) where there was close to a 50:50 balance between arrivals and departures; this is no longer the case. These hours are now further restricted by the number of arrivals within those hours in the single runway operation. Page 82

83 Hour (UTC+8) Grown Arrival ATMs Grown Departure ATMs Table 19. Current capacity index profile for Runway 03. Grown Capacity Index ATMs Total The total number of movements across the day has increased to 524 movements for Runway 03. The breakdown of where additional movements have been accommodated for Runway 03 is shown in Table 20. Page 83

84 Hour (UTC+8) Arrivals (%) Grown Arrival Movements Additional Arrival Movements Hour (AEST) Departures (%) Grown Departure Movements Additional Departure Movements Total 261 Total 263 Table 20. Placement of additional movements in the current capacity index profile for Runway 03. A total of 36 arrival movements were added to the observed average daily traffic profile across the whole day. Of these, 21 (58%) were added during the arrival biased hours and (UTC+8). As with Runways 21/24, there is limited potential for additional arrival movements during periods where observed arrival traffic was already high. Specifically, no movements were added during the period where the Runway 03 baseline from Section already exceeded the 10-minute delay criterion, during the 1800 and 1900 hours (UTC+8). A total of 34 departure movements were added to the observed average daily traffic profile across the whole day. Of these, 20 (59%) were added during the departure biased hours and (UTC+8). It was not possible to grow the departure demand by the same magnitude during the peak departure periods ( and (UTC+8)) as it was for Runways 21/24, again due to departures being more restricted within this single runway configuration. The simulated arrival and departure delays under the grown traffic profile for Runway 03 are shown in Figure 54 and Figure 55 respectively. On both figures, the blue line represents the baseline simulated average arrival delay and the orange line is the simulated current capacity index. Page 84

85 Figure 54. Simulated arrival delay for the current capacity index profile for Runway 03. Figure 55. Simulated departure delay for the current capacity index profile for Runway 03. When comparing the simulated delays for the grown traffic profiles for Runway 03 against the baseline, it can be seen that the levels of delay do not always exceed that from the baseline. This is particularly clear during the evening peak period for arrival demand, during the and hours (UTC+8). Page 85

86 Differences in demand that were observed on individual days, and which exist in the average daily profiles in the baseline simulation, make the average arrival and departure delays more volatile. The baseline profile is comprised of a number of days of observed traffic demand. On some of those days during the 1800 hour there were high demand values that exceeded the service rate despite the average remaining at 21 movements. When the service rate was exceeded, large delays were observed. The arrival demand simulated in the current capacity profile at this time was also 21 movements but does not exhibit this variability. This explains why the current capacity profile results in lower delays. Average arrival delay is 4.45 minutes over the 18 hour period (UTC+8), with the peak average arrival delay of minutes occurring during the (UTC+8) period. Average simulated departure Delay is 2.93 minutes over the 18 hour period (UTC+8), with a peak average departure delay of 9.51 minutes occurring during the (UTC+8) hour. Whilst the 10 minute average delay criteria is to a small degree exceeded for arrivals, during these peak hours the average delay for all ATMs (arrivals and departures) still remains below 10 minutes. Page 86

87 8.4 Current Capacity Index Summary From the observations of the current operation at Perth Airport a peak hourly movement rate of 37 was achieved comprising 3 arrivals and 34 departures ( UTC+8). In assessing the capacity index under observed operational performance an agreed delay criteria of no more than 10 minutes average delay for arrivals and departures in any given half-hour period was applied. This led to the following key results : The current capacity index for Runways 21/24 was estimated to deliver 590 ATMs per day; The current capacity index for Ruwnay 03 was estimated to deliver 524 ATMs per day; The maximum achievable service rate using Runways 21/24 was estimated to be 42 movements. This was observed in the model in 3 separate hours; (UTC+8) where there is an arrival departure bias close to 50:50 and and where there is a bias towards departures; The maximum achievable service rate using Runway 03 was estimated to be 40 movements; During periods of high arrival demand, the maximum number of arrivals that can be handled in any given hour is 25 (based on the performance observed during the NATS visit); In the current capacity index profile for Runways 21/24, more than 30 ATMs are consistently handled each hour from (UTC+8), with 40 or more ATMs handled in six different hours across the day. Table 21 shows a comparison between the average observed hourly traffic profile and the current capacity index for Runways 21/24; Table 22 shows a comparison between the average observed hourly traffic profile and the current capacity index for Runway 03. It should be noted that no consideration has been given to the capability of the airspace, terminal, or taxiway infrastructure to meet the levels of demand simulated here. Furthermore, the proportion of night-time operations has remained consistent with the current operation. Page 87

88 Hour (UTC+8) Observed Current Capacity Index Arrivals Departures Combined Arrivals Departures Combined Total Table 21. Comparison between the average observed hourly traffic profile and grown current capacity index for Runways 21/14. Page 88

89 Hour (UTC+8) Observed Current Capacity Index Arrivals Departures Combined Arrivals Departures Combined Total Table 22. Comparison between the average observed hourly traffic profile and grown current capacity index for Runway 03. Page 89

90 9 Target Capacity Index In Section 7 a number of key recommendations for implementation in the target capacity index were identified. Specifically these recommendations related to the near-term opportunities to release the latent capacity of the airfield. In this section these recommendations are tested against the agreed 10 minute delay criterion and the capacity index re-evaluated to provide a target for future operations. 9.1 Target Capacity Assumptions ICAO Wake Turbulence Minima for Arrival-Arrival Separations Recommendation 7.1 for the target capacity index is to reduce the arrival-arrival separations to the ICAO wake turbulence minima at threshold. These values are shown in Table 14. For the purposes of presenting these separations to the HERMES simulation they have been translated into time separations. This is achieved using an average approach speed of 150kts, which has been established through many years of analysis in the UK and has been found to be robust to a wide variety of aircraft types. The ICAO minimum threshold spacing between pairs of wake turbulence constrained arrivals, translated into time separations, are presented in Table 23. Radar minimum spacing for non-wake turbulence constrained pairs of arrivals is assumed to be 3nm (72 seconds). Table 23. ICAO minimum threshold spacing between wake turbulence constrained arrivals, translated into time separations. All arrival-departure-arrival separations in both models have been simulated with a separation of 150 seconds, as described in recommendation 7.2. Page 90

91 9.1.2 Impact of Arrival Runway Occupancy Times on Reduced Arrival-Arrival Separations As discussed in Section 6.1.1, arrival runway occupancy times for some exit taxiways could constrain growth. For example, the ICAO wake turbulence minimum separation for a pair of medium wake turbulence category aircraft is 3nm, or 72 seconds. If the lead aircraft had an arrival runway occupancy time of 48 seconds or greater, then the following aircraft would be less than 1 nautical mile from the threshold (assuming around 24 seconds to cover 1nm on final approach) when the lead aircraft vacates the runway. This would be likely to cause the follower aircraft to have a late landing clearance and it may have to go-around if the arot for the lead aircraft was as high as some of those that were observed. The average arot for aircraft arriving on Runway 21 and vacating at taxiway L1 is 72.5 seconds (Section ). Similarly the average arot for aircraft arriving on Runway 24 and vacating at taxiway V is 68 seconds. On Runway 03, 66% of traffic uses either runway exit taxiway D1 or W, which have average arots of 70 and 95.6 seconds respectively. In these cases the arots would not support the level of reduction in the arrival spacing that was shown in Table 23 for either Runways 21/24 or Runway Runways 21/24 Table 24 shows the arrival-arrival spacing that can be achieved on Runways 21 and 24 without any reduction in arrival runway occupancy times. Table 24. Arrival-arrival separations presented to the HERMES simulation for the Runways 21/24 model. Average arot values for light aircraft are lower than those for other aircraft types, so a 72 second spacing can still be applied when a light aircraft is the leading aircraft type. Note that, where required, spacing has been increased in 1nm (24 second) increments. Arrival runway occupancy times on Runways 21 and 24 have been compared to those at Airport A which operates in a HIRO environment. This has shown that there are limited opportunities for improvements that would enable any further reductions in arrival spacing. The impact of implementing these changes on the average observed daily profile is considered in Section to demonstrate that departure delay is not negatively affected by reduced arrival spacing. The capacity index under the proposed spacing regime is then presented in Section Runway 03 Table 25 shows the arrival-arrival spacing that can be achieved on Runway 03 without reduction to the arrival runway occupancy times. Page 91

92 Table 25. Arrival-arrival separations presented to the HERMES simulation for the Runway 03 mixed mode model. Spacing for arrival pairs where the leading aircraft is either a medium or a heavy wake turbulence category aircraft have not been reduced as far as those in the Runway 21/24 configuration due to the large average arrival runway occupancy time for medium and heavy aircraft vacating at exit D1 and W. The impact of implementing these changes on the average observed daily profile is considered in Section to demonstrate that departure delay is not negatively affected by reduced arrival spacing. A preliminary assessment of the capacity index under this proposed spacing regime is then presented in Section In Section it was identified that arots on Runway 03 were higher than those at comparable airports that service consistently high levels of demand. In Section the impact of improving arrival runway occupancy on Runway 03 will be considered which leads to a further evaluation of the target capacity index in Section Arrival-Departure Balancing In order to address recommendation 7.3, changes were made to the configuration of the simulation to allow greater consideration of the arrival and departure queue lengths. Analysis of the simulation results showed little operational benefit arising from these changes, due to the shape of the traffic profile at Perth Airport. Each hour of the day is heavily biased towards either arrivals or departures, or alternatively is balanced with lower levels of traffic. As a result, any preference to favour either arrivals or departures in order to reduce overall delays provides little benefit with the current traffic profile Reduced Pilot Reaction Time In Section the opportunity to improve the consistency of the operation through reduced pilot reaction times was identified. Average pilot reaction times at both comparison Airport A and Airport B are around 6 seconds, whilst at Perth the average pilot reaction time is 11.7 seconds (Section 5.5.3). The effect of reducing average pilot reaction time has been considered in Section for Runways 21/24 and in Section for Runway 03. Page 92

93 9.2 Target Capacity Index Assessment for Runways 21 and Impact of Reduced Arrival-Arrival and Arrival-Departure-Arrival Separations for Runway Mode Runway 21 (Arrivals/Departures) and Runway 24 (Arrivals only) The simulated arrival and departure delays for Runways 21 and 24 are shown in Figure 56 and Figure 57. In both charts, the blue line is the simulated delay for the observed average daily traffic profile as shown in Section 8.2.1, while the black line is the simulated delay after implementing recommendations 7.1 and 7.2. Figure 56. Simulated arrival delays for Runways 21 and 24 following implementation of recommendations 7.1 and 7.2. It is clear that the reduced AA and ADA spacing have had a marked impact on arrival delay. Indeed, simulated average arrival delay over the 18 hour period (UTC + 8) has been reduced from 3.08 minutes to 1.34 minutes. Peak average arrival delay has been reduced from 8.91 minutes to 4.72 minutes. Page 93

94 Figure 57. Simulated departure delays for Runway 21 following implementation of recommendations 7.1 and 7.2. Furthermore, Figure 57 demonstrates that AA and ADA spacing can be reduced without having an adverse impact on departure delays Target Capacity Index for Runways 21 and 24 Following the implementation of the reduced AA and ADA spacing, the traffic has been grown using the process described in Section to obtain the target capacity index traffic profile shown in Table 26. The peak hourly arrival demand has been grown to 33 movements, with the peak hourly departure demand remaining at the current capacity index peak of 39 movements. The peak capacity hour in this profile is (UTC+8), with 55 movements, split 40:60 between arrivals and departures. In addition, two consecutive hours of 53 movements have been achieved in the simulation, between 1200 and 1359 (UTC+8). Each of these hours has an arrival-departure balance close to 50:50. It is noteworthy that this differs from the current capacity index (Section 8.3.2) where (UTC+8) had the highest number of movements and was heavily biased towards departures. While the assessment of the current capacity found arrival bias to be a limiting factor to traffic growth, in the target capacity assessment (following the implementation of recommendations reducing arrival spacing) departure demand becomes the limiting factor for further growth. Page 94

95 Hour (UTC+8) Grown Arrival ATMs Grown Departure ATMs Table 26. Target capacity index profile for Runways 21 and 24 Grown Capacity Index ATMs Total The total number of movements across the day has been increased to 696 movements. The breakdown of where these additional movements have been accommodated is shown in Table 27. Page 95

96 Hour (UTC+8) Arrivals (%) Grown Arrival Movements Additional Arrival Movements Hour (UTC+8) Departures (%) Grown Departure Movements Additional Departure Movements Total 348 Total 348 Table 27. Placement of additional movements in the target capacity index profile for runways 21/24 (relative to the observed traffic). A total of 123 arrival movements have been added to the observed average daily traffic profile across the whole day. Of these, 69 were added during the arrival biased hours and (UTC+8). In particular it has been possible to grow the traffic during periods where observed arrival traffic was already high. This is in contrast to the current capacity index profile (Section 8.3.2), where there was limited possibility for traffic growth in these peak hours. For example, the hour (UTC+8) was observed to have on average 21 arrival movements. In the current capacity index it was possible to grow this to 22 arrivals, but in the target capacity index it has been possible to increase this to 32 arrivals, within the defined delay criteria. A total of 119 departure movements have been added to the observed average daily traffic profile across the whole day. Of these, 66 were added during the departure biased hours and (UTC+8). The peak departure hour in the current capacity index profile (Section 8.3.2) of (UTC+8) has remained the peak hour in the target capacity index, with the number of departures unchanged. The simulated average arrival and departure delays associated with the traffic profile shown in Table 26 are shown in Figure 58 and Figure 59. In the charts below the blue and black lines represent the simulated delay based on the observed average daily traffic profile, before and after implementing the recommendations. The orange line is the average delay associated with the target capacity index profiles shown in Table 26. The green bars show the number of ATMs in each hour contained within the target capacity index profile. Page 96

97 Figure 58. Simulated arrival delay for the target capacity index profile for Runways 21 and 24. Figure 59. Simulated departure delay for the target capacity index profile for Runways 21 and 24. Average arrival delay over the 18 hour period (UTC+8) was simulated to be 3.70 minutes, with a peak average arrival delay of 8.85 minutes occurring during the (UTC+8) period. Page 97

98 Average departure delay over the 18 hour period (UTC+8) was simulated to be 5.24 minutes, with a peak of minutes in the period (UTC+8). Whilst this peak average departure delay exceeds the 10 minute average delay criterion to a small degree, the average delay for all ATMs (arrivals and departures) during this hour remains below 10 minutes Reduced Pilot Reaction Time In order to assess the potential benefits of reducing pilot reaction time, the target capacity index traffic demand has been simulated with the PRT reduced by 5 seconds, bringing it to a similar level to those observed at Airport A and Airport B. Figure 60 and Figure 61 show the simulated arrival and departure delays respectively. The orange line represents the simulated delay for the target capacity index profile as presented in Section and the green line represents the simulated delays associated with the target capacity index profile after reducing average pilot reaction time. Figure 60. Simulated arrival delay for the target capacity index profile for Runways 21 and 24, after reducing average pilot reaction time by 5 seconds. Page 98

99 Figure 61. Simulated departure delay for the target capacity index profile for Runways 21 and 24, after reducing average pilot reaction time by 5 seconds. The effect on the simulated arrival delay is negligible; however reducing average pilot reaction time has had a greater impact on departure delays. Simulated departure delay for the target capacity index profile for Runways 21 and 24 has been reduced from an average of 5.24 minutes to 5.04 minutes (across the 18-hour period , UTC+8), with peak departure delay brought down from minutes to 9.81 minutes. The reductions in average arrival and departure delays do not allow for the addition of any further movements to the target capacity index profile, however reduced pilot reaction time may lead to a more consistent and therefore resilient operation during periods of high demand. 9.3 Target Capacity Assessment for Runway Impact of Partially Reduced Arrival-Arrival and Arrival-Departure- Arrival Separations The simulated arrival and departure delays for Runway 03 after reducing arrivaldeparture-arrival separations to 150 seconds and arrival-arrival separations to those shown in Table 25, are shown in Figure 62 and Figure 63 respectively. The blue line represents the simulated delay based on the observed traffic levels, from Section 8.2.2, with the black line representing the same scenario after the reduced arrival spacing has been implemented. Page 99

100 Figure 62. Simulated arrival delay for Runway 03 following the implementation of recommendations 7.1 and 7.2. Figure 63. Simulated departure delay for Runway 03 following the implementation of recommendations 7.1 and 7.2. Simulated average arrival delay for the observed operation during the 18 hour period (UTC+8) has reduced from 3.64 minutes to 1.65 minutes, with peak arrival delay now below the 10 minute delay criterion to 6.91 minutes. Page 100

101 As with the Runway 21/24 model, recommendations 7.1 and 7.2 have had a limited effect on departure delay. The peak morning period (UTC+8) remains unchanged due to the low number of arrivals at this time. There is a reduction in average delay during the afternoon peak, when there is a higher frequency of ADA separations, which have now been reduced from the observed average of 179 seconds to 150 seconds as per recommendation 7.2. During the main arrival biased periods of the day ( and (UTC+8)) there is an increase in the average departure delay. This is due to the significant reduction in arrival delay seen in these hours as a result of the reduced AA and ADA separations. The simulation makes the most efficient use of the runway during these periods and reduces the average delay for all ATMs (arrivals and departures) by making departures wait slightly longer at the holding point in favour of arriving aircraft. Overall simulated departure delay over the 18 hour period (UTC+8) has increased from 2.72 minutes in the observed operation to 2.86 minutes after implementing the recommendations. Since arrival delay is currently the most constraining factor to growth in the current operation, the substantial reduction in arrival delay achieved through the reductions to AA and ADA spacing is considered to offset these minor increases in departure delay Preliminary Target Capacity Index for Runway 03 Following the implementation of the reduced AA and ADA spacing in Section 9.3.1, the traffic has been grown using the process described in Section to obtain the target capacity index traffic profile shown in Table 28. The peak arrival demand is 25 movements ( and UTC+8) and the peak departure demand is 38 movements ( UTC+8). The peak hourly movement rate has been grown to 42 movements in an hour at (UTC+8) with 17 arrivals and 25 departures. This is 13 movements fewer than was achieved in the Runways 21/24 target capacity index. Two consecutively busy hours between 1200 and 1359 (UTC+8) achieved hourly rates of 40 movements. Page 101

102 Hour (UTC+8) Grown Arrival ATMs Grown Departure ATMs Grown Capacity Index ATMs Total Table 28. Preliminary target capacity index profile for Runway 03. The total number of movements across the day has increased to 578 movements. The breakdown of where additional movements have been accommodated is shown in Table 29. Page 102

103 Hour (UTC+8) Arrivals (%) Grown Arrival Movements Additional Arrival Movements Hour (UTC+8) Departures (%) Grown Departure Movements Additional Departure Movements Total 289 Total 289 Table 29. Placement of additional movements in the preliminary target capacity index profile for Runway 03 (relative to the observed traffic profile). A total of 64 arrival movements have been added to the observed average daily traffic profile across the whole day. Of these, 37 (58%) were added during the arrival biased hours and (UTC+8). A total of 60 departure movements were added to the observed average daily traffic profile across the whole day. Of these, 39 (65%) were added during the departure biased hours and (UTC+8). As in the Runway 21/24 simulation, it has not been possible to increase the number of departures in the peak hours. For example, (UTC+8) remains the peak departure hour, with 38 movements in the target capacity index. The simulated average arrival and departure delays associated with the traffic profile in Table 28 are shown in Figure 64 and Figure 65 respectively. The blue and black lines represent the simulated delays based on the observed levels of traffic, before and after implementing the reduced separations. The pink line is the average delay associated with the preliminary target capacity index profile in Table 28. The green bars show the number of ATMs in each hour. Page 103

104 Figure 64. Simulated arrival delay for the preliminary target capacity index profile for Runway 03. Figure 65. Simulated departure delay for the preliminary target capacity index profile for Runway 03. Average arrival delay over the 18 hour period (UTC+8) was simulated to be 2.53 minutes, with a peak average delay of 9.44 minutes occurring in the hour (UTC+8). Average departure delay over the 18 hour period (UTC+8) was simulated to be 5.29 minutes, with a peak average delay of minutes Page 104

105 occurring in the hour (UTC+8). Whilst this peak average departure delay exceeds the 10 minute average delay criteria to a small degree, the average delay for all ATMs (arrivals and departures) during this hour remains below 10 minutes Impact of Reduced Arrival Runway Occupancy Times In Section it was identified that arots on Runway 03 were higher than those at comparable airports that service consistently high levels of demand. Indeed, over half of all aircraft observed arriving on Runway 03 had a runway occupancy time of 70 seconds or greater. Whilst the taxiway infrastructure around the J, A, and the Runway 06/24 intersection is clearly complex, the potential benefits of reduced arot on Runway 03 have been simulated here. For the purposes of modelling improved runway occupancy, the average arot for all medium and light aircraft has been reduced to 55 seconds which is comparable with Airport A and Airport B in high intensity operations. In addition, all heavy aircraft have been assumed to be capable of vacating via D1 (during the observation period 10% of heavy aircraft were seen to vacate at W). These reductions in runway occupancy would enable AA spacing to be reduced from those shown in Table 25 to those shown in Table 30 (equal to those simulated for Runway 21/24 and shown in Table 24). Table 30. Arrival-arrival separations used to simulate the potential benefits of reducing arot times on Runway 03. The effect of these changes to the simulated arrival and departure delays are shown in Figure 66 and Figure 67. The pink lines represent the delay lines presented in Section for the preliminary target capacity index. The purple line represents the delay for the same traffic profile, but with the reduced arots and AA spacing described above. Page 105

106 Figure 66. Simulated arrival delay for the preliminary target capacity index profile for Runway 03, with reduced arrival runway occupancy times and reduced arrival-arrival separations. Figure 67. Simulated departure delay for the preliminary target capacity index profile for Runway 03, with reduced arrival runway occupancy times and reduced arrivalarrival separations. These extra improvements have reduced the average arrival delay for the preliminary target capacity index for the 18 hour period (UTC+8) from 2.53 minutes to 2.15 minutes, whilst bringing the peak delay down from 9.44 minutes to 7.0 minutes. Significant reductions were also seen in the departure delay, with average delay over the 18 hour period reduced from 5.29 Page 106

107 minutes to 3.3 minutes. In addition peak departure delay has been reduced to 8.17 minutes, from minutes Target Capacity Index for Runway 03 The reductions in the arrival and departure delay simulated in Section have enable opportunities to further grow the traffic above the levels obtained in the preliminary target capacity index. Table 31 shows the target capacity index for Runway 03 that may be obtained through both reduced arots and AA spacing. Hour (UTC+8) Grown Arrival ATMs Grown Departure ATMs Table 31. Target capacity index profile for Runway 03. Grown Capacity Index ATMs Total The total number of movements across the day has now been increased to 622. This represents an increase of 44 movements on the preliminary target capacity index. Table 32 shows the breakdown of where additional movements have been added into the observed average daily traffic profile to create the target capacity profile. Page 107

108 Hour (UTC+8) Arrivals (%) Grown Arrival Movements Additional Arrival Movements Hour (UTC+8) Departures (%) Grown Departure Movements Additional Departure Movements Total 311 Total 311 Table 32. Placement of additional movements in the target capacity index for Runway 03 (relative to the observed traffic profile). When compared to the preliminary target capacity index profile, further traffic growth has been possible in every hour between 0500 and 2100, with the exceptions of the 0700 and 2100 hours (UTC+8). A total of 86 arrival movements have been added to the observed average daily traffic profile across the day. Of these, 53 (62%) were added during the arrival biased hours and (UTC+8). A total of 82 departures have been added to the observed average daily traffic profile across the whole day. Of these, 47 (57%) have been added during the departure biased hours and (UTC+8). The simulated average arrival and departure delays associated with the traffic profile in Table 31 are shown in Figure 68 and Figure 69 respectively. The purple line represents the simulated delay based upon the preliminary target capacity index traffic profile with the reduced arot and associated AA spacing, as shown in Section The yellow line represents the simulated delay accompanying the target capacity index profile presented in Table 31. The green blocks show the number of movements in each hour, in the target capacity index. Page 108

109 Figure 68. Simulated arrival delay for the target capacity index profile for Runway 03. Figure 69. Simulated departure delay for the target capacity index profile for Runway 03. Average arrival delay over the 18 hour period (UTC+8) for the target capacity index was simulated to be 3.39 minutes, with a peak average arrival delay of 8.63 minutes occurring during the (UTC+8) period. Average departure delay over the 18 hour period (UTC+8) for the target capacity index was simulated to be 5.24 minutes, with a peak average departure delay of 9.74 minutes occurring during the (UTC+8) period. Page 109

110 9.3.5 Reduced Pilot Reaction Time Operational Performance and Airport In order to assess the potential benefits of reducing pilot reaction time, the target capacity index traffic demand has been simulated with the PRT reduced by 5 seconds, bringing it to a similar level to those observed at Airport A and Airport B. Figure 70 and Figure 71 show the arrival delays and the departure delays from the simulation respectively. On these charts the yellow line represents the simulated delay for the target capacity index profile as presented in Section 9.3.4, the green line represents the simulated delays associated with the target capacity index profile after reducing average pilot reaction time. Figure 70. Simulated arrival delay for the target capacity index profile for Runway 03, after reducing average pilot reaction time by 5 seconds. Page 110

111 Figure 71. Simulated departure delay for the target capacity index profile for Runway 03, after reducing average pilot reaction time by 5 seconds. The effect of reduced pilot reaction time on arrival delay is negligible; however reducing average pilot reaction time has had a greater impact on departure delay. Simulated average departure delay for the target capacity index profile for Runway 03 over the 18 hour period (UTC+8) has been reduced from 5.24 minutes to 4.94 minutes. Peak delay has been reduced from 9.74 minutes to 9.2 minutes. As a result of these reductions in departure delay it is now possible to accommodate four additional movements in the target capacity index: An arrival at (UTC+8); An arrival at (UTC+8); A departure at (UTC+8); A departure at (UTC+8). 9.4 Target Capacity Index Summary From observation of the current operation at Perth Airport, a peak average hourly movement rate of 37 movements was achieved comprising 3 arrivals and 34 departures ( , UTC+8). Other departure-biased hours have similarly high movements. The peak average hourly movement rate in an arrival-biased hour was 30, comprising 21 arrivals and 9 departures ( , UTC+8). In Sections 9.2 and 9.3, the likely capacity index that would result from the implementation of the recommendations in Section 7 were assessed. In order to fully realise the potential of recommendation 7.1 for Runway 03, a substantial reduction in arrival runway occupancy times was required. As a result a preliminary target capacity index was produced detailing the benefits of Page 111

112 reducing separations with the current runway occupancy times, before reducing separations and runway occupancy times fully to assess the target capacity index. The following key results were obtained: The target capacity index for Runways 21/24 was estimated to deliver 696 ATMs per day; The preliminary target capacity index for Runway 03 (without reduced arrival runway occupancy) was estimated to deliver 578 ATMs per day; The target capacity index for Runway 03 was estimated to deliver 622 ATMs per day (this could be raised to 626 with improvements in pilot reaction time); The peak simulated achievable service rate for Runway 21/24 is 55 movements in an hour ( , UTC+8); The peak simulated achievable service rate for Runway 03 is 45 movements in a single hour ( , UTC+8) with reduced arot and 42 for the same hour without reduced arot; The peak simulated number of departures that was observed in the model in a single hour was 39 for Runways 21 and 24. This number can also be attained on Runway 03, but is reduced to 38 without reduced arot times. This occurred in the (UTC+8) hour for both modes. Whilst respecting the current arrival/departure balance and the need to have similar numbers of arrivals and departures in the day, the peak simulated number of arrivals in a single hour was 32 when operating Runways 21 and 24 ( , UTC+8), reduced to 28 when operating Runway 03 alone ( , UTC+8). This number is further reduced to 25 without the reduced arot values. The observed operation at Perth Airport allows for in excess of 20 movements per hour in every hour between 0600 and After recommendations this number can be increased to 27 each hour for Runway 03 operations before reducing arot values, 30 in each hour for Runway 03 operations with reduced arot s and at least 31 each hour for Runway 21/24 operations. Note that this figure includes hours where the operation is heavily arrival biased. In hours where it is departure biased or the arrival-departure balance is closer to 50:50, in excess of 40 movements can regularly be achieved when Runways 21 and 24 are in operation. A summary of the capacity indices for the two modes assessed is shown in Table 33 and Table 34 for Runways 21/24 and Runway 03 respectively. Page 112

113 Hour (UTC+8) Observed Current Capacity Index Target Capacity Index Arrivals Departures Combined Arrivals Departures Combined Arrivals Departures Combined Total Table 33. Comparison between the average hourly observed traffic profile, the current capacity index profile and the target capacity index profile for Runways 21/24. Page 113

114 Hour (UTC+8) Observed Current Capacity Index Preliminary Target Capacity Index Target Capacity Index Arrivals Departures Combined Arrivals Departures Combined Arrivals Departures Combined Arrivals Departures Combined Total Table 34. Comparison between the average hourly observed traffic profile, the current capacity index profile, the preliminary target capacity index and the target capacity index profile for Runway 03. Page 114

115 10 Glossary Of Terms ACC ACE A-CDM ADA AIP APC ASMGCS AMAN AMSL ATC ATCO ATFM ATM ATMs ATS CATC CCD CDA CONOPS CROPS EAT EASA EFPS EUROCONTROL FAA FL FLOPC FOD FIR FIS GMC HERMES HIRO IATA ICAO ILS LSA LCA LVP m MATS NM NOC NOTAM NVCR PANS-OPS Area Control Centre Airport Capacity Enhancement Airport Collaborative Decision Making Arrival Departure Arrival Air Pilot Document Approach Control Advanced Surface Movement Guidance and Control System Arrival Manager (ATFM position) Above Mean Sea Level Air Traffic Control Air Traffic Control Officer Air Traffic Flow Management Air Traffic Management Air Traffic Movements Air Traffic Services College of ATC Continuous Climb Departures Continuous Descent Approach Concept of Operations Crossing Runway Operations Expected Approach Time European Aviation Safety Agency Electronic Flight Progress Strips European Organisation for the Safety of Air Navigation United states Federal Aviation Agency Flight Level Flight Operations Performance Committee Foreign Object Debris Flight Information Region Flight Information Service Ground Movement Control/ler Heuristic Runway Movement Event Simulation High Intensity Runway Operations International Air Transport Association International Civil Aviation Organisation Instrument Landing System Localiser Sensitive Area Localiser Critical Area Low Visibility Procedure Metres Manual of Air Traffic Services Nautical miles National Operations Centre Notice to Airmen New Visual Control Room Procedures for Air Navigation Services Aircraft Page 115

116 PSS OSCET OJT RDP RPIG R/T RET RNAV RSL ROT SID SPC SRO STAR SOIR SMC SRO SVFR TAAM TCAS TDB TMA TOR TSF VCR VCF VFR VMC Operations Position Staffing Schedule Operational Safety and Capacity Enhancement Team On The-Job Training Radar Data Processing Runway Performance Improvement Group Radio Telephone Rapid Exit Taxiway Area Navigation Runway Slot Committee Runway Occupancy Time Standard Instrument Departure Slot Performance Committee Single Runway Operations Standard Instrument Arrival Simultaneous Operations on Parallel or Near-Parallel Instrument Runways Surface Movement Control Single Runway Operation Special VFR Total Airport and Airspace Model Traffic Collision Avoidance System Track Data Block Terminal Manoeuvring Area Terms of Reference Target Sector Flow Visual Control Room Virtual Contingency Facility Visual Flight Rules Visual Meteorological Conditions Page 116

117 11 References Operational Performance and Airport [1] PH TCU Local Instructions, Section [2] UK AIP, AD 2-EGKK-1-10, h. Minimum Runway Occupancy Time D/EG_AD_2_EGKK_en [3] UK AIP, AD 2-EGKK-1-15, 5. Detailed Procedures D/EG_AD_2_EGKK_en [4] MATS Chapter Wake Turbulence Heavy Wake Turbulence Separation [5] UK CAA CAP493 MATS Part 1. Section 1, Chapter 3, Page 12, Para &mode=detail&id=222 [6] Perth Airport HERMES Simulation Modelling, Current Operational Assumptions, Version 0.4 (6 th March 2012) Page 117

118 Appendix A (Aerodrome Chart) Page 118