Lisbon Airport Capacity Enhancement · bem como por me receberem e integrarem no 11º encontro de...
102
Lisbon Airport Capacity Enhancement Airspace capacity estimation and enhancement Jonas Amaral Rodrigues Thesis to obtain the Master of Science Degree in Aerospace Engineering Supervisors: Prof. Pedro da Graça Tavares Álvares Serrão Sr. Américo Gomes Dias de Melo Examination Committee Chairperson: Prof. Filipe Szolnoky Ramos Pinto Cunha Supervisor: Prof. Pedro da Graça Tavares Álvares Serrão Members of the Committee: Prof. António José Nobre Martins Aguiar Prof. Rodrigo Martins de Matos Ventura November 2014
Transcript of Lisbon Airport Capacity Enhancement · bem como por me receberem e integrarem no 11º encontro de...
Lisbon Airport Capacity Enhancement
Airspace capacity estimation and enhancement
Jonas Amaral Rodrigues
Thesis to obtain the Master of Science Degree in
Aerospace Engineering
Supervisors: Prof. Pedro da Graça Tavares Álvares Serrão
Sr. Américo Gomes Dias de Melo
Examination Committee
Chairperson: Prof. Filipe Szolnoky Ramos Pinto Cunha
Supervisor: Prof. Pedro da Graça Tavares Álvares Serrão
Members of the Committee: Prof. António José Nobre Martins Aguiar
Prof. Rodrigo Martins de Matos Ventura
November 2014
i
Agradecimentos
Em primeiro lugar, gostaria de agradecer ao meu orientador, Prof. Pedro Serrão, por toda a simpatia
e disponibilidade demonstradas ao longo da execução desta tese, bem como por me ter concedido a
oportunidade de trabalhar em colaboração com a NAV Portugal. Nesta, tive a oportunidade de
conhecer o meu co-orientador, Américo Melo, bem como Jesus Conde e Luís Martins, aos quais
também quero agradecer pelo à vontade com que me colocaram, por todo o apoio e
acompanhamento e, acima de tudo, pelo entusiasmo com que sempre encararam este trabalho.
Gostaria também de fazer referência a Ian Crooke, Kenneth Martin e Sandrine Molton da ISA
Software por todo o apoio e prontidão na resposta a alguns pedidos e dúvidas colocados por e-mail,
bem como por me receberem e integrarem no 11º encontro de utilizadores do RAMS, realizado em
Junho de 2014 em Madrid. Um enorme obrigado a António Ruas do EUROCONTROL, especialista no
método CAPAN, que tive oportunidade de conhecer em Madrid e sempre demonstrou total
disponibilidade e vontade de ajudar. Nas suas breves passagens por Lisboa tive oportunidade de me
reunir com ele, e inclusivamente trazê-lo ao IST, para mostrar os avanços no trabalho, esclarecer
dúvidas e conhecer o funcionamento de algumas funções mais complexas do RAMS.
Um agradecimento à Cátia, que fez uma revisão exaustiva do texto, e à Marta que apoiou na
correcção. Uma palavra também aos meus colegas e amigos com quem trabalhei e partilhei as
aventuras e desventuras de um curso de engenharia numa instituição como o IST.
E, como este trabalho marca o fim de uma etapa, queria agradecer a todos aqueles que me
acompanharam e apoiaram ao longo de todo este percurso académico, que se iniciou 20 anos antes,
a cerca de 1500 quilómetros de distância do local onde agora termina, na freguesia de Arrifes, ilha de
São Miguel. Um obrigado muito especial aos meus pais por tornarem tudo isto possível e por, em
todos os instantes, em todos os bons e maus momentos, me terem apoiado e proporcionado muito
mais do que aquilo que alguma vez poderia exigir. Uma palavra também para a minha irmã que se viu
privada da companhia e sábios conselhos do irmão ao longo de grande parte do seu crescimento.
Não esqueço também o apoio da minha madrinha, no mínimo tão entusiasta da aviação quanto eu. E,
obviamente, à Lili, por estes quase dois anos e meio de carinho, amizade e partilha, bem como por
toda a ajuda, paciência e incentivo para “trabalhar mais um bocadinho” ao longo das últimas semanas
deste desafio. Este trabalho é dedicado a todos vocês, na certeza de que sem o vosso apoio o
mesmo não teria sido possível. Muito obrigado!
ii
Resumo
Este trabalho tem como objectivo principal a estimativa da capacidade do espaço aéreo terminal do
Aeroporto de Lisboa, recorrendo ao software de simulação RAMS Plus. Foi realizado com a
colaboração da NAV Portugal.
Inicialmente, é feita uma abordagem aos conceitos de capacidade de um aeroporto e de um espaço
aéreo, bem como às cargas de trabalho dos controladores e sua relevância na quantidade de tráfego
que um sector pode receber. É também feita uma introdução ao mundo das simulações em tempo
real e acelerado de controlo de tráfego aéreo, bem como uma exposição do software RAMS Plus e
respectivo modo de funcionamento.
O ponto de partida para as simulações foi a configuração actual do espaço aéreo terminal do
Aeroporto de Lisboa, tendo para isso sido estudada em mais detalhe e reproduzida, tão
fidedignamente quanto possível, em dois cenários, uma para a pista 03 e outro para a 21, que viriam
a servir não só de referência mas também de calibração do modelo. As amostras de tráfego utilizadas
nas simulações baseiam-se no tráfego real do verão de 2013.
Com a capacidade actual estimada, analisaram-se então outros cenários alternativos, sempre
levando em conta as duas pistas mais utilizadas (03/21), chegando-se à conclusão de que com
algumas modificações no desenho actual dos sectores, seria possível aumentar a capacidade do
espaço aéreo.
Palavras chave: Capacidade, Carga de trabalho, Controlo de Tráfego Aéreo, Espaço Aéreo, RAMS
Plus, Simulação.
iii
Abstract
The aim of this work is to estimate the capacity of Lisbon Airport Terminal Airspace, using the
simulation software RAMS Plus. It was realized with the collaboration of NAV Portugal.
Initially, it is presented an approach to the concepts of airport and airspace capacity, as well as
controller workload and its relevance on the amount of traffic that an ATC sector can receive.
Afterwards, an introduction to the world of ATC real time and fast time simulations is also presented,
as well as an exposition of the software RAMS Plus and how it works.
The starting point of this work was Lisbon Airport Terminal Airspace current configuration. For this
reason it was studied in further detail and reproduced, as close as possible, in two scenarios, one for
RWY 03 and another of RWY 21, which would be used not only for reference but also for model
calibration. The traffic samples used in the simulations are based in real traffic from the summer of
2013.
With the current capacity estimated, alternative scenarios were analysed, always having in mind the
most used runways (03/21), having concluded that, with some modifications in the design of the
current sectors, it would be possible to increase the capacity of the airspace.
Keywords: Airspace, Air Traffic Control, Capacity, RAMS Plus, Simulation, Workload.
iv
Table of Contents
Agradecimentos .........................................................................................................................................i
Resumo .................................................................................................................................................... ii
Abstract.................................................................................................................................................... iii
Table of Contents .................................................................................................................................... iv
List of Figures ........................................................................................................................................ viii
List of Tables ........................................................................................................................................... xi
List of Acronyms ..................................................................................................................................... xii
1 Introduction ........................................................................................................................................... 1
1.1 Motivation and Objectives .............................................................................................................. 1
1.2 TMA overview ................................................................................................................................ 2
1.3 Methodology .................................................................................................................................. 2
1.4 Thesis overview ............................................................................................................................. 3
2. Airport Capacity ................................................................................................................................... 4
2.1 Runway Capacity ........................................................................................................................... 4
2.2 Airspace Capacity .......................................................................................................................... 5
2.2.1 Controller Workload ................................................................................................................ 6
2.2.2 Controller Positions ................................................................................................................. 8
3. ATC Simulations and Techniques ..................................................................................................... 10
3.1 ATC Simulations .......................................................................................................................... 10
3.1.1 Fast Time Simulations ........................................................................................................... 10
3.1.2 Real Time Simulations .......................................................................................................... 11
3.2 Capacity Estimation: CAPAN Method .......................................................................................... 11
4. RAMS Plus ........................................................................................................................................ 14
4.1 Airspace Sectors .......................................................................................................................... 14
4.2 Aircraft Performance .................................................................................................................... 15
4.3 Flights........................................................................................................................................... 15
4.4 Airports and Runways .................................................................................................................. 16
4.5 Routes and Holdstacks ................................................................................................................ 17
v
4.6 Controllers .................................................................................................................................... 18
4.6.1 Controller Windows ............................................................................................................... 18
4.7 Conflict Detection and Resolution ................................................................................................ 19
4.8 Tasks and Workload .................................................................................................................... 21
4.9 Stochastic Variations ................................................................................................................... 22
4.10 ATM Analyser ............................................................................................................................ 22
5. Lisbon’s Airport and Airspace ............................................................................................................ 23
5.1 Runways ...................................................................................................................................... 23
5.1.1 Runway Usage ...................................................................................................................... 23
5.1.2 Runway Occupancy Times ................................................................................................... 24
5.2 Airspace ....................................................................................................................................... 25
5.2.1 Sectors .................................................................................................................................. 26
5.2.2 Military areas ......................................................................................................................... 26
5.2.3 Routes ................................................................................................................................... 27
5.2.4 Separations ........................................................................................................................... 28
5.3 Traffic ........................................................................................................................................... 29
6. Simulations ........................................................................................................................................ 31
6.1 Reference Scenario RWY 03 ....................................................................................................... 32
6.1.1 Overview ............................................................................................................................... 32
6.1.2 Airspace ................................................................................................................................ 32
6.1.3 Routes ................................................................................................................................... 32
6.1.4 Runways, Taxiways and Gates ............................................................................................. 33
6.1.5 Traffic .................................................................................................................................... 33
6.1.6 Conflict Resolution Rules ...................................................................................................... 33
6.1.7 Workload ............................................................................................................................... 38
6.1.8 Simulation Procedures .......................................................................................................... 39
6.1.9 Results .................................................................................................................................. 39
6.2 LPPT_03_B .................................................................................................................................. 42
6.2.1 Overview ............................................................................................................................... 42
6.2.2 Results .................................................................................................................................. 43
6.3 LPPT_03_C ................................................................................................................................. 44
vi
6.3.1 Overview ............................................................................................................................... 44
6.3.2 Results .................................................................................................................................. 44
6.4 LPPT_03_D ................................................................................................................................. 47
6.4.1 Overview ............................................................................................................................... 47
6.4.2 Results .................................................................................................................................. 47
6.5 LPPT_03_E .................................................................................................................................. 50
6.5.1 Overview ............................................................................................................................... 50
6.5.2 Results .................................................................................................................................. 51
6.6 LPPT_03_F .................................................................................................................................. 52
6.6.1 Overview ............................................................................................................................... 52
6.6.2 Results .................................................................................................................................. 53
6.7 Reference Scenario RWY 21 ....................................................................................................... 56
6.7.1 Overview ............................................................................................................................... 56
6.7.2 Results .................................................................................................................................. 57
6.8 LPPT_21_B .................................................................................................................................. 60
6.8.1 Overview ............................................................................................................................... 60
6.8.2 Results .................................................................................................................................. 61
6.9 LPPT_21_C ................................................................................................................................. 62
6.9.1 Overview ............................................................................................................................... 62
6.9.2 Results .................................................................................................................................. 63
6.10 LPPT_21_D ............................................................................................................................... 65
6.10.1 Overview ............................................................................................................................. 65
6.10.2 Results ................................................................................................................................ 65
6.11 LPPT_21_E ................................................................................................................................ 67
6.11.1 Overview ................................................................................................................................. 67
6.11.2 Results ................................................................................................................................ 68
6.12 LPPT_21_F ................................................................................................................................ 68
6.12.1 Overview ............................................................................................................................. 68
6.12.1 Results ................................................................................................................................ 68
7. Discussion ......................................................................................................................................... 72
7.1 Scenarios operating with RWY 03 ............................................................................................... 72
vii
7.2 Scenarios operating with RWY 21 ............................................................................................... 74
7.3 Workload ...................................................................................................................................... 75
8. Design Proposal ................................................................................................................................ 76
8.1 Description ................................................................................................................................... 76
8.2 Improvements .............................................................................................................................. 77
9. Conclusions ....................................................................................................................................... 80
References ............................................................................................................................................ 81
Appendix A – Conflict Resolution Flowchart ......................................................................................... 86
Appendix B – Task List .......................................................................................................................... 87
viii
List of Figures
Figure 1: First Control Tower (Croydon, early 1920s) [8] ........................................................................ 1
Figure 2: Transport Capacities [16] ......................................................................................................... 6
Figure 3: ATCo System [25] .................................................................................................................... 7
Figure 4: Controller Workload [24] .......................................................................................................... 7
Figure 5: Example of ATM Validation Process [28] ............................................................................... 10
Figure 6: Typical output data from RAMS ............................................................................................. 13
Figure 7: CAPAN Regression Analysis ................................................................................................. 13
Figure 8: Sector construction................................................................................................................. 15
Figure 9: Flight Profile ........................................................................................................................... 16
Figure 10: Example of RAMS Airport scheme [42] ............................................................................... 17
Figure 11: SIDs (red), STARs (blue) and holdstacks (black) ................................................................ 18
Figure 12: Controller Windows working scheme [32] ............................................................................ 19
Figure 13: Conflict Detection Models [32] ............................................................................................. 19
Figure 14: Shape overlapping (no conflict) [32] .................................................................................... 20
Figure 15: Shape overlapping (conflict not detected) [10] ..................................................................... 20
Figure 16: Shape overlapping (conflict detected) [10] ........................................................................... 20
Figure 17: Lisbon Runways [48] ............................................................................................................ 23
Figure 18: Runway Usage ..................................................................................................................... 24
Figure 19: Lisboa TMA .......................................................................................................................... 25
Figure 20: Lisboa TMA sectorization ..................................................................................................... 26
Figure 21: Most significant military areas around Lisbon ...................................................................... 27
Figure 22: RWYs 03 / 35 SIDs (red) and STARs (blue) ........................................................................ 28
Figure 23: RWY 21 SIDs (red) and STARs (blue) ................................................................................. 28
Figure 24: Traffic in Lisbon by trimester (year 2013) ............................................................................ 29
Figure 25: Movements of July 2013 ...................................................................................................... 29
Figure 26: Entry and exit points - 1st of July (RWY 03) ........................................................................ 30
Figure 27: Entry and exit points - 12th of July (RWY 21) ...................................................................... 30
Figure 28: Sectorization of reference scenario RWY 03 ....................................................................... 32
Figure 29: SIDs and STARs of reference scenario RWY03 .................................................................. 32
Figure 30: Ground representation ......................................................................................................... 33
Figure 31: Detail of the STAR chart [47] ............................................................................................... 34
Figure 32: Short downind RWY 03 [58] ................................................................................................. 35
Figure 33: Extended downwind RWY 03 [58]........................................................................................ 35
Figure 34: Simulation detail - base turns ............................................................................................... 35
Figure 35: Short approach from west [58] ............................................................................................. 35
Figure 36: Long approach from west [58] .............................................................................................. 35
Figure 37: Short approach from east [58] .............................................................................................. 36
ix
Figure 38: Long approach from east [58] .............................................................................................. 36
Figure 39: Simulation detail - ADSAD and EKMAR sequencing ........................................................... 36
Figure 40: South approach direct to final [58] ....................................................................................... 36
Figure 41: South approach via EKMAR [58] ......................................................................................... 36
Figure 42: Simulation detail - Approach routes from south ................................................................... 37
Figure 43: STARs of reference scenario RWY 03 ................................................................................ 38
Figure 44: Graphic of key indicators – Lisboa APP - Reference Scenario RWY 03 ............................. 39
Figure 45: Regression analysis - Lisboa APP - Reference Scenario RWY 03 ..................................... 40
Figure 46: Graphic of key indicators - Lisboa TMA - Reference Scenario RWY 03 ............................. 40
Figure 47: Graphic of RWY movements - Reference Scenario RWY 03 .............................................. 41
Figure 48: Four sectors scenario - Sector's vertical profile ................................................................... 42
Figure 49: Graphic of key indicators – Lisboa APP 1 - Four Sector Scenario RWY 03 ........................ 43
Figure 50: Regression analysis - Lisboa APP 1 – Four Sectors Scenario RWY 03 ............................. 43
Figure 51: Graphic of key indicators - Lisboa APP 2 - Four Sector Scenario RWY 03......................... 44
Figure 52: Graphic of key indicators - Lisboa TMA lower - Four Sector Scenario RWY 03 ................. 44
Figure 53: Graphic of key indicators - Lisboa TMA upper - Four Sector Scenario RWY 03 ................. 44
Figure 54: Graphic of RWY movements – Four Sector Scenario RWY 03 ........................................... 44
Figure 55: Graphic of key indicators – Lisboa APP – Scenario with UMUPI holdstack RWY 03 ......... 45
Figure 56: Regression analysis - Lisboa APP – Scenario with UMUPI holdstack RWY 03 .................. 45
Figure 57: Graphic of key indicators – Lisboa APP – Scenario with UMUPI holdstack RWY 03 ......... 46
Figure 58: Graphic of RWY movements – Scenario with UMUPI holdstack RWY 03 .......................... 46
Figure 59: Graphic of key indicators – Lisboa APP –Reduced RWY occupancy time RWY 03 ........... 48
Figure 60: Regression analysis - Lisboa APP – Reduced RWY occupancy time RWY 03 .................. 48
Figure 61: Graphic of key indicators – Lisboa TMA –Reduced RWY occupancy time RWY 03 ........... 49
Figure 62: Graphic of RWY movements – Reduced RWY occupancy times RWY 03 ......................... 49
Figure 63: Minimum separation between departing aircraft (45 degrees divergence) [52] ................... 50
Figure 64: Minimum separation between departing aircraft (same track) [52] ...................................... 50
Figure 65: LPPT_03_E - New SIDs ....................................................................................................... 51
Figure 66: Detail of military areas .......................................................................................................... 51
Figure 67: Graphic of key indicators – Lisboa APP – New SIDs RWY 03 ............................................ 51
Figure 68: Regression analysis - Lisboa APP – New SIDs RWY 03 .................................................... 52
Figure 69: Lisboa APP North and South ............................................................................................... 53
Figure 70: Graphic of key indicators – Lisboa APP North – New APP sectors RWY 03 ...................... 54
Figure 71: Graphic of key indicators – Lisboa APP South – New APP sectors RWY 03...................... 54
Figure 72: Graphic of key indicators – Lisboa TMA – New APP sectors RWY 03 ................................ 55
Figure 73: Regression analysis – Lisboa TMA – New APP sectors RWY 03 ....................................... 55
Figure 74: Graphic of RWY movements – New APP sectors RWY 03 ................................................. 56
Figure 75: SIDs and STARs RWY 21 .................................................................................................... 57
Figure 76: Alternate STARs RWY 21 .................................................................................................... 57
Figure 77: Graphic of key indicators – Lisboa APP– Reference scenario RWY 21 .............................. 58
x
Figure 78: Regression analysis – Lisboa APP – Reference Scenario RWY 21 .................................... 58
Figure 79: Graphic of key indicators – Lisboa TMA– Reference scenario RWY 21 ............................. 59
Figure 80: Graphic of RWY movements – Reference Scenario RWY 21 ............................................. 59
Figure 81: Graphic of key indicators – Lisboa APP 1 – 4 sector scenario RWY 21 .............................. 61
Figure 82: Regression analysis – Lisboa APP 1 – 4 sector scenario RWY 21 ..................................... 61
Figure 83: Graphic of key indicators – Lisboa APP 2 – 4 sector scenario RWY 21 .............................. 62
Figure 84: Graphic of key indicators – Lisboa TMA lower – 4 sector scenario RWY 21....................... 62
Figure 85: Graphic of key indicators – Lisboa TMA upper – 4 sector scenario RWY 21 ...................... 62
Figure 86: Graphic of RWY movements – 4 sector scenario RWY 21 ................................................. 62
Figure 87: STARs of scenario LPPT_21_C........................................................................................... 63
Figure 88: Graphic of key indicators – Lisboa APP – LPPT_21_C ....................................................... 64
Figure 89: Regression analysis – Lisboa APP – LPPT_21_C .............................................................. 64
Figure 90: Graphic of key indicators – Lisboa APP – Reduced RWY times RWY21 ............................ 65
Figure 91: Regression analysis – Lisboa APP – Reduced RWY times RWY 21 .................................. 66
Figure 92: Graphic of key indicators – Lisboa TMA – Reduced RWY times RWY21 ........................... 66
Figure 93: Graphic of RWY movements – Reduced RWY times RWY 21 ........................................... 67
Figure 94: LPPT_21_E: New SIDs RWY 21 ......................................................................................... 67
Figure 95: Detail of military areas .......................................................................................................... 68
Figure 96: Graphic of key indicators – Lisboa TMA – New sectors RWY21 ......................................... 69
Figure 97: Regression analysis – Lisboa APP – New sectors RWY 21 ................................................ 69
Figure 98: Graphic of key indicators – Lisboa APP North – New sectors RWY21 ............................... 70
Figure 99: Graphic of key indicators – Lisboa APP South – New sectors RWY21 ............................... 70
Figure 100: Graphic of RWY movements – New sectors RWY 21 ....................................................... 71
Figure 101: Total Cycle [60] .................................................................................................................. 74
Figure 102: Nav221HS holdstack .......................................................................................................... 76
Figure 103: Graphic of key indicators - Lisboa TMA – Final design RWY 03 ....................................... 77
Figure 104: Regression analysis – Lisboa TMA – Final design RWY 03 .............................................. 77
Figure 105: Graphic of key indicators - Lisboa TMA – Final design RWY 21 ....................................... 78
Figure 106: Regression analysis – Lisboa TMA– Final design RWY 21 ............................................... 78
Figure 107: Graphic of key indicators - Lisboa APP_North - Final Design RWY 03 ............................. 78
Figure 108: Graphic of key indicators - Lisboa APP_South- Final Design RWY 03 ............................. 78
Figure 109: Graphic of key indicators - Lisboa APP_North - Final Design RWY 21 ............................. 78
Figure 110: Graphic of key indicators - Lisboa APP_South - Final Design RWY 21 ............................ 78
Figure 111: Graphic of RWY movements - Final design RWY 03 ........................................................ 79
Figure 112: Graphic of RWY movements - Final design RWY 21 ........................................................ 79
xi
List of Tables
Table 1: Runway Capacity Measures [21] ............................................................................................... 5
Table 2: EUROCONTROL workload thresholds [9] ................................................................................ 8
Table 3: Planning and Tactical Controller Roles [32] .............................................................................. 9
Table 4: Task Definition Example .......................................................................................................... 22
Table 5: ROTA by RWY ........................................................................................................................ 24
Table 6: ROTD key indicators [49] ........................................................................................................ 25
Table 7: Wake vortex separations ......................................................................................................... 28
Table 8: List of scenarios and characteristics ....................................................................................... 31
Table 9: Reference Scenario RWY 03 - Conflict resolution summary .................................................. 41
Table 10: Reference Scenario RWY 03 - Runway conflict resolution summary ................................... 41
Table 11: Reference Scenario RWY 03 - Conflict resolution impacts ................................................... 42
Table 12: Reference Scenario RWY 03 - Holdstack data ..................................................................... 42
Table 13: Holdstack usage - Scenario with UMUPI holdstack RWY 03 ............................................... 47
Table 14: LPPT_03_D - New RWY Occupancy Times ......................................................................... 47
Table 15: LPPT_03_C – Holdstack data ............................................................................................... 50
Table 16: Conflict Resolution Impacts - New APP sectors scenario ..................................................... 56
Table 17: Holdstack data - New APP sectors scenario ......................................................................... 56
Table 18: Reference Scenario RWY 21 - Conflict resolution summary ................................................ 60
Table 19: Reference Scenario RWY 21 - Runway conflict resolution summary ................................... 60
Table 20: Reference Scenario RWY 21 - Conflict resolution impacts ................................................... 60
Table 21: Reference Scenario RWY 21 - Holdstack data ..................................................................... 60
Table 22: Restricted alternative STARs usage ..................................................................................... 63
Table 23: LPPT_21_C - Conflict resolutions applied ............................................................................ 63
Table 24: LPPT_21_D - New RWY Occupancy Times ......................................................................... 65
Table 25: Conflict Resolution Impacts - New APP sectors scenario RWY 21 ...................................... 71
Table 26: Holdstack data - New APP sectors scenario RWY 21 .......................................................... 71
Table 27: Holdstack data – Final Design RWY 03 ................................................................................ 79
Table 28: Holdstack data – Final Design RWY 21 ................................................................................ 79
xii
List of Acronyms
AIP - Aeronautical Information Publication
ANSP – Air Navigation Service Provider
APP – Approach
ATC – Air Traffic Control
ATCo – Air Traffic Controller
ATM – Air Traffic Management
CAPAN – Capacity Analyser (EUROCONTROL Method)
CAST – Comprehensive Airport Simulation Technology
CTR – Control Zone
EAM – EUROCONTROL Airspace Model
EUROCONTROL – European Organization for the Safety of Air Navigation
FAA – Federal Aviation Administration
FL – Flight Level
FRLC – Flight crew reaction for line-up clearance;
FRTT – Flight crew reaction for take-off clearance;
FTS – Fast Time Simulation
ICAO - International Civil Aviation Organization
ILS – Instrument Landing System
kt - Knot
LUPT – Line-up time;
MTC – Maximum Throughput Capacity
PC – Planning Controller
PHCAP – Practical Hourly Capacity
RAMS – Reorganized ATC Mathematical Simulator
ROTA – Arrival Runway Occupancy Time
ROTD – Departure Runway Occupancy Time
RT – Radio/Telephone
RTS – Real Time Simulation
xiii
RWY – Runway
SID – Standard Instrument Departure
STAR – Standard Terminal Arrival Route
TAAM – Total Airspace & Airport Modeller
TC – Tactical Controller
TMA – Terminal Area
TOCD – Take-off clearance delivery
TOFT – Take-off roll time.
1
1 Introduction
1.1 Motivation and Objectives
At the beginning of the 20th century, when the Wright brothers made the first controlled flight and
changed forever the concept of travel, they did not file a flight plan nor asked permission to take off.
Their aircraft was the only one in the air. In the following years more aircraft were crossing the skies
but the traffic was still low enough for the pilots alone to take the necessary measures to avoid other
aircraft. With the amazing increase in the number of air movements, pilots gradually lost the capability
to assure all the measures needed to ensure the safety of the flight. Consequently, air traffic control
(ATC) began to evolve. First to keep aircraft apart during ground manoeuvers and watch them over
while in the air (to know their location and allow a fast intervention in case something went wrong) and
later to avoid mid-air collisions, ATC became fundamental [1].
While waving a red flag to authorize the take-off, those first controllers probably never imagined that
one day, in a not so far future, the sky would be congested with thousands of airplanes and, despite
the state of the art technology employed, airspace capacity would have to be pushed to its limits in
order to deal with the increasing traffic demand. Statistics say air traffic is doubling every fifteen years
[2]. All over the world, controllers are struggling to deal with more and more congested airspace,
especially in the terminal areas of some busy airports.
Figure 1: First Control Tower (Croydon, early 1920s) [8]
The case of Lisbon Portela Airport is no exception. With a growth rate of 3.1% per year between 2008
and 2012 in passengers numbers [3] and expecting a traffic increase of 13% between 2014 and 2019
[43], much has been said about the capacity of the airport and its supposed lack of capability to deal
with this ever-increasing demand made by airlines [4][5]. At a time when the construction of a new
airport seems definitely out of question for the upcoming years [6] and focus is on improving the
2
capacity of Portela Airport [7], it became more relevant than ever to analyse the capacity of the airport,
which is now limited by its terminal airspace1 (TMA) and Approach (APP) sectors [43].
With this in mind, the aim of this thesis is to perform a model based simulation of Lisboa TMA and
APP to analyse its current capacity and conditioning factors, as well as to evaluate possible
improvements that could increase these sectors performance.
1.2 TMA overview
TMAs are connections between airports and en-route airspace. In this confined part of the airspace,
and in short periods of time, aircraft are constantly changing altitude, speed and direction before
landing or after take-off.
The management of arrival streams into airports is a fundamental part of each Air Traffic Management
(ATM) system. In a period of, typically, 20 to 30 minutes, an aircraft descends from cruise flight level to
runway altitude. In this phase of the flight, the aircraft is more likely to have other aircraft in its
neighbourhood conditioning its own flight plan and execution, while dealing with reduced space and
time to compensate hypothetic deviations from its schedule. When bottlenecks occur in the arrival
system, there is a strong probability they will have an impact on other parts of the ATM system as well:
departures might have to be held on ground because the RWY is occupied by the intense (and
probably delayed) approaching traffic and en-route sectors that feed the TMA may face traffic backlog
due to the incapability of the referred sector to receive more aircrafts. Meanwhile, the ATC has to deal
with airlines preferred profiles, environmental restrictions which must be respected and often military
areas around airports reducing the volume of airspace available [26]. Unpredictable events, such as
missed approaches, also have to be considered.
In the light of all this, it is clear that TMAs are complicated to study, regarding not only its complexity
but also the various entities involved (airports, airlines, ATC, militaries, citizens around the airport,
etc.), and that new solutions focused on more efficient procedures are arduous to find.
1.3 Methodology
The present study will follow the guidelines from the European Organization for the Safety of Air
Navigation (EUROCONTROL) for airspace design [27], specifically the ones for terminal areas
(chapter 5).
A model based simulation will be presented using an air traffic mathematical simulator: RAMS Plus.
The study consists of three phases: the first phase aims at reproducing the current Lisbon airport
scenario, with the intent of calibrating the model and using it as a benchmark against which the new
1 Terminal Airspace (EUROCONTROL) or Terminal Control Area (ICAO).
3
scenarios can be compared. The second phase will evaluate 12 distinct scenarios, 6 for each of the
analysed runways, while taking into account different changes for the third phase, which, combined
with the improvements found in the second phase, will result in the proposal of a final scenario (third
phase).
1.4 Thesis overview
The present document is organized in the following way:
Chapter 1 – Introduction: the aim of this thesis is presented and a contextualization of its subject-
matter is done. There is also an introduction to the complex environment of a TMA sector.
Chapter 2 – Airport Capacity: prevailing definitions of capacity are presented and discussed. There
is also reference to the controller workload subject.
Chapter 3 – ATC Simulations and Techniques: fast time and real time simulations are introduced
and there is an overview of different software to perform fast time simulations. A synopsis of the
CAPAN Method, a methodology used by EUROCONTROL to estimate sector capacity, is also
presented.
Chapter 4 – RAMS Plus: description of the software employed in this thesis.
Chapter 5 – Lisbon Airport: The current situation of Lisbon Portela Airport is reviewed, combined
with a description of present-day airspaces and routes. Traffic flows are also analysed based on
current traffic samples.
Chapter 6 – Simulations: In this chapter the simulations performed are presented. For the Reference
Scenario, a detailed explanation of the ATC system constructed is offered. For the other scenarios, an
explanation of the differences between them and the reference scenario is given.
Chapter 7 – Discussion: The results of the previous chapter will be analysed and interpreted
Chapter 8 – Design Proposal: From the knowledge and results obtained from chapters 6 and 7, a
scenario aggregating the improvements obtained is proposed.
Chapter 9 – Conclusions: Conclusions and future work.
4
2. Airport Capacity
Airport capacity refers to its ability to accommodate aircraft and is expressed in the number of
movements (arrivals and departures), typically in the period of one hour (movements per hour) [12]. It
is conditioned by four distinct elements: airspace, airfield, terminal and ground access [11].
Despite being a commonly used term, the exact definition of “airport capacity” is disputable. The
reason for that is simple: not only is it conditioned by different elements, but it also depends on various
aspects such as weather conditions (wind, visibility…), traffic (type of aircraft, mixed operations...) and
even the proficiency of air traffic controllers (ATCos) and pilots working at the moment. Some of these
factors vary with time, making capacity time-variable, and others cannot even be measured
quantitatively [14].
It is unquestionable that every airport has a limited capacity and, if traffic demand exceeds that
capacity for a certain time, greater delays occur as a consequence. However, there is a curious
relation between capacity and delays: usually, priority is given to the arriving flights and, in the gaps
between them, the departures. To use this gap between arrivals, the controller needs a queue of
departing aircraft waiting close to the runway ready for take-off. This results in a paradox: if an airport
works above its capacity it produces delays, but delays are needed so the airport can explore its entire
capacity [14]. Thus, a question is raised: what is an acceptable delay and what is the limit above which
the airport is overloaded.
In early 1960’s, the Federal Aviation Administration (FAA) proposed the Practical Hourly Capacity
(PHCAP), an estimate of the number of movements that could be performed without exceeding an
average delay of four minutes. Nowadays, major airports have considerable higher delays during
peak-traffic periods [15], making this method outdated.
Taking into account the difficulties to define the capacity of an airport, and regarding the work that will
later be described, two distinct capacities will be considered: airspace capacity and runway capacity.
2.1 Runway Capacity
There are a number of different techniques to evaluate the capacity of a runway system. Some of the
most common measures are the following: Maximum Throughput Capacity (MTC), Practical Hourly
Capacity (PHCAP), Declared Capacity and Sustained Capacity [15]. The following table resumes
these methods.
5
Definition % of MTC
MTC Number of movements performed per hour.
100%
PHCAP Number of movements performed per hour, with an average delay of 4 minutes.
80-90%
Declared Capacity Number of movements per hour with a reasonable level of service.2
85-90%
Sustained Capacity
Number of movements per hour that can be reasonably sustained over periods of several hours.
90% with good weather MTC;
100% with bad weather MTC.
Table 1: Runway Capacity Measures [21]
It is important to notice that, as mentioned previously for the case of an airport, runway capacity is no
more than a probabilistic quantity depending on the circumstances involved. The following factors can
be considered as the most influent ones:
Number and geometric layout of the runways;
Separation requirements between aircraft imposed by the ATM system;
Mix of movements using the runway (arrivals only, departures only or mixed), sequencing of
movements as well as traffic type;
Type and location of taxiway exits from the runway(s).
MTC is the fundamental measure of the capacity, simply quantifying the number of movements that
can be performed in one hour if the runway is used to its maximum potential, without violation of ATM
rules. It also employs some assumptions [22]:
There is a continuous aircraft demand for arrivals and/or departures;
Static fleet mix (i.e. runway is solicited by a similar mix of aircraft types along time);
Procedures do not change.
All things considered, MTC gives an average capacity of a runway, independent of concerns with the
level of service provided.
2.2 Airspace Capacity
For most airports dealing with capacity issues, but not limited by their runway(s), it is usually the
airspace that constitutes a major constraint.
2 Level of service is a quality measure of the service provided to the traffic. In the case of air traffic, the level of service is highly related with delays.
6
The capacity of an ATC sector depends on a whole chain of characteristics such as shape, connection to adjacent airspaces, the function of the airspace itself, traffic type, etc. [17] However, in areas with high traffic density, the capacity is not determined uniquely by spatial and aircraft performance constraints. Unlike road or rail transportation, the capacity of an ATC sector is also conditioned by a different factor: ATCo workload (
Figure 2: Transport Capacities [16]).
Figure 2: Transport Capacities [16]
Taking into account that experience in Europe usually shows airspace capacities determined by ATCo
workload, airspace capacity (or capacity of an ATC sector) can be defined as the maximum number of
aircraft that can be controlled in the sector during a period of time, while still allowing an acceptable
level of controller workload [13]. With this in mind, ATCo tasks and procedures become the most
significant factors for the analysis, although airspace characteristics cannot be neglected, since they
affect the complexity of the sector [18][19][20]. Because of its relevance, the controller’s workload
subject will be addressed with more detail further ahead.
2.2.1 Controller Workload
ATCos work in a complex person-machine environment in which they are subject to multiple demands
and tasks over time. The workload in response to those solicitations will be a function of their
capacities and of what they must do in order to maintain a safe and expeditious traffic flow [23].
Although it can be measured and estimated, workload is affected by a complex interaction of [13]:
Airspace situation: features of both sector and air traffic;
State of the equipment: design, reliability, accuracy;
CAPACITY
Veichles per hour
CAPACITY
Veichles per hour
CAPACITY
Aircraft per hour
ATCo Workload
ATC sector characteristics
ATC traffic characteristics
Street / junction
geometry
Traffic flow mixture
aggregates
Signal system
geometrics
Types of vehicles / services
7
State of the controller: age, mood, experience, decision-making strategies, etc.
All things considered, we end up with a complex system around the ATCo (Figure 3).
Figure 3: ATCo System [25]
The major tasks to be performed by controllers in a sector consist of standard communications, traffic
monitoring, traffic evolution prediction and detection and resolution of conflicts. In Figure 4 we have
the relation of these tasks (the reader is advised to take into account that the workload values are only
indicative).
Figure 4: Controller Workload [24]
It is clear that the workload associated with standard communications, monitoring and trajectory
prediction is close to a linear function of the number of aircraft in sector but, when the scenario
8
includes conflicts, the workload increases exponentially. This can be observed in some real-time
simulations where even a small increase of traffic density (e.g. 2%) can turn the working atmosphere
from relaxed to chaotic [24].
As mentioned before, the workload can be estimated. This typically occurs through empirical
experiences, by measuring the time controllers take to perform their tasks. With that information, it is
possible to estimate the working time of a controller during one hour, simulating the control positions.
Then, once again with empirical experimentation, these numbers can be turned into qualitative values.
For EUROCONTROL, workload shall not exceed the 70% threshold, i.e., the tasks carried out by a
controller shall occupy him for a maximum of forty two minutes per hour. Table 2 aggregates workload
thresholds and interpretations for EUROCONTROL.
Threshold Interpretation Working time per hour
70% or above Overload 42 minutes +
54% – 69% Heavy Load 32 – 41 minutes
42% - 53% Medium Load 25 – 31 minutes
18% - 41% Light Load 11 – 24 minutes
0% - 17% Very Light Load 0 – 10 minutes
Table 2: EUROCONTROL workload thresholds [9]
2.2.2 Controller Positions
Although there are differences from centre to centre, the most common configuration of ATC position
implies two controllers per sector: one planning controller and one tactical controller [1]. Despite
having different tasks and responsibilities, both work within the same purpose: to guide the traffic
crossing their sector in a safe and efficient manner.
The planning controller has to manage the entries (and exits) of aircraft into (and from) his sector,
coordinating with adjacent sectors. Before an aircraft enters his sector, the planning controller receives
a request with an entry point and a flight level. Considering the traffic already in the sector and other
aircraft approaching it, the controller then accepts or makes a counter-proposal until agreement is
attained with previous sector. Similarly, he coordinates aircraft leaving his airspace with the
subsequent sector [32][41].
While the planning controller communicates with other sectors, the tactical controller (also known as
radar controller, radar executive or executive controller [1]) contacts directly with pilots. He gives
instructions to assure separation between aircraft in his sector and to achieve exit conditions arranged
by the planning controller [41].
Table 3 resumes the roles of both controllers.
9
Planning Controller Tactical Controller
Aircraft considered
Aircraft entering into and exiting from sector
Only aircraft within the sector
Time Frame 15-20 minutes look ahead. 5-10 minutes look ahead
Coordination requirements
With TC within sector and PCs of adjacent sectors.
Primarily with PC of the sector
Prime Task Agree entry/exit conditions with adjacent sectors.
Maintain aircraft separation distances.
Table 3: Planning and Tactical Controller Roles [32]
10
3. ATC Simulations and Techniques
3.1 ATC Simulations
In order to estimate a sector’s capacity, reproduction of the ATC environment is needed. To do that,
Air Navigation Service Providers (ANSP) resort to two types of simulations: Fast Time Simulations
(FTS) and Real Time Simulations (RTS) [28]. When changes are made to the ATC system, different
validation methodologies can be used:
Figure 5: Example of ATM Validation Process [28]
3.1.1 Fast Time Simulations
To calculate controller workload and estimate sector capacities, ANSPs frequently use computer
models to simulate the ATC system or, in other words, fast time simulation (FTS) techniques, which
consist in an attempt to model the entire ATC system in a mathematical and logical language,
describing the behaviour of all parties involved in the process [39]. This represents a cost-effective
method to provide support for high-level decisions that would require longer time and greater financial
costs to be simulated in actual environment, making it possible to evaluate the benefits of both
physical and procedural modifications on the airport and/or airspace in a relatively short period of time
and at a considerable lower cost. From the simulation outputs, it is usually obtained a formula between
the number of entries in the sector and the controller workload [1].
Nowadays, there is a considerable number of fast-time simulation tools available. EUROCONTROL,
on its Experimental Centre, currently uses four different software [34]:
AirTOp – Air Traffic Optimization, from Airtopsoft [35];
RAMS Plus – Re-organized ATC Mathematical Simulator, from ISA Software [36];
TAAM – Total Airspace & Airport Modeller, from Jeppesen [37];
CAST – Comprehensive Airport Simulation Technology, from Airport Research Centre GmbH
[38].
11
Other methods worth mentioning are the Sector Design and Analysis Tool (SDAT), developed by FAA
and the Performance and Usability Modeling in ATM (PUMA), from UK’s National Air Traffic Services
(NATS). The last one is the unique model which considers controller’s cognitive aspects [32], but it is
not fully validated for use [1].
There are, though, some limitations about FTS, because it is unprovided of human aspects like
judgement and thinking (with the exception of PUMA). There is the consciousness that workload
experienced by controllers cannot be explained purely as a function of the amount of traffic they
handle [1]. At this phase, there is the need to resort to real time simulations.
3.1.2 Real Time Simulations
Due to the lack of human elements relatively to FTS, there is the need to perform Real-Time
Simulations (RTS). As the name suggests, the simulation process takes the same time as the real
operation in the ATC system [39].
In RTS the operational environment is replicated as close as possible to the reality, including the new
technology or procedure(s) to be tested. Professional controllers perform their functions in positions
which represent the actual controller positions in the ATC system and there are even pilots available
for communication.
The main disadvantage of RTS is the cost, since it requires trained personnel, equipment and
significant simulation time. Nevertheless, they are fundamental to estimate with great accuracy the
impact of new technologies and procedures on controller workload and capacity [1]. EUROCONTROL
uses three different platforms (ESCAPE, eDEP and MCS) to validate complex airspace organizations,
new tools and concepts [40].
3.2 Capacity Estimation: CAPAN Method
The Capacity Analyser (CAPAN) method was developed by EUROCONTROL to evaluate sector
capacities and new airspace designs. Initially, CAPAN was integrated in the software
EUROCONTROL Airspace Model (EAM) as a module to estimate sector capacities but, being the
EAM an outdated software from the eighties, it was combined by the EUROCONTROL CAPAN Team
with the RAMS Plus software. Nowadays this method is still used with RAMS Plus, with
EUROCONTROL using a version of RAMS with the necessary modifications. With RAMS working as a
simulation tool to determine the workload generated in a sector for a given traffic sample, CAPAN
generates values expressing the loading in the simulated control positions. From the empiric
experience, these values can be classified according to Table 2.
With CAPAN, the capacity of an ATC sector is considered to be the maximum number of aircraft that
can enter into it, in the period of an hour, without exceeding an acceptable level of workload for the
controller (i.e. the 70% threshold) [9]. With this in mind, it is necessary to attain a formula that
12
correlates the workload as a function of the number of sector entries. The method employed is as
follows:
1. The sector and its procedures are reproduced in RAMS;
2. An appropriate3 traffic sample is added to the simulation;
3. RAMS runs the simulations and generates traffic flows and task events;
4. Flight data and workload are registered;
5. Analysing the data, it is possible to relate workload values with the number of sector entries;
6. Plotting the values obtained for the 24 hour period and performing a regression analysis, the
function 𝑊𝐿 = 𝐹(𝑁𝑠𝑒𝑐𝑡𝑜𝑟 𝑒𝑛𝑡𝑟𝑖𝑒𝑠) is obtained;
7. The capacity value is obtained intersecting the obtained function with the 70% threshold.
Some important aspects to refer about the process:
CAPAN uses the busiest or the most constrained scenario to simulate the worst combination
expected;
There is an unregulated demand of traffic and resources, being possible to record values over
the 70% limit;
The data used shall be the average of a series of simulation runs (10 to 25 iterations typically).
This way, RAMS performs simulations with different entry times and slightly changes in the
performance of the aircraft;
The 18 minutes of “free” time left from the 70% threshold are available for non-objective tasks
like radar surveillance or recuperation. RAMS only records workloads of objective tasks.
Figure 6 shows the data that is usually extracted from RAMS for the CAPAN Analysis. The blue line
represents the Tactical Controller (TC) workload, whereas the green one shows the number of sector
entries and the brown one the number of conflicts, all in the last 60 minutes. For example, at 20:50 the
TC workload recorded is 45% and the number of entries is 23. This means that from 19:50 to 20:50 23
flights have entered the sector and the controller had a workload of 45% (27 minutes). The yellow line
shows the instantaneous number of flights inside the sector. Additionally, there might be a line similar
to the blue one (TC workload) related to the Planning Controller (PC). Regarding the fact that, when
workload is the limiting factor, that workload is from the TC [1][29], it was opted for omission of the
information.
3 A traffic sample consistent with the traffic observed in the sector: arrivals, departures, aircraft types, etc. Ideally a real traffic sample.
13
Figure 6: Typical output data from RAMS
Bellow, Figure 7 displays the regression analysis with workload as a function of the number of sector
entries.
Figure 7: CAPAN Regression Analysis
Having obtained the equation, it is trivial to estimate the capacity of this sector:
𝑊𝐿 = 0.0202𝑁𝑠𝑒𝑐𝑡𝑜𝑟 𝑒𝑛𝑡𝑟𝑖𝑒𝑠2 + 1.2118𝑁𝑠𝑒𝑐𝑡𝑜𝑟 𝑒𝑛𝑡𝑟𝑖𝑒𝑠 ,
with the 70% threshold and solving the equation for 𝑁𝑠𝑒𝑐𝑡𝑜𝑟 𝑒𝑛𝑡𝑟𝑖𝑒𝑠 it is equal to:
𝑁𝑠𝑒𝑐𝑡𝑜𝑟 𝑒𝑛𝑡𝑟𝑖𝑒𝑠 = 36.1 .
As a result, for this sector, the estimated capacity is of 36 flights per hour.
As it is easily observed in Figure 7, when traffic demand is low the workload almost resembles a linear
function of sector entries. Consequently, if the traffic sample used for the simulation is considerably
below the capacity of the sector, the capacity obtained may be too high, since the equation obtained
from the regression will only represent the linear behaviour. In such situation, it is necessary to re-
simulate with a more intense traffic sample.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Example Scenario- Runway 00X Flights - 20 iterations
TC (%) Flights Peak Occ. Conflicts
y = 0,0202x2 + 1,2118xR² = 0,978
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35 40
Wo
rklo
ad (
%)
No. Sector Entries
Example Scenario- Runway 00X Flights - 20 iterations
TC (%) Polynomial (TC (%))
14
4. RAMS Plus
In the early nineties, EUROCONTROL started the development of a simulation tool: the Reorganized
ATC Mathematical Simulator (RAMS) [9]. This software would later become a commercial product
being further developed by ISA Software. Its current designation is RAMS Plus. It works as a fast-time
simulation model and it is used worldwide by some of the biggest ANSPs to measure a wide variety of
ATM parameters such as workload, delays, capacity of the airport and airspace [10].
RAMS works as a discrete-event simulation model of the ATC system, providing a high fidelity model.
At the beginning, RAMS was a runway-to-runway simulator but, in 2003, with the release 5.00 a
groundside module was added, upgrade which improved the definition of RAMS to a gate-to-gate
model [31]. The most recent version is the 6.00, launched in December 2013 and its features include
4D flight profile calculation, 4D sectorization, 4D spatial conflict detection, conflict resolution process
based on a rulebase, 4D resolution manoeuvring, workload assignment, time-based metering, TMA
runway/holdstacks and airport ground movements [10]. With extensive outputs, it is possible to extract
information about a variety of parameters. Three stand out as the most important ones: flight events,
sector/controller workload and conflicts [30].
4.1 Airspace Sectors
In RAMS, the air traffic, imitating reality, is simulated through an airspace divided in ATC sectors.
Each sector is defined by airspace corners and segments forming a polygon to which a height is
given, generating a 3D space and thus creating a portion of airspace. The cross-sectional area of an
airspace frequently changes with altitude: an APP sector, for example, usually excludes a portion of
volume on its lower altitude corresponding to the airport control zone (CTR). When this occurs, the
sector can be defined by different layers. Other information like separation minima and restricted
zones shall be added to the simulation to make the model as close as possible to the actual ATC
system.
Each sector is assigned to a control centre with an associated schedule. This provides a fourth
dimension (time) and gives the option of a re-sectorization functionality: different sectorizations are
possible to schedule so that existing sectors are collapsed or expanded. Resectorization allows the
airspace to be divided into more and smaller sectors during rush hours, as well as less and bigger
sectors for the periods of low demand. It is important to notice one thing about the resectorization:
RAMS does it exclusively based on scheduling, not on dynamic conditions [10][32].
15
Figure 8: Sector construction
4.2 Aircraft Performance
The standard aircraft database of RAMS has 300 different aircraft. Each aircraft model is
characterized by two aspects: aircraft group and performance group.
Aircraft groups define the wake vortex category (light, medium, heavy and helicopter) and the
respective wake turbulence separation minima of each group. Performance groups, as the name
suggests, defines the performance of the aircraft. For in-flight, the performance is associated with level
bands and for each one there is a climb rate, descent rate, climb speed, cruise speed and descent
speed. The behaviour of the aircraft on the ground is also characterized by acceleration and
deceleration rates (for dry and wet cases), touchdown and lift off speeds, runway block time after take-
off and before landing as well as normal and fast exit speeds. Additionally, it is possible to define the
fuel consumption of the aircraft in each level band and flight phase (climb, cruise and descent).
The information contained in this database, like others in RAMS, can be manipulated. It may be
upgraded with more precise values, new models or performance groups can be added, new level
bands, etc. [10]
4.3 Flights
Each flight is defined by an entry time in the simulation, entry level, cruise level, exit level, aircraft
model, departure and destination airports and route (or a list of navaids). For a TMA scenario, the
flight shall be assigned with a runway and SID or STAR. When pretending to perform a gate-to-gate
simulation, there is an option for definition of gates and taxiways. It is also possible to clone existing
16
flights using the tool “Traffic Generation”, which is particularly useful when creating scenarios with
possible traffic increase [10].
Figure 9: Flight Profile
4.4 Airports and Runways
Airports are defined in the scenario by a latitude/longitude location. They can be defined with more or
less detail depending on the simulation’s aim. A fully detailed airport will contain runways, taxiways
and gates, being possible to simulate ground movements like taxi and pushback.
Runways are assigned to the respective airports. They are defined by two pairs of coordinates (start
and end) and can be adjusted for a given heading and length. The touchdown marker is also defined
and can be connected with a stochastic distribution to reproduce the uncertainty of touchdown
location.
Runway occupancy times are defined by one of the following parameters (top to lowest):
Acceleration/Deceleration rates – in this case RAMS uses aircraft performance to measure the
time it stays on the runway. Although this is the most realistic method, it has an inconvenient: when an
aircraft lands it decelerates to exit speed as fast as possible and continues taxiing on runway until the
exit. As a consequence, if the aircraft takes an exit near the end of the runway, with a maximum exit
speed of 15kt, it will have a landing run at that speed after landing and decelerating until arriving at the
exit node, largely increasing the occupancy time.
Distribution by performance group – a time distribution is given for each of the above-
mentioned performance groups.
17
Default time distributions – two time distributions are assigned to each runway (one for
landing, another for take-off) regardless of the aircraft.
Each runway has its own schedule. Before landing or take-off, a flight requests runway reservation
and a time slot is assign to it. If a conflict is raised during the schedule reservation, runways will resort
to their own rulebase for conflict resolution. Typically, this consists on reducing speed of arrival flights,
holding the departure on ground, alternating route for the arrival and dispatching of the arrival to a
holdstack assigned to that runway. For airports with runways that can operate simultaneously, there
might be the option of sending the flight to an alternative runway.
Another feature related to the runways is “runway dependencies”, which gives room for the insertion of
highly configurable time separations, depending on the various aspects of the operation. For example,
it is possible to impose a one hundred and twenty seconds separation between departures. As a
result, the runway becomes unavailable in the two minutes period after a take-off for other departures.
Nevertheless, the runway can receive landings during that time. The restrictions can also depend on
the routes flown, since runways have assigned SIDs and STARs [10].
Figure 10: Example of RAMS Airport scheme [42]
4.5 Routes and Holdstacks
Different types of routes, from airways to SIDs, STARs and restricted segments, can be constructed
on RAMS through a sequence of navaids. Each pair of them defines a route segment to which altitude
and speed restrictions can be added.
There is the possibility to create holdstacks along the routes. They are particularly used in terminal
routes where flights are sent to them in case of runway scheduling conflicts. Holdstacks are defined by
a point (navaid) and restricted to a given number of flight levels [10].
18
Figure 11: SIDs (red), STARs (blue) and holdstacks (black)
4.6 Controllers
Each sector has a controller team, consisting of a planning and a tactical controller, responsible for the
safe movement of aircraft as they pass through their sector. Controllers have to assure that
separations between aircraft are fulfilled and, otherwise, intervene to resolve the conflict. Typically,
aircraft enter first in the planning controller for long term conflict prediction, while the tactical one takes
control of it closer to the sector pierce and takes care of short-term conflicts. Each controller has its
own separation standards.
4.6.1 Controller Windows
Controller windows are lists of flights and RAMS works with three of them: planning controller
information window, tactical controller information window and tactical controller handoff window. The
entries in planning information and tactical handoff windows are determined by a time offset
distribution before sector pierce, whereas the entry in tactical information is defined by a distance
offset.
Controller information windows contain a list of all flights controllers of that sector know about. The
same flight can be in more than one information window at a time (different sectors windows). The
tactical handoff window contains a list of the flights under control of the tactical controller. A flight
cannot be in more than one tactical handoff window simultaneously, since only one controller controls
it at a time [10]. When a new flight enters the window, the controller compares its flight plan with those
of the other flights already in the window. If there is a conflict, it will be identified at this time [32].
19
Usually, as an aircraft flies, its passage through controller windows occurs as described in the
following picture:
Figure 12: Controller Windows working scheme [32]
4.7 Conflict Detection and Resolution
With the defined separation parameters, RAMS creates a conflict zone around each flight. There are
different shapes available for conflict detection (Figure 13), with the aircraft in its centre. Combining
these shapes with the separation in altitude, a volume is created around the flight.
Figure 13: Conflict Detection Models [32]
By definition, a conflict exists when two aircraft are contained in each other’s shape. This means that
there is a possibility for the existence of some overlapping zones without a conflict situation warning.
20
The situation represented in Figure 14 does not constitute a conflict because, despite the existence of
an area where shapes overlap, none of the aircraft is violating the other’s separation distances.
Figure 14: Shape overlapping (no conflict) [32]
The case depicted in Figure 15 also does not represent a conflict situation. Although Flight 1 has
entered in Flight’s 2 shape, Flight 2 is not in Flight 1’s shape. In other words, only Flight 2 has its
separation values violated. When only one flight is contained in the other’s shape, RAMS classifies it
as an invalid conflict, producing no detection (nor resolution consequently).
Figure 15: Shape overlapping (conflict not detected) [10]
Finally, in Figure 16, there is a conflict detected situation, with both aircraft contained in each other’s
shape.
Figure 16: Shape overlapping (conflict detected) [10]
When a conflict is detected, the simulator tries different manoeuvers to keep the aircraft separated.
With this in mind, RAMS has conflict resolution rules established in a rule base either predefined or
user modified. It is possible to assign different rules to each controller in every single sector. Runways
have their own resolution rulebases.
21
If a conflict occurs, the rulebase is triggered and, taking into account the information of flight
characteristics, a manoeuvre is proposed. Its feasibility is weighted according to the following three
questions: 1.Can the aircraft perform it? 2. Does it solve the conflict? 3. Does it cause another
conflict? The manoeuvre is then applied or, if it does not solve or creates another conflict, the rulebase
tries to Figure out another solution [10][32].
Conflict resolution is an advanced simulation mode of RAMS and possesses some limitations,
particularly in complex sectors such as TMAs. As a consequence, the simulator may not be able to
find a solution for some conflicts. When none of the manoeuvres work, the original flight path of the
involved aircraft remains unchanged and the controller only monitors them. Additionally, there might
be some cases where the output data presents an “unkown” solution. This happens when the rulebase
has no information about how to change a conflict of that type.
A flowchart of the conflict resolution process can be found in Appendix A.
4.8 Tasks and Workload
Controller’s tasks can be simulated in RAMS. This is fundamental to estimate controller workload and,
consequently, sector capacity.
The different tasks are defined by a unique name and assigned to a category group and actor. The
simulator is pre-defined with five task categories: Conflict Search, Coordination, Flight Data
Management, Radar Activity and Radio/Telephone (RT) Communications. In fact, task names and
categories have no significance within the simulation, but are useful when analysing the output data,
making it possible to categorise the workload like in Figure 4. The actor is usually the controller that
performs the task, but there are other options. For example, if there is interest in studying the
telephone usage, there is the possibility of creating tasks with “telephone” as the actor. Each task is
then given a weight that, in a controller workload study, is the time (in seconds) to perform the task.
The weight can also be useful for other applications, like cost-benefit studies. In this case, it would
represent a cost.
All tasks are activated by a trigger, which is the event that leads to the task, and a time offset to
determine the moment when it occurs (negative if prior to the trigger, zero if realized at the trigger,
positive if after the trigger). RAMS recognizes more than 200 triggers.
There is the option to associate the trigger with dynamic conditions. In this case, the task will be
performed (and recorded) only if all the conditions are met at the triggering moment. The conditions
may be related with flight status (stable, unstable, climb, descent), airspace characteristics and, if the
trigger is connected with a conflict situation, the type of conflict and its resolution.
It is also possible to create multiple relationships between tasks and specific airspace objects. For
example, a task may only occur if a flight departs from a certain airport, and/or the aircraft is of a
specific model, and even if it has a specific destination and uses a certain airspace sector [10][32][33].
22
To better elucidate how tasks are created in RAMS, the following table presents the definition of a
simple one, where the ATCo vectors an aircraft to the base turn.
Comments
Task name T30
Trigger ATCNavAid Task occurs when flights fly through the NavAid
“PT402” Object PT402
Category Radar Activity
Activity Radar Vectoring for Base Turn
Actor Tactical Controller
Time offset 0 No time offset
Weight 8 Controller needs 8 seconds to perform the task
Table 4: Task Definition Example
4.9 Stochastic Variations
Because there is always some randomness in any system behaviour and because fast-time
simulations are based on the concept that, as the number of simulations increases (each one with
some degree of randomisation), the average of them becomes closer to reality, simulations require
some random inputs. In RAMS, there is an option of using distributions to provide this stochastic
variation. These distributions can be used in some functionalities such as: aircraft performance, task
workload weights, controller window entry and exit, runway occupancy times, traffic cloning and traffic
iterations generation [10].
4.10 ATM Analyser
RAMS generates extensive output files with detailed information. This data can be easily accessed
with ATM Analyser, a report generator included in RAMS Plus. It has a large number of pre-defined
reports and allows users to change them in order for them to obtain the desired information [10].
Nevertheless, some RAMS users prefer to analyse the results directly from the output files, often
creating “Macros” to access the information in a faster and more organized way.
23
5. Lisbon’s Airport and Airspace
Opened in 1940, and receiving commercial traffic since 1942 [44], the Lisbon Airport is the largest
Portuguese airport. Last year (2013), its two civil terminals received just over 16 million passengers
[45].
In 2006, it had a capacity of 36 movements per hour [46] but, since then, it was subjected to an
expansion plan and now can receive 38 movements per hour. It is claimed that this number could go
up to 41 movements per hour, if it was not constrained by its terminal airspace [43].
5.1 Runways
Lisbon Airport has two crossing runways (RWY): 03/21 and 17/35. RWYs 03 and 21 are equipped with
ILS (CAT I and CAT III respectively) [47].
Figure 17: Lisbon Runways [48]
Being impossible to operate in more than one runway simultaneously, and to enhance its capacity, the
airport receives the majority of the traffic in an arrival – departure – arrival sequence [49].
5.1.1 Runway Usage
The RWY in use is preferentially the 03/21, since operations in RWY 17/35 require coordination with
some restricted areas around the airport and, consequently, depend on military conditions. In general,
the active RWY is the 03, switching to RWY 21 whenever the weather justifies it. RWY 35 is
occasionally requested. From a traffic sample of July 2013, it is possible to verify this pattern in Figure
18.
24
Figure 18: Runway Usage
During the period in analysis, RWY 03 was in use in almost 82% of the time. Although some traffic
used RWY 35 in the period analysed, it was significant enough to be taken into account in this work.
At this point, the reader might be thinking about what makes the most used RWY have a less
sophisticated instrument approach system. The answer is simple: when operating under adverse
meteorological conditions, the wind usually comes from south/southwest [53]. For this reason, it is
important to notice that Figure 18 represents a study based on a summer month traffic sample and
may not reflect the behaviour throughout the year. In 2010, a data collection exercise took place
between November and December [49] recording 38% of the movements in RWY 21 (twice of the
observed in July 2013).
5.1.2 Runway Occupancy Times
The runway occupancy times used in this study were determined in 2010 [49].
The arrival runway occupancy time (ROTA) is defined as the interval between the aircraft crossing the
threshold4 and vacating the runway [50]. It is largely dependent of the exit point used. Average ROTA
values of both RWYs 03 and 21 are presented in Table 5.
RWY 03 RWY 21
Heavy 57.5s 52.9s
MedJet 51.1s 49.1s
MedProp 53.6s 42.9s
Table 5: ROTA by RWY
4 ICAO defines the threshold as the beginning of the portion of RWY usable for landing.
25
The departure runway occupancy time (ROTD), the time between crossing the stop bar for entering
the RWY and lifting the main gear from the ground [50], is more complex to calculate, since it includes
different key indicators:
FRLC – Flight crew reaction for line-up clearance;
LUPT – Line-up time;
TOCD – Take-off clearance delivery
FRTT – Flight crew reaction for take-off clearance;
TOFT – Take-off roll time.
The average times for the above mentioned intervals can be found in Table 6.
FRLC LUPT TOCD FRTT TOFT ROTD
RWY 03 16.8s 32.5s 4.1s 14.3s 31.4s 89.9s
RWY 21 12.9s 37.3s 8.4s 12.9s 31s 86.6s
Table 6: ROTD key indicators [49]
5.2 Airspace
The terminal control area comprises the airspace represented in Figure 19 (Lisboa TMA), excluding
the portion of Lisboa CTR within these limits. Its vertical lower limit varies from 1000ft above
ground/mean sea level to FL055, while its upper limit is FL245. Lisbon TMA has an unusual shape
due to military zones constraints.
Figure 19: Lisboa TMA
26
5.2.1 Sectors
Lisboa TMA, according to the Portuguese Aeronautical Information Publication (AIP) [47] can be
divided in up to four sectors (Figure 20):
Lisboa APP Sector 1 – An arc of circle centred on ARP with 30NM radius, from 2000ft to
FL085;
Lisboa APP Sector 2 – An arc of circle centred on ARP with 30NM radius, from 1000ft (on the
first 9NM radius) or 1500ft (between 9 and 30NM radius) to 2000ft;
Lisboa TMA Lower Sector – the area represented in Figure 19 from FL055 to FL145,
excluding the portion of Lisboa APP sector within its limits;
Lisboa TMA Upper Sector – the area represented in Figure 19, from FL145 to FL245.
Figure 20: Lisboa TMA sectorization
Usually, Lisboa TMA works with the configuration of two sectors (Lisboa TMA and Lisboa APP
sectors). In periods of low traffic demand, the sectors may be collapsed into a single one (Lisboa
TMA). Presently, the configuration of four sectors is very rarely used.
5.2.2 Military areas
“The Air Force accepts the concept that in peacetime the civilian air traffic, mainly the air
carriers, having in mind the economic vector, may deserve a prerogative. It means that
the Air Force requires high operational freedom but is conscious of Civil Aviation technical
requirements in airspace and the high effect in national economy. However, as well
known, all we have in peacetime to prepare at all levels our own devices to face the times
of crisis, tension and war.” (Capt. Luis G. Rocha) [51]
There are several military areas around Lisbon, Lisbon Airport being mainly affected by Alcochete,
Montijo, Sintra and Monte Real (Figure 21). They influence the design of the airspace, as well as the
Lisboa TMA
Lisboa TMA Sector
Lisboa TMA Lower Sector
Lisboa TMA Upper Sector
Lisboa APP Sector
Lisboa APP Sector 1
Lisboa APP Sector 2
27
approach and departure routes. The area of Montijo also conditions the use of RWY 35, due to its
proximity to the beginning of the RWY.
Figure 21: Most significant military areas around Lisbon
For a long time, there has been a bit of a dispute between civil and military authorities because of the
airspace around Lisbon. Although the above mentioned opinion of Capt. Luis Rocha may look recent,
it dates from 1982.
Given the configuration of the referred airspace, a good collaboration and an effective coordination
between both authorities is fundamental.
5.2.3 Routes
There are SIDs and STARs published for, respectively, traffic departing from and arriving at Lisbon.
RWYs 03 and 35 share the standard procedures. These two RWYs have eleven SIDs and twelve
STARs. For RWY 21, there are twelve departure procedures and nineteen STARs. Most of these
procedures are for RNAV equipped aircraft and can be shortened by radar vectoring or instructions to
follow specific waypoints whenever the conditions justify it. Some of this routes are pending on military
activity, since they use restricted airspace. There are no standardized approach or departure
procedures for RWY 17.
Along the routes there are some holdstacks defined. ADSAD is assigned for RWY 03, RINOR for
RWY 21, while EKMAR and UMUPI are for both RWYs 03 and 21. The last one is also dependent on
military activity, being located in the restricted area LPR42B.
For the inbound traffic, there are speed constraints. These are the maximum speeds allowed:
280KT IAS between FL245 and FL100;
250KT IAS at and below FL100;
28
220KT IAS at and below FL070
200KT IAS at and below 4000ft;
Between 160 and 180KT IAS when established on final approach segment;
160KT IAS until 4NM from RWY threshold.
More information about these procedures can be obtained in the Portuguese AIP [47].
Figure 22: RWYs 03 / 35 SIDs (red) and STARs
(blue)
Figure 23: RWY 21 SIDs (red) and STARs (blue)
5.2.4 Separations
The minimum horizontal separation in Lisboa TMA is 8NM, with the exception of the Lisboa Airport
Radar Vectoring Area where a minimum of 3NM is applied below FL155. This area is similar to the
Lisboa APP sector [47]. Although there is the possibility to reduce the separation until 3NM, the APP
sector usually applies a minimum of 5NM. The vertical separation minimum is 1000ft, in accordance
with Doc. 4444 [52]. Additionally, wake vortex separations are imposed according to the aircraft’s
category, as presented in Table 7.
Heavy (leading) Medium (leading) Light (leading)
Heavy (trailing) 3NM 3NM 3NM
Medium (trailing) 4NM 3NM 3NM
Light (trailing) 6NM 4NM 3NM
Table 7: Wake vortex separations
Since none of the existing SIDs allows departing aircraft to fly in a track diverging 45 degrees or more
from the previous one, there is the need to impose a two minute separation between departures as
per ICAO PANS-ATM Doc. 4444 chapter 5.6 [52].
29
5.3 Traffic
In 2013, Lisbon Airport received approximately 145.000 movements [54][55][56][57]. The summer
represents an obvious peak of the traffic, while the first trimester is the period with fewer movements.
Figure 24: Traffic in Lisbon by trimester (year 2013)
The traffic sample used in this study is from July 2013. During this period, the airport operated always
above the four hundred movements per day (Figure 25).
Figure 25: Movements of July 2013
From that sample, the traffic from July 1st and 12th was analysed in more detail. These days were
selected because only one RWY was used during the 24 hours (RWY 03 in July 1st and RWY 21 on
July 12th) and because it presented more movements and highest peak hours. Given that this sample
will be used to calibrate the scenario, there was interest an interest in using a pattern that not only
represents a typical flow of traffic but also produces peaks as close as possible to the declared
capacity limit. In this case, the controller workload should be around 70%.
0
10000
20000
30000
40000
50000
1st Trimester 2nd Trimester 3rd Trimester 4th Trimester
Movem
ents
Traffic Year 2013
Movements Average
200
250
300
350
400
450
500
Movements July 2013
TOTAL
ARRIVALS
DEPARTURES
30
In terms of geographic distribution, more than half of the daily traffic arrives and departs from and to
the north zone of the TMA, by the points XAMAX, INBOM and IXIDA. Together with the movements
from the east side, with IDBID and EXONA, they represent around 75% of the traffic (Figure 26 and
Figure 27). In terms of aircraft, more than 80% of the flights are done with medium sized jets.
Figure 26: Entry and exit points - 1st of July
(RWY 03)
Figure 27: Entry and exit points - 12th of July
(RWY 21)
31
6. Simulations
In this section, the different scenarios simulated will be described and the results will be presented.
Reference scenarios will have a more detailed description, where all important simulation inputs will
be explained. Since the remaining scenarios are variants from their references, their description will
consist of the differences between them and the reference scenario.
The following table presents the list of scenarios that will be studied and their main characteristics.
Scenario Name
RWY Route structure
Traffic Sectorization Purpose
Reference Scenario 03
03 AIP 01 July 2013 + 3%
APP + TMA (AIP) Reproduce the current situation and model calibration
LPPT_03_B 03 AIP 01 July 2013 + 3%
APP 1 + APP 2 + TMA Lower +
TMA Upper (AIP)
Analyse eventual benefits of this sectorization
LPPT_03_C 03 AIP including routes and holdstack
pending on military activity
01 July 2013 + 5%
APP + TMA (AIP) Calculate the impact of the routes and holdstack pending
on military activity.
LPPT_03_D 03 AIP 01 July 2013 + 8%
APP + TMA (AIP) Evaluate the impact of a reduction in RWY occupancy
times
LPPT_03_E 03 New SIDs APP + TMA (AIP) Decrease the separation between departures from 2 to
1 minute
LPPT_03_F 03 AIP 01 July 2013 + 25%
APP North + APP South + TMA
New sectorization with approach and departure
sectors
Reference Scenario 21
21 AIP 12 July 2013 +10%
APP + TMA (AIP) Reproduce the current situation and model calibration
LPPT_21_B 21 AIP 12 July 2013 +10%
APP 1 + APP 2 + TMA Lower +
TMA Upper (aip)
Analyse eventual benefits of this sectorization
LPPT_21_C 21 AIP including routes and holdstack
pending on military activity
12 July 2013 +10%
APP + TMA (AIP) Calculate the impact of the routes and holdstack pending
on military activity.
LPPT_21_D 21 AIP 12 July 2013 +10%
APP + TMA (AIP) Evaluate the impact of a reduction in RWY occupancy
times
LPPT_21_E 21 New SIDs APP + TMA (AIP) Decrease the separation between departures from 2 to
1 minute
LPPT_21_F 21 AIP APP North + APP South + TMA
New sectorization with approach and departure
sectors
Table 8: List of scenarios and characteristics
32
6.1 Reference Scenario RWY 03
6.1.1 Overview
In this scenario, an attempt is made to reproduce the current Lisbon Airport Terminal Airspace, as well
as its operating mode. The airspace configuration, routes and procedures are replicated as close as
possible to the ones published in the AIP and described in previous chapter.
6.1.2 Airspace
In this simulation, the airspace configuration is the most commonly used in Lisbon, with two sectors
(APP and TMA), as described in 5.2.1. It is represented in Figure 28.
Figure 28: Sectorization of reference scenario RWY 03
6.1.3 Routes
The standard routes are the ones published in the AIP, with exception to some changes in STARs
structure that will be explained later in 6.1.6. Also, the two holdstacks assigned to RWY 03 were
designed. The holdstack UMUPI was not included since it is located in a military area, its utilization
being conditioned.
Figure 29: SIDs and STARs of reference scenario RWY03
33
6.1.4 Runways, Taxiways and Gates
Since this study was confined to the airspace, the ground facilities of the airport were only grossly
designed to assure that the arriving traffic would not stay on the runway. They consist of a “super
gate” with capacity for all planes that enter in the simulation, and taxiways to and from the runway.
Initially, the runways were the only ground element planned to be designed but, while running the
simulation, it was noticed that some arriving flights were staying for a while on the runway for the
turnaround, causing conflicts and delays.
Figure 30: Ground representation
Due to the problem with runway’s taxi speed described in section 4.4, it was decided to configure
runway occupancy times according to performance groups. Thus, the landing occupancy time is
assigned according to the aircraft’s group with the times presented in 5.1.2. For departures, it was
considered an average of 43 seconds for all groups.
6.1.5 Traffic
The traffic sample used for this simulation is the real one from the 1st of July of 2013, which consist in
a total of 466 movements (241 arrivals and 225 departures). Because this traffic was not enough to
saturate any of the sectors, it was increased using the traffic generator tool of RAMS in about 3%,
obtaining a total of 481 flights.
6.1.6 Conflict Resolution Rules
One of the biggest challenges faced during the realization of this work was the formulation of a rule
system capable of solving conflicts inside the terminal airspace.
As mentioned before in section 4.7, conflict resolution is an advanced simulation mode of RAMS and
presents some limitations. Although the pre-defined rules may work well enough for the majority of en-
route situations, in complex scenarios such TMAs they can struggle to solve some conflicts. This
happens mostly because, once an aircraft starts flying in a SID or STAR, the simulator does not
change any of the flight parameters, except in cases of runway conflict. While this shall not be a big
34
issue inside the same SID/STAR, admitting that at the entrance point the flights are with an adequate
separations, it originates problems in crossing or merging points of STARs. Because this reference
scenario was pretended to be as close as possible to the reality, it was necessary to define rules that
would result in a realistic traffic behaviour.
The first to be addressed will be the runway conflict resolution rules. When the conflict is between an
arrival and a departure, the tendency is to penalise the aircraft on ground. However, it is important to
find a sensible middle to avoid large departure queues with consequent long delays. For this reason,
and to guarantee enough separation between aircraft on final approach, an interval of two minutes
between arrivals is imposed, allowing the airport to operate in an arrival – departure – arrival system in
peak traffic hours.
To make the two minutes interval between arrivals respected, it is necessary to coordinate the aircraft
approaching Lisbon Airport. The first manoeuvre defined in the runway conflict resolution rules is
“speed reduction to runway”. Although this is the manoeuvre that causes less negative impact on
flights, its effectiveness is limited. Combining aircraft’s minimum approach speed with the limits
imposed in STARs makes it only possible to reduce slightly the speed of arriving aircraft.
Consequently, it is often necessary to make aircraft fly a longer distance.
Since most of the traffic approaches from north, the method used by Lisbon’s controllers is simply to
extend the downwind, vectoring the traffic for the base turn when at a distance capable of assuring the
separation minima. This procedure is predicted in AIP [47], with waypoints that make it possible to
reduce or extend the route (Figure 31: Detail of the STAR chart [47]). This procedure is easily verified
observing aircraft’s trajectories (Figure 32: Short downind RWY 03 [58] and Figure 33).
Figure 31: Detail of the STAR chart [47]
35
Figure 32: Short downind RWY 03 [58]
Figure 33: Extended downwind RWY 03 [58]
To reproduce that, ten different STARs (for each original STAR) were designed in RAMS, each one
extending the previous one in 2NM. This applies to the original standard arrivals from north (INBOM
and XAMAX) and makes it possible to represent the variable turning point to the base leg. The second
rule of runway conflict resolution is, of course, “Alternate STAR”.
Figure 34: Simulation detail - base turns
For the approach routes from west (BUSEN), as well as the route from east (EXONA), the aircraft are
sequenced by routing them for a shorter or longer final approach, depending on the traffic conditions
at the moment, as it is possible to observe in the following pictures.
Figure 35: Short approach from west [58]
Figure 36: Long approach from west [58]
36
Figure 37: Short approach from east [58]
Figure 38: Long approach from east [58]
To make the simulation able to reproduce this, the waypoints EKMAR (for west approaches) and
ADSAD (for east approaches) work as sequencing points and the route flown by the aircraft from then
onwards depend on other traffic that may also be approaching Lisbon at that time.
Figure 39: Simulation detail - ADSAD and EKMAR sequencing
Last, the approach routes from south. In this case, if the traffic allows it, the aircraft is routed directly to
the final approach segment (Figure 40). If not, the aircraft flies into one of the above mentioned
sequencing points (EKMAR for approaches starting in LIGRA and UNPOT, ADSAD for NAKOS,
TROIA and GAIOS as established in AIP) (Figure 41). The reproduction of these routes in the
simulation is presented in Figure 42.
Figure 40: South approach direct to final [58]
Figure 41: South approach via EKMAR [58]
37
Figure 42: Simulation detail - Approach routes from south
When neither the speed reduction nor the alternate STAR procedure solves the conflict, then the third
rule, which consists in putting the flight waiting in a holdstack, is triggered. There is also a “hold on
ground” rule. This rule holds departures at the departure queue when the runway is occupied, due to
an aircraft landing or another taking off.
With these rules solving runways’ conflicts, it is now necessary to address other conflict situations
inside the TMA due to separation minima violation. It would be expected that at the entrance of the
TMA the traffic would have enough separation. Nevertheless, the traffic introduced for the simulation
does not always respect this. That happens because, in the original traffic samples, some flights were
vectored directly to a point inside the TMA, not entering in it by the defined points. This traffic sample
was modified to make all flights approach the airport via standard routes and, as a result, some flights
ended timed to enter in the TMA without enough separation between them. To solve this problem,
each point of entrance in the TMA also works as a metering point, in an attempt to give enough
distance between aircraft entering the same STAR.
Another spot of conflict were STARs merging points. As mentioned at the beginning of this section,
once the aircraft is at a STAR, the simulation only changes flights’ parameters if there is a conflict on
the runway. Because of that, manoeuvres such as speed change or altitude change are not a
hypothesis to solve separation conflicts. The best way found to deal with this problem was to use the
alternate STAR rule once again. For this, some additional STARs were defined in the simulation,
giving some flexibility at merge points. The final system of STARs used for the reference scenario of
RWY 03 is presented in Figure 43.
38
Figure 43: STARs of reference scenario RWY 03
When none of the rules from the rulebase is capable of solving the conflict, it is monitored. This task
(conflict monitor) was assigned a high weight to allocate the time required by the controller to solve the
conflict. For example, this time must be sufficient to vector the aircraft in a way that solves the conflict.
6.1.7 Workload
Because each TMA has its own particularities, each workload study needs a specific list, with the rules
played by that sector’s controllers. Given that, at the moment, there is no data for Lisbon ATC, it was
necessary to create a list of tasks assigned to the controllers and their characteristics (such as offset,
weight and trigger). The most appropriate mode to do that would be to observe the controllers working
at their positions, analysing and timing the tasks performed, and then discuss and adjust them with a
team of experienced ATCos. This is a common method used in EUROCONTROL CAPAN studies.
However, the control room being a sterile environment with restricted access, the execution of a study
of this type was beyond the scope of this thesis.
The solution was to build a task list based on others already existent. The major source used consists
in a study developed for a hypothetical scenario of a new airport in Ota. Additionally, it was
complemented with data from the study of Prague TMA [33]. After discussed and analysed together
with experts from NAV Portugal, the task list presented in Appendix B was obtained.
A remark for the points inserted in the base turns, near to the turning point from the downwind. There
was the need to define a task that consisted in radar vectoring the aircraft for the base turn and the
obvious trigger to that would be the point along the downwind where the turn starts. But, because all
aircraft fly through those points while waiting for the order to turn, the task would keep being triggered
by all aircraft flying through the downwind. In order to avoid that, the new points were inserted in the
base turns and the task, instead of being triggered in the downwind, is triggered at the base turn with
an offset. The same logic is applied to the points on route segments between ADSAD and the final
approach.
39
6.1.8 Simulation Procedures
The simulation was initially run with the original traffic of the 1st of July 2013. Because it did not
saturate any airspace sector, more traffic was generated in accordance with the procedure of CAPAN
Method. With a sample of 481 flights, the 70% threshold was attained in the APP sector.
6.1.9 Results
In Figure 44, with data according to the one described in section 3.2, it is possible to verify the tactical
controller workload attaining the 70% threshold in a period of time around midday. The relation
between workload and flights entering the sector is clearly observable, with workload being almost a
linear function during periods of low demanding and increasing considerably in periods of intense
traffic. The relation of the workload with the number of conflicts is also unquestionable. The number of
conflicts presented is only relative to the tactical controller.
Figure 44: Graphic of key indicators – Lisboa APP - Reference Scenario RWY 03
Transposing these results for a graphic of workload as a function of the number of sector entries, and
performing a regression analysis, it is possible to confirm the proximity of the function with a quadratic
function.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa APP- Reference Scenario - Runway 03Traffic 01 July 2013 + 3% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
40
Figure 45: Regression analysis - Lisboa APP - Reference Scenario RWY 03
With the formula obtained from the regression, a capacity estimation of 38 movements per hour is
obtained for this sector.
The other sector, Lisboa TMA, presented considerable lower values of workload and was far from
saturation. The capacity of this sector was not calculated, since these results would not give a reliable
estimation, but it is obviously higher than APP capacity and consequently does not introduce any
constraint.
Figure 46: Graphic of key indicators - Lisboa TMA - Reference Scenario RWY 03
Relatively to the runway, it is possible to observe in Figure 47 its usage, with a peak of 36 movements
around midday, coincident with the maximum attained in the number of sector entries in APP sector.
y = 0,0155x2 + 1,2256xR² = 0,977
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35 40
Wo
rklo
ad (
%)
No. Flights
Lisboa APP- Reference Scenario - Runway 03Traffic 01 July 2013 + 3% - 5 Iterations
TC (%) Polynomial (TC (%))
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa TMA - Reference Scenario - Runway 03Traffic 01 July 2013 + 3% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
41
Figure 47: Graphic of RWY movements - Reference Scenario RWY 03
For the number of separation conflicts and their resolutions, the average results are presented in the
following table.
Meter_EXONA Meter_INBOM Meter_LIGRA LisboaAPP LisboaTMA
AlternateSTAR - - - 22 79
Meter 3 15 1 - -
Monitor - - - 85 20
Unknown - - - - 2
TOTAL 3 15 1 107 101
Table 9: Reference Scenario RWY 03 - Conflict resolution summary
The conflict resolutions applied to runway conflicts are resumed in Table 10: Reference Scenario
RWY 03 - Runway conflict resolution summary.
TowerTController_LPPT
AlternateSTAR 2
HoldOnGroundRunway 151
HoldStackToFrom 18
Monitor -
SpeedReductionToRunway 67
TOTAL 238
Table 10: Reference Scenario RWY 03 - Runway conflict resolution summary
All these interventions in the original flight plans have obvious impacts in the time and distance
travelled by the aircraft. They are resumed in Table 11.
0
5
10
15
20
25
30
35
40
00
:00
00
:50
01
:40
02
:30
03
:20
04
:10
05
:00
05
:50
06
:40
07
:30
08
:20
09
:10
10
:00
10
:50
11
:40
12
:30
13
:20
14
:10
15
:00
15
:50
16
:40
17
:30
18
:20
19
:10
20
:00
20
:50
21
:40
22
:30
23
:20
Runway 03 - Reference Scenario - Arrivals and DeparturesTraffic 01 July 2013 - 20 Iterations
Departures Arrivals Total
42
Sector Distance Impact (NM)
Time Impact (minutes)
LisboaAPP 14.6 4.3
LisboaTMA 0.8 1.2
Departures delay n.a. 1.8
Table 11: Reference Scenario RWY 03 - Conflict resolution impacts
The last information presented about this scenario is the utilization of the holdstacks in terms of
number of flights sent to them and the average time they stayed there (Table 12).
Holdstack No. Flights Received Average holding time (seconds)
ADSAD 12 68
EKMAR 6 225
Table 12: Reference Scenario RWY 03 - Holdstack data
6.2 LPPT_03_B
6.2.1 Overview
This scenario introduces the four sectors airspace configuration described in 5.2.1 and pretends to
evaluate if it improves the capacity of the two sector Reference Scenario.
The inputs and procedures are the same of the Reference Scenario, with the exception of the sectors,
which in this simulation acquire the vertical profile of Figure 48.
Figure 48: Four sectors scenario - Sector's vertical profile
In this scenario, it was imposed a minimum separation of one minute between traffic entering the
simulation. Thus, it is expected to obtain a slightly lower workload for this scenario as a result of
possible avoided conflicts with the minimum separation imposed.
43
6.2.2 Results
The results obtained for this scenario were, as expected, close to the reference scenario, with a
slightly low workload. The capacity of Lisboa APP Sector 1, which in fact receives the same traffic as
Lisboa APP in the previous scenario, is the limiting sector and its capacity remains unchanged in the
38 movements per hour.
Figure 49: Graphic of key indicators – Lisboa APP 1 - Four Sector Scenario RWY 03
Figure 50: Regression analysis - Lisboa APP 1 – Four Sectors Scenario RWY 03
No significant information is extracted from the data of the remaining sectors, since they are far from
saturation. The graphic of the runway is identical to the one from the Reference Scenario.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
APP 1 - LPPT_03_B - Runway 03Traffic 01 July 2013 + 3% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
y = 0,0165x2 + 1,1776xR² = 0,9802
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35 40
Wo
rklo
ad (
%)
No. Flights
APP_1 - LPPT_03_B - Runway 03Traffic 01 July 2013 + 3% - 5 Iterations
TC (%) Polynomial (TC (%))
44
Figure 51: Graphic of key indicators - Lisboa
APP 2 - Four Sector Scenario RWY 03
Figure 52: Graphic of key indicators - Lisboa
TMA lower - Four Sector Scenario RWY 03
Figure 53: Graphic of key indicators - Lisboa
TMA upper - Four Sector Scenario RWY 03
Figure 54: Graphic of RWY movements – Four
Sector Scenario RWY 03
In terms of conflict detection and resolution, the situation is identical to the reference scenario, since
the modifications made for this scenario do not affect this subject.
6.3 LPPT_03_C
6.3.1 Overview
Comparatively to the reference scenario, this simulation introduces the holdstack UMUPI assigned to
RWY 03. It pretends to evaluate the impact of this holdstack in the capacity of airspace and whether
military exercises in this area may constrain the operation of the airport.
The traffic sample was subject to a 5% increment relatively to the original.
6.3.2 Results
Once again, and as expected, the first sector to attain its maximum capacity was Lisboa APP sector.
45
Figure 55: Graphic of key indicators – Lisboa APP – Scenario with UMUPI holdstack RWY 03
Figure 56: Regression analysis - Lisboa APP – Scenario with UMUPI holdstack RWY 03
Proceeding in the same way, a capacity of 38 movements per hour is estimated. The workload in
Lisboa TMA sector is, once again, well below the threshold.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa APP - LPPT_03_C - Runway 03Traffic 01 July 2013 + 5% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
y = 0,0166x2 + 1,1935xR² = 0,9833
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35 40
Wo
rklo
ad (
%)
No. Flights
Lisboa APP- LPPT_03_C - Runway 03Traffic 01 July 2013 + 5%- 5 Iterations
TC (%) Polynomial (TC (%))
46
Figure 57: Graphic of key indicators – Lisboa APP – Scenario with UMUPI holdstack RWY 03
In terms of runway movements, there are no significant differences, with a peak of 37 movements in
an hour at midday.
Figure 58: Graphic of RWY movements – Scenario with UMUPI holdstack RWY 03
With the difference between this scenario and the reference one consisting in an additional holdstack,
there is interest in analyzing how it influences the holdstack occupancy. Because of that, the data
concerning the holdstacks is presented in the following table.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa TMA - LPPT_03_C - Runway 03Traffic 01 July 2013 + 5% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
0
5
10
15
20
25
30
35
40
00
:00
00
:50
01
:40
02
:30
03
:20
04
:10
05
:00
05
:50
06
:40
07
:30
08
:20
09
:10
10
:00
10
:50
11
:40
12
:30
13
:20
14
:10
15
:00
15
:50
16
:40
17
:30
18
:20
19
:10
20
:00
20
:50
21
:40
22
:30
23
:20
Runway 03 - LPPT_03_C - Arrivals and DeparturesTraffic 01 July 2013 + 5% - 5 Iterations
Departures Arrivals Total
47
Holdstack No. Flights Received Average holding time (seconds)
ADSAD 3 176
EKMAR 4 203
UMUPI 12 23
Table 13: Holdstack usage - Scenario with UMUPI holdstack RWY 03
It is possible to notice that, despite not increasing the capacity, the holdstack “UMUPI” it’s the most
used and has a low average holding time.
The route XAMAX4C, which also would be object of study in this scenario, only receives one flight
along the day, and therefore does not justify a detailed analysis.
6.4 LPPT_03_D
6.4.1 Overview
The aim of this scenario is to evaluate the effects of a reduction in the runway occupancy time. The
new times are presented in Table 14.
Old RWY Occupancy Times New RWY Occupancy Times
Arrivals Heavy
58s 54s
Arrivals MedJet
51s 48s
Arrivals MedProp
54s 51s
Departures 43s 40s
Table 14: LPPT_03_D - New RWY Occupancy Times
The traffic was increased by 8% to achieve saturation in one of the sectors.
6.4.2 Results
The reduction of the runway occupancy times did not have effects on the capacity of the airport, since
the capacity of the sector Lisboa APP is still of 38 movements per hour.
48
Figure 59: Graphic of key indicators – Lisboa APP –Reduced RWY occupancy time RWY 03
Figure 60: Regression analysis - Lisboa APP – Reduced RWY occupancy time RWY 03
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa APP- LPPT_03_D - Runway 03Traffic 01 July 2013 + 8% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
y = 0,0172x2 + 1,1651xR² = 0,9737
0
10
20
30
40
50
60
70
80
90
0 5 10 15 20 25 30 35 40 45
Wo
rklo
ad (
%)
No. Flights
Lisboa APP- LPPT_03_D - Runway 03Traffic 01 July 2013 + 8%- 5 Iterations
TC (%) Polynomial (TC (%))
49
Figure 61: Graphic of key indicators – Lisboa TMA –Reduced RWY occupancy time RWY 03
As expected, the peak number of runway movements did not change considerably, because of the
airspace constraint.
Figure 62: Graphic of RWY movements – Reduced RWY occupancy times RWY 03
The only considerable improvement obtained from the reduction of the RWY occupancy time is a
decrease in the average time on hold of flights where this resolution was applied, as it is possible to
observe in Table 15.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa TMA - LPPT_03_D - Runway 03Traffic 01 July 2013 + 8% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
0
5
10
15
20
25
30
35
40
45
00
:00
00
:50
01
:40
02
:30
03
:20
04
:10
05
:00
05
:50
06
:40
07
:30
08
:20
09
:10
10
:00
10
:50
11
:40
12
:30
13
:20
14
:10
15
:00
15
:50
16
:40
17
:30
18
:20
19
:10
20
:00
20
:50
21
:40
22
:30
23
:20
LPPT_03_D - Arrivals and DeparturesTraffic 01 July 2013 + 8% - 5 Iterations
Departures Arrivals Total
50
Holdstack No. Flights Received Average holding time (seconds)
ADSAD 12 59
EKMAR 6 201
Table 15: LPPT_03_C – Holdstack data
The average delay of late departures remains in the 1.8 minutes.
6.5 LPPT_03_E
6.5.1 Overview
In this scenario, the standard departure procedures are modified.
As mentioned in 5.2.4, all traffic taking off in Lisbon Airport must have a minimum time separation of
two minutes since there are no departure procedures with tracks diverging 45º or more. The aim in this
scenario was to introduce new routes to operate departures with one minute separation as per ICAO
PANS-ATM Doc. 4444 chapter 5.6 [52]. The idea was to create an alternate SID for each of the
existent ones, with the minimum divergence of 45º between the tracks (Figure 63 and Figure 64).
There were concerns related with the restricted area of Alcochete, which may interfere with these
procedures.
Figure 63: Minimum separation between departing aircraft (45 degrees divergence) [52]
Figure 64: Minimum separation between departing aircraft (same track) [52]
51
With some reorganization of the current SIDs (Figure 65), it was possible to introduce the alternatives
without conflicting with the military areas, as presented in detail in Figure 66 where the restricted areas
are marked in brown. Although the departure routes cross the area of Montijo, the aircraft have an
altitude higher than the 2000ft of vertical limit imposed while flying over that area.
Figure 65: LPPT_03_E - New SIDs
Figure 66: Detail of military areas
6.5.2 Results
The capacity of the APP sector in this scenario decreases to 37 movements per hour, due to the
complexity added with the new design applied in the departure routes, which includes two merging
points.
Figure 67: Graphic of key indicators – Lisboa APP – New SIDs RWY 03
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa APP- LPPT_03_E- Runway 03Traffic 01 July 2013 + 5% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
52
Figure 68: Regression analysis - Lisboa APP – New SIDs RWY 03
The strength of this configuration lies on the low delays occurred with departures, with an average of
1.1 minutes only considering the late takeoffs.
6.6 LPPT_03_F
6.6.1 Overview
This version presents a new airspace configuration. Since the sector Lisboa APP limits the airspace
capacity, and the division analysed in 6.2 does not improve it, another division is tried, this time in
North and South instead of Lower and Upper.
The new sectorization predicts an APP Approach Sector (which, for RWY 03, will be APP South) and
an APP Departure Sector (APP North), as presented in Figure 69.
y = 0,0194x2 + 1,1578xR² = 0,9827
0
10
20
30
40
50
60
70
80
90
0 5 10 15 20 25 30 35 40
Wo
rklo
ad (
%)
No. Flights
Lisboa APP- LPPT_03_E - Runway 03Traffic 01 July 2013 + 5%- 5 Iterations
TC (%) Polynomial (TC (%))
53
Figure 69: Lisboa APP North and South
Some aircraft departing for the south side of the TMA, namely those with lower climbing
performances, may enter the APP Sector South, here working as the approach sector. Since this
occurs in a very brief period of time (usually less than half a minute) in the transition between Lisboa
APP North and Lisboa TMA, a “sector clip” was introduced here. As a result, if a departure flight enters
the South Sector and will exit in less than a minute, it is handed off straight to the control of Lisboa
TMA, instead of being handed to the control of this sector.
Although the North Sector is supposed to work as the departure sector, it is still crossed by part of the
traffic approaching Lisbon via STARs INBOM and XAMAX, depending on their altitude profile. Since
the average crossing time of these flights is considerably higher than in the case of the departures, the
sector clip is not applied here.
6.6.2 Results
With this new configuration, the approach sectors do not constraint the airspace capacity, presenting a
workload far from the threshold. The division introduced in the original Lisboa APP sector was not only
able to increase the capacity, but also to split the workload in a very similar proportion between the
two new sectors.
54
Figure 70: Graphic of key indicators – Lisboa APP North – New APP sectors RWY 03
Figure 71: Graphic of key indicators – Lisboa APP South – New APP sectors RWY 03
The sector that reaches saturation in this scenario is Lisboa TMA, which remain unchanged relatively
to the original sector. Its estimated capacity is 48 movements per hour.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa APP North- LPPT_03_F - Runway 03Traffic 01 July 2013 + 25% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa APP South - LPPT_03_F - Runway 03Traffic 01 July 2013 + 25% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
55
Figure 72: Graphic of key indicators – Lisboa TMA – New APP sectors RWY 03
Figure 73: Regression analysis – Lisboa TMA – New APP sectors RWY 03
The runway has a peak of 44 movements per hour, which is almost 10% lower than the capacity of the
airspace. This will be object of analysis in chapter 7.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa TMA - LPPT_03_F - Runway 03Traffic 01 July 2013 + 25% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
y = 0,0046x2 + 1,2159xR² = 0,9672
0
10
20
30
40
50
60
70
80
0 10 20 30 40 50
Wo
rklo
ad (
%)
No. Flights
Lisboa TMA - LPPT_03_F - Runway 03Traffic 01 July 2013 + 25% - 5 Iterations
TC (%) Polynomial (TC (%))
56
Figure 74: Graphic of RWY movements – New APP sectors RWY 03
The following table presents the average delays and distance impact due to the conflict resolution
manoeuvres. Comparing them with the values from the reference scenario, it is possible to observe an
increase on the different parameters.
Sector Distance Impact (NM) Time Impact (minutes)
Lisboa APP North 25.9 7.2
Lisboa APP South 0 0.9
Lisboa TMA 1.4 1.8
Departures delay n.a. 2.8
Table 16: Conflict Resolution Impacts - New APP sectors scenario
The number of flights sent to holdstacks, as well as the average time they stayed there, also presents
higher values.
Holdstack No. Flights Received Average holding time
(seconds)
ADSAD 33 118
EKMAR 13 233
Table 17: Holdstack data - New APP sectors scenario
6.7 Reference Scenario RWY 21
6.7.1 Overview
This scenario is in all equivalent to the presented in 6.1, but this time the RWY in use will be the 21
and, of course, SIDs and STARs will be the ones published in AIP for this RWY.
0
5
10
15
20
25
30
35
40
45
50
00
:00
00
:50
01
:40
02
:30
03
:20
04
:10
05
:00
05
:50
06
:40
07
:30
08
:20
09
:10
10
:00
10
:50
11
:40
12
:30
13
:20
14
:10
15
:00
15
:50
16
:40
17
:30
18
:20
19
:10
20
:00
20
:50
21
:40
22
:30
23
:20
LPPT_03_F - Arrivals and DeparturesTraffic 01 July 2013 + 25% - 5 Iterations
Departures Arrivals Total
57
Figure 75: SIDs and STARs RWY 21
For conflict resolution, alternate STARs were designed using the same principles applied for the case
of RWY 03. The final result is presented in Figure 76.
Figure 76: Alternate STARs RWY 21
The traffic sample is from the 12th of July 2013, with an increase of 10%, totalizing 528 flights, to
saturate one of the sectors.
6.7.2 Results
For the case of RWY 21, the APP sector is also the one that provides capacity constraints.
58
Figure 77: Graphic of key indicators – Lisboa APP– Reference scenario RWY 21
Figure 78: Regression analysis – Lisboa APP – Reference Scenario RWY 21
The capacity is estimated at 38 movements per hour, as it was for RWY 03. Although it was necessary
to increase the traffic up to a number of movements considerably higher than for the case of RWY 03,
this doesn’t mean that this scenario has greater capacity. In fact, the reason for that is related with the
traffic sample that, for the case of RWY 21, had more traffic distributed along the day.
The major difference consists in the sector of Lisboa TMA. Although it never attains the 70%
threshold, the workload values are significantly more than in RWY 03.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa APP- Reference Scenario - Runway 21Traffic 12 July 2013 + 10%- 5 iterations
TC (%) Flights Peak Occ. Conflicts
y = 0,0067x2 + 1,5495xR² = 0,8939
0
10
20
30
40
50
60
70
80
90
0 5 10 15 20 25 30 35 40
Wo
rklo
ad (
%)
No. Flights
Lisboa APP- Reference Scenario - Runway 21Traffic 12 July 2013 + 10%- 5 Iterations
TC (%) Polynomial (TC (%))
59
Figure 79: Graphic of key indicators – Lisboa TMA– Reference scenario RWY 21
In terms of RWY movements, the results show a peak of 39 movements, this value being obtained in
the periods of overloading in APP sector. Once again, there is evidence of a superior capacity of the
RWY when compared with the airspace.
Figure 80: Graphic of RWY movements – Reference Scenario RWY 21
The number of separation conflicts detected in this scenario is roughly the same of the reference
scenario of RWY 03, when pondered with the amount of traffic. Nevertheless, there is an important
difference to be noticed, with regards to the resolution. There is a clear increment of “monitor”
resolutions, indicating that it is harder for the simulator to find conflict resolutions. The number of
conflicts solved by alternate STARs decreased around 50%.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa TMA - Reference Scenario - Runway 21Traffic 12 July 2013 - 5 iterations
TC (%) Flights Peak Occ. Conflicts
0
5
10
15
20
25
30
35
40
45
00
:00
00
:50
01
:40
02
:30
03
:20
04
:10
05
:00
05
:50
06
:40
07
:30
08
:20
09
:10
10
:00
10
:50
11
:40
12
:30
13
:20
14
:10
15
:00
15
:50
16
:40
17
:30
18
:20
19
:10
20
:00
20
:50
21
:40
22
:30
23
:20
Runway 21 - Reference Scenario - Arrivals and DeparturesTraffic 12 July 2013 + 10% - 5 Iterations
Departures Arrivals Total
60
Meter EXONA
Meter INBOM
Meter LIGRA
Meter XAMAX
LisboaAPP LisboaTMA
AlternateSTAR - - - - 10 38
Meter 1 8 2 2 - -
Monitor - - - - 163 49
Unknown - - - - - -
TOTAL 1 8 2 2 173 87
Table 18: Reference Scenario RWY 21 - Conflict resolution summary
In this scenario, RWY’s rulebase has to resort more in holding manoeuvres, maintaining more
departing traffic on the ground and sending a higher number of arriving aircraft to holdstacks.
TowerTController_LPPT
AlternateSTAR 7
HoldOnGroundRunway 188
HoldStackToFrom 28
Monitor -
SpeedReductionToRunway 95
TOTAL 318
Table 19: Reference Scenario RWY 21 - Runway conflict resolution summary
The impact of the conflict resolution procedures assumes greater proportions, with more delays for
departing aircraft and more time on holdstacks for arriving traffic.
Sector Distance Impact (NM)
Time Impact (minutes)
LisboaAPP 18.4 7.0
LisboaTMA 4.3 2.1
Departures delay n.a. 2.4
Table 20: Reference Scenario RWY 21 - Conflict resolution impacts
Holdstack No. Flights Received Average holding time (seconds)
EKMAR 3 328
RINOR 25 152
Table 21: Reference Scenario RWY 21 - Holdstack data
6.8 LPPT_21_B
6.8.1 Overview
This scenario is a variant of LPPT_03_B presented in section 6.2 applied to RWY 21.
61
6.8.2 Results
The capacity of Lisbon APP Sector 1 increased to 39 movements per hour, according to the
information provided in the following graphics. This apparent capacity’s improvement will be discussed
in chapter 7.
Figure 81: Graphic of key indicators – Lisboa APP 1 – 4 sector scenario RWY 21
Figure 82: Regression analysis – Lisboa APP 1 – 4 sector scenario RWY 21
As in scenario LPPT_03_B, no other significant information is obtained neither from the other sectors
nor from the runway.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa APP_1 - LPPT_21_B - Runway 21Traffic 12 July 2013 + 10% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
y = 0,0172x2 + 1,1024xR² = 0,977
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35 40 45
Wo
rklo
ad (
%)
No. Flights
Lisboa APP_1- LPPT_21_B - Runway 21Traffic 12 July 2013 + 10% - 5 iterations
TC (%) Polynomial (TC (%))
62
Figure 83: Graphic of key indicators – Lisboa
APP 2 – 4 sector scenario RWY 21
Figure 84: Graphic of key indicators – Lisboa
TMA lower – 4 sector scenario RWY 21
Figure 85: Graphic of key indicators – Lisboa
TMA upper – 4 sector scenario RWY 21
Figure 86: Graphic of RWY movements – 4
sector scenario RWY 21
6.9 LPPT_21_C
6.9.1 Overview
This scenario pretends to evaluate the impact of the holdstack UMUPI and the STARs BUSEN2D,
GAIOS2D, LIGRA2D, NAKOS2D, TROIA2D, UNPOT2D and XAMAX4D, which use restricted
airspace. Therefore, these STARs were inserted in the simulation.
63
Figure 87: STARs of scenario LPPT_21_C
All other parameters remain equal to the ones of the reference scenario.
6.9.2 Results
The new routes receive some flights. On average, 15% of the flights assigned to STARs that have a
restricted alternative were sent to it by the conflict resolution rules of the simulation.
Flights using conditioned route
Total flights entering in point
%
BUSEN 2 18 11.1%
GAIOS 1 10 10%
LIGRA 0 15 0%
NAKOS 1 7 14.3%
TROIA 0 18 0%
UNPOT 3 20 15%%
XAMAX 12 39 30.1%
TOTAL 19 127 15%
Table 22: Restricted alternative STARs usage
These new STARs lead to a reduction on the number of flights sent to holdstacks and improve the
separation inside TMA as presented in the following table, where it is possible to compare the new
values with the ones obtained in the reference scenario.
LisboaAPP LisboaTMA Tower
AlternateSTAR 14(+4) 60(+22) 6 (-1)
Monitor 148(-15) 52(+3) -
HoldOnGroundRunway - - 188(=)
HoldStackToFrom - - 27(-1)
Table 23: LPPT_21_C - Conflict resolutions applied
64
The UMUPI holdstack is not relevant, receiving only one flight along the day. The holding time in
EKMAR holdstack decreases more than 50% (to 140 seconds), whilst in RINOR the average
increases to 196 seconds.
Overall, and despite the improvements in conflict resolution, the capacity remains 38 movements per
hour, conditioned by the APP sector. It is possible to notice a small reduction in the workload, but it is
not sufficient to increase the capacity of the sector. The STARs in the restricted areas may help and
facilitate traffic separation, but do not lead to an increased capacity.
Figure 88: Graphic of key indicators – Lisboa APP – LPPT_21_C
Figure 89: Regression analysis – Lisboa APP – LPPT_21_C
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa APP- LPPT_21_C - Runway 21Traffic 12 July 2013 + 10% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
y = 0,0173x2 + 1,1547xR² = 0,983
0
10
20
30
40
50
60
70
80
90
0 5 10 15 20 25 30 35 40 45
Wo
rklo
ad (
%)
No. Flights
Lisboa APP - LPPT_21_C - Runway 21Traffic 12 July 2013 + 10% - 5 Iterations
TC (%) Polynomial (TC (%))
65
6.10 LPPT_21_D
6.10.1 Overview
This scenario introduces a reduction in the occupancy times of RWY 21 according to what is
presented in Table 24.
Old RWY Occupancy Times New RWY Occupancy Times
Arrivals Heavy
53s 49s
Arrivals MedJet
49s 46s
Arrivals MedProp
43s 40s
Departures 41s 38s
Table 24: LPPT_21_D - New RWY Occupancy Times
6.10.2 Results
This reduction in the RWY occupancy had strong impact in the average holding time of holdstack
EKMAR, decreasing from 328 seconds to 184s. In RINOR the average time increased to 168s, but the
number of aircraft using it is reduced to an average of 23 flights.
Overall, the capacity is kept within the 38 movements, constrained by the APP sector.
Figure 90: Graphic of key indicators – Lisboa APP – Reduced RWY times RWY21
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa APP - LPPT_21_D - Runway 21Traffic 12 July 2013 + 10% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
66
Figure 91: Regression analysis – Lisboa APP – Reduced RWY times RWY 21
Figure 92: Graphic of key indicators – Lisboa TMA – Reduced RWY times RWY21
The runway presents a behavior similar to the reference scenario, with peaks above the 38
movements corresponding to periods of sector overloading.
y = 0,0173x2 + 1,1934xR² = 0,9785
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35 40 45
Wo
rklo
ad (
%)
No. Flights
Lisboa APP - LPPT_21_D - Runway 21Traffic 12 July 2013 + 10% - 5 Iterations
TC (%) Polynomial (TC (%))
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa TMA - LPPT_21_D - Runway 21Traffic 12 July 2013 + 10% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
67
Figure 93: Graphic of RWY movements – Reduced RWY times RWY 21
6.11 LPPT_21_E
6.11.1 Overview
Using the logic of LPPT_03_E, new SIDs were designed for aircraft departing Lisbon Airport via RWY
21 (Figure 94).
Figure 94: LPPT_21_E: New SIDs RWY 21
In this case, the aircraft departing in the new routes have to fly over the restricted area of Montijo in an
early stage of the flight, as shown in Figure 95.
0
5
10
15
20
25
30
35
40
45
00
:00
00
:50
01
:40
02
:30
03
:20
04
:10
05
:00
05
:50
06
:40
07
:30
08
:20
09
:10
10
:00
10
:50
11
:40
12
:30
13
:20
14
:10
15
:00
15
:50
16
:40
17
:30
18
:20
19
:10
20
:00
20
:50
21
:40
22
:30
23
:20
Runway 21 - LPPT_21_D - Arrivals and DeparturesTraffic 12 July 2013 + 10% - 5 Iterations
Departures Arrivals Total
68
Figure 95: Detail of military areas
6.11.2 Results
The experience of LPPT_03_E, presented in section 6.5 and discussed in 7.1, suggests that this new
configuration, despite reducing departure delays, will not result in increased airspace capacity. For this
scenario there is also the conditioning from the area of Montijo, which would introduce performance
restrictions in the new routes, limiting even more its benefits.
All things considered, it was decided not to submit this scenario into a deeper analysis.
6.12 LPPT_21_F
6.12.1 Overview
This scenario uses the same sectorization of LPPT_03_F, this time applied to RWY 21. In this case,
the APP North sectors works as the approach sector, whilst APP South sector is the departure sector.
Unlike LPPT_03_F, where the departure sector was also crossed by arriving traffic approaching from
north and there was the need to use sector clip in approach sector to deal with departing aircraft
entering it for a short period of time, in the case of this scenario these situations happen very rarely.
The exceptions are flights approaching from south that, because of their low performance or due to
conflict resolution manoeuvres applied, are flying below FL085 when crossing the area of the
departure sector and, consequently, entering it.
6.12.1 Results
This sectorization increases the capacity for up to 44 movements per hour, at which the sector of
Lisboa TMA attains saturation.
69
Figure 96: Graphic of key indicators – Lisboa TMA – New sectors RWY21
Figure 97: Regression analysis – Lisboa APP – New sectors RWY 21
With the APP sector divided in APP Sector North (approach sector) and APP Sector South (departure
sector), the workload that in the previous sectors would lead to saturation is now on charge of two
controller teams. As it is possible to see in Figure 98 and Figure 99, the new sectors do not saturate.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa TMA - LPPT_21_F - Runway 21Traffic 12 July 2013 + 15% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
y = 0,0131x2 + 0,9909xR² = 0,9699
0
10
20
30
40
50
60
70
80
0 10 20 30 40 50
Wo
rklo
ad (
%)
No. Flights
Lisboa TMA - LPPT_21_F - Runway 21Traffic 12 July 2013 + 15% - 5 Iterations
TC (%) Polynomial (TC (%))
70
Figure 98: Graphic of key indicators – Lisboa APP North – New sectors RWY21
Figure 99: Graphic of key indicators – Lisboa APP South – New sectors RWY21
With the traffic sample used in this simulation the RWY presents a peak of only 40 movements.
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa APP - LPPT_21_F - Runway 21Traffic 12 July 2013 + 15% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
0
10
20
30
40
50
60
70
80
90
100
00
:00
01
:00
02
:00
03
:00
04
:00
05
:00
06
:00
07
:00
08
:00
09
:00
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
24
:00
No
. Flig
hts
, N
o. C
on
flic
ts o
r P
erce
nta
ge
Hours
Lisboa APP South - LPPT_21_F - Runway 21Traffic 12 July 2013 + 15% - 5 iterations
TC (%) Flights Peak Occ. Conflicts
71
Figure 100: Graphic of RWY movements – New sectors RWY 21
Contrary to what happened in RWY 03, the average impacts of conflict resolution did not raise so
clearly relatively to the reference scenario. Some parameters even present improvements.
Sector Distance Impact (NM) Time Impact (minutes)
Lisboa APP North 11.5 3.9
Lisboa APP South 65.45 27
Lisboa TMA 1.4 2.1
Departures delay n.a. 2.4
Table 25: Conflict Resolution Impacts - New APP sectors scenario RWY 21
As for the holdstack usage, although the average time in EKMAR decreased considerably, the
increased traffic of this scenario lead, inevitably, to a considerably higher number of aircraft held at
RINOR.
Holdstack No. Flights Received Average holding time
(seconds)
EKMAR 3 209
RINOR 38 158
Table 26: Holdstack data - New APP sectors scenario RWY 21
5 Only one flight has contributed to this value.
0
5
10
15
20
25
30
35
40
45
00
:00
00
:50
01
:40
02
:30
03
:20
04
:10
05
:00
05
:50
06
:40
07
:30
08
:20
09
:10
10
:00
10
:50
11
:40
12
:30
13
:20
14
:10
15
:00
15
:50
16
:40
17
:30
18
:20
19
:10
20
:00
20
:50
21
:40
22
:30
23
:20
Runway 21 - LPPT_21_F - Arrivals and DeparturesTraffic 12 July 2013 + 15% - 5 Iterations
Departures Arrivals Total
72
7. Discussion
7.1 Scenarios operating with RWY 03
The reproduction of the current Lisbon Airport Terminal Airspace and the calibration of the scenario
allowed to obtain a reference scenario with a capacity of 38 movements, consistent with the declared
values of the airport. Despite the peak of 37 movements entering the APP sector between 11:00 a.m.
and 12:00 p.m., the RWY presented a maximum of only 36 movements. That does not necessarily
means that it is saturated and not able to perform the same 37 operations in one hour. The data is
analysed using a sliding window with 10 minutes steps and, unlike the workload generated by a flight
which is distributed along time, a RWY departure or arrival is an “instantaneous” event. Consequently,
the 10 minutes step may not capture the highest peak.
The conflict resolution summary presents a high number of conflicts solved by monitor in APP sector.
This does not always mean that they are unsolved conflicts. An example of a situation when this might
occur is in case of having 3 flights, where flight 2 is in conflict with flight 1 and flight 3 is in conflict with
flight 2. While solving the conflict between flights 2 and 1, applying a manoeuvre to flight 2, it also
terminates the problem between flights 2 and 3. As the software detected a conflict between those 2
flights, and did not solve it directly, it monitors the situation. The fundamental part of the conflict
resolution procedure is the one related with the RWY, to avoid flights entering the RWY out of their
time. Once that is successfully done, and because of the complexity of a TMA, it is not unusual
between RAMS users to do not even try to solve separation conflicts, only detecting them and
allocating time to account for the workload that would be generated when solving it.
The division of the airspace into the four sectors declared in AIP, as done in the scenario LPPT_03_B,
only increases the capacity of the TMA sector. Although it splits the saturated APP sector into two, the
APP Sector 2, being only between 1000ft and 2000ft, is practically unused and only departing traffic
with low climb performance enters it for a very short period of time. For arriving aircraft, being the FAP
assigned to an altitude of 3000ft, this sector is never used. In actual fact, the “new” APP Sector 1 has
exactly the same work as the former and saturated APP sector. This division exists due to former
procedures related with Cascais Aerodrome (LPCS) and is currently outdated, not adding capacity to
the airspace.
In LPPT_03_C, despite the addiction of the holdstack UMUPI and the STAR XAMAX4C, the capacity
remained unchanged. Because of the traffic pattern, with the majority of the traffic approaching from
STARs INBOM and XAMAX, it is possible to separate it with the simplified trombone existing for the
approach of RWY 03, putting the alternative to the original XAMAX4A into secondary status. The
holdstack UMUPI receives considerable traffic, but mostly because of its location. Instead of deviating
the flights to EKMAR, they are held earlier in UMUPI, in a less congested zone. Although this might
simplify the work of the controller, it does not seem to be enough to increase the capacity of the
73
sector. It is also possible to notice the reduction of the time in hold of traffic sent to UMUPI. The
reason for that is mostly its location which, in consequence of being further from the original flight
paths, creates a delay itself on the distance that aircraft fly between its original flight plans into the
holdstack.
With a reduction of RWY occupancy times, in LPPT_03_D, no improvements where obtain apart from
a reduction in the average time of flights in holdstacks. The average delay of late departures remains
in the 1.8 minutes of the reference scenario, suggesting that it is caused by the separations between
takeoffs and landing and not because of the RWY occupancy times. This result is consistent with the
declared capacities of the airport, where is stated that it is constrained because of the TMA and not
because of lack of RWY capacity.
In LPPT_03_E, the new SIDs created to reduce time separation between departures did not contribute
to improve the airspace. Instead, the extra complexity generated by the new routes brought the
capacity down to the 37 movements. The conflicts caused by the merging of some departure routes
could be easily avoided introducing altitude restrictions, but this would create the need to stop the
climb of some flights, with obvious economic and environmental consequences. Furthermore, because
of the dominant type of operation performed in Lisbon Airport (arrival – departure – arrival), the
reduction of the separation to one minute would not bring gains during most periods of high
demanding. The excellent average departure delay presented is obtained from the periods with fewer
arrivals and more departures, reducing their time in departure queues. Typically, these are not periods
of peak workload.
The new sectorization, introduced in scenario LPPT_03_F, divides the APP sector in two. Although
the sectorization is imposed in a geographical basis, it has in mind a functional sectorization, creating
an approach sector and a departure sector. In this case, the constraint caused by the baseline
sectorization is eliminated and the capacity of the airspace is enhanced. The considerable increase in
the average stack holding time, together with an apparent incapability to deal with all the flights
entering in the APP sectors, with peaks up to 47 movements combining the two sectors, denotes a
RWY saturation. To confirm it, a quick estimate will be made.
The total cycle time, i.e., the typical time interval for each landing and takeoff cycle, of a runway
operating in mixed mode consists of three factors [60]:
1) Time needed to guarantee a minimum separation of 3NM between an aircraft starting the
takeoff roll and an approaching flight;
2) ROTA;
3) ROTD.
Considering a 160kt speed on final approach, the first parameter is calculated in (1)
60∗3
160= 1.125 min = 67.5𝑠 (1)
Estimating the ROTA with 80% medium jets, 15% heavies and 5% medium propellers:
74
51.1 × 0.8 + 57.5 × 0.15 + 53.6 × 0.05 = 52.185 ≅ 52𝑠 (2)
and with a ROTD of 43 seconds, the total cycle time is obtained:
67.5 + 52 + 43 = 162.5𝑠 (3)
Figure 101: Total Cycle [60]
The capacity of the RWY, with mixed operations and for the period of an hour is then:
𝑅𝑊𝑌𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 =3600
162.5× 2 = 44 𝑚𝑜𝑣𝑒𝑚𝑒𝑛𝑡𝑠 (4)
which confirms that the runway attained its maximum capacity in LPPT_03_F simulation. All
considered, the capacity of this scenario is 44 movements, despite the 48 allowed by the airspace.
Introducing the RWY occupancy times of LPPT_03_D, the capacity of the RWY could be increased to
up to 46 movements per hour.
7.2 Scenarios operating with RWY 21
The results obtained for RWY 21 are in all similar to the ones of RWY 03. The most considerable
difference between them lays in the resolution of conflicts, with higher recurrence to the holdstacks.
This behavior was expected due to the orientation of the RWY and the dominant traffic patterns. With
the majority of traffic entering the TMA from north and heading almost directly towards the RWY, it is
more difficult to sequence and give it the necessary separations. Some changes in the heading of the
approaching traffic may allow the controller to gain some time, but they are always limited because of
the shape of the TMA, with military areas on the one side and SIDs on the other side of the approach
routes, making it susceptible of conflicts between arriving and departing traffic. At the end, the best
compromise is to hold aircraft and, hence, the use of holdstacks is considerably higher for this RWY.
In LPPT_21_B, it seems to exist no immediate logical justification for the increase of the capacity to 39
movements in APP Sector 1. In a deeper analysis, the reason for this increment was found in a
software model for vertical trajectory. Although the final approach point of RWY 21 (FAP21) has
defined a minimum altitude of 4000ft, the altitude restrictions are occasionally not observed, as well as
75
speed limitations. Because of that, eventually one or another flight do not respect the altitude
restrictions and enters Lisboa APP 2, which it should not due to the limitation of 4000ft on FAP (it
should exit from APP 1 directly to Lisboa CTR). As a result, the workload of the controller in APP
Sector 1 is, mistakenly, alleviated and therefore the capacity increases.
The introduction of the alternative STARs using restricted airspace is more relevant in this case of
RWY21, as the higher number of routes of this type could suggest. These alternatives received an
average of 15% of the traffic entering the TMA via STARs with alternatives pending on military
exercises. Although the estimated capacity remains 38 movements, the reduction in the number of
conflicts monitored suggests a decrease in the complexity of the airspace. It is possible to notice a
reduction on the total workload of the tactical controller. Despite not being the solution for a capacity
enhancement, these routes may play an important role and provide a useful alternative to the narrow
air corridor between the airport and the military area of Sintra.
Regarding scenarios LPPT_21_D and LPPT_21_E, similar conclusions could be drawn as for the
equivalent cases of RWY 03. The alternative SIDs for RWY 21 would have an even more limited
beneficial impact because of the restrictions that would have to be applied due to the restricted area of
Montijo.
In LPPT_21_F, the new sectorization also improved the airspace capacity. As for the scenario
LPPT_03_F, it is limited by the TMA sector, but with a lower capacity of 44 movements. Studying the
configuration of the airspace, the cause for this lower capacity is easily identified. Unlike the
configuration for RWY 03, the most used holdstack for RWY 21 is in the TMA sector, creating more
workload for its controller team.
Using the previously employed method to estimate RWY capacity, it is possible to find a capacity of 45
movements per hour for RWY 21 with the standard runway occupancy times (if the reduced times
were applied, it could go up to 46 movements). Considering that the traffic sample used never attained
these peak values, the average delays and impacts of conflict resolution remained at levels similar to
the reference scenario.
7.3 Workload
The workload calculation in RAMS is determined by the task list. Therefore, the design of a task list
and the estimations of each task duration (weight) requires great precision and shall be based on
observation of controllers during performance in a RTS environment [59]. Because of that, the values
of the new capacities shall be carefully treated. Although the model was calibrated with the reference
scenario, the weighing between standard tasks and extra workload created by conflict resolution may
not have been accurately determined. Nevertheless, the analysis has the merit of accessing the
scenario relative changes impact to the workload, being possible to compare them in a qualitative
way.
76
8. Design Proposal
8.1 Description
The simulations performed in chapter 6, combined with the analysis done in chapter 7, allowed to
understand the impact of different changes in the current airspace of Lisbon Airport and their eventual
benefits. It is time now to, from the knowledge obtained, suggest a final scenario which maximizes the
capacity.
From the previous scenarios, the most beneficial change was the new sectorization, with the former
APP sector being divided into approach and departure sectors. As far as RWY 03 is concerned, this
would generate sufficient airspace capacity to keep up with the capacity of the RWY, even if the
reduction of the average occupancy times analysed in scenario LPPT_03_D was attained.
The same was not verified for RWY 21 where, although the new sectors would raise the airspace
capacity for a value close to the one estimated for the runway with the present occupancy times, an
eventual future reduction in those times would, once again, put the airspace capacity below the
runway capacity. The main reason for this limitation is the location of the most used terminal holdstack
of RWY 21, which is positioned in the TMA, unlike the holdstacks of RWY 03 which are located in the
APP sector. Because of that, the TMA presents a lower capacity when RWY 21 is operating. To solve
this, the first change considered was to increase the diameter of the APP sector in order to include the
RINOR holdstack. However, this would imply the APP sectors to apply different separations inside the
sector, since the Lisboa Airport Radar Vectoring Area is confined to the 30NM distance. As an
alternative, the RINOR holdstack was replaced by a new holdstack in the waypoint Nav221, already
inside the APP sector and about 3.5NM away from the original RINOR holdstack.
Figure 102: Nav221HS holdstack
The alternative STARs using restricted areas studied in LPPT_03_C and LPPT_21_C, although not
leading to a capacity increase, allowed a better traffic sequencing and reduced the number of
conflicts, as well as the use of holdstacks, particularly in RWY 21. A good operational coordination is
necessary in order to, for example, liberate this portion of airspace to be used by civil aviation during
77
peak traffic demand periods and restricting it for military activities in the remaining time. These routes
were also included in this scenario.
The RWY occupancy times used in this simulation consider the reduction studied in LPPT_03_D and
LPPT_21_D to explore the capacity of this scenario at the maximum.
8.2 Improvements
Overall, these changes allowed to remove the constraints created by the airspace in the current
design and give enough margin to deal with eventual reductions in runway occupancy times and the
consequent runway capacity increase obtained.
By removing the RINOR holdstack from the TMA sector, the capacity of the airspace under operation
of RWY 21 increased from the 44 movements of LPPT_21_F up to 48 movements, equaling the
capacity estimated for the airspace when RWY 03 is operating.
The function that gives the workload versus the number of sector entries has a very low quadratic
component in these cases. As mentioned in chapter 2, the origin of the quadratic behavior is mostly in
the conflicts. Given that in this case they are mostly solved in the APP sectors, the function acquires
an almost linear shape.
Figure 103: Graphic of key indicators - Lisboa
TMA – Final design RWY 03
Figure 104: Regression analysis – Lisboa TMA –
Final design RWY 03
78
Figure 105: Graphic of key indicators - Lisboa
TMA – Final design RWY 21
Figure 106: Regression analysis – Lisboa TMA–
Final design RWY 21
The APP sectors in both cases are far from the saturation, as it is shown by the following graphics.
Figure 107: Graphic of key indicators - Lisboa
APP_North - Final Design RWY 03
Figure 108: Graphic of key indicators - Lisboa
APP_South- Final Design RWY 03
Figure 109: Graphic of key indicators - Lisboa
APP_North - Final Design RWY 21
Figure 110: Graphic of key indicators - Lisboa
APP_South - Final Design RWY 21
79
Observing the graphics of RWY movements, it is possible to identify a peak of 46 movements in RWY
03, coinciding with its estimated capacity. For RWY 21, the maximum presented is of 41 movements,
since the used traffic sample, unlike what happens in the one from RWY 03, never presented
simultaneous peaks of arrivals and departures.
Figure 111: Graphic of RWY movements - Final
design RWY 03
Figure 112: Graphic of RWY movements - Final
design RWY 21
In general, the average delays do not differ substantially from the reference scenarios. The averages
in stack holding times are presented in tables 27 and 28.
Holdstack No. Flights Received
Average holding time (seconds)
ADSAD 5 402
EKMAR 2 205
UMUPI 2 27
Table 27: Holdstack data – Final Design RWY 03
Holdstack No. Flights Received
Average holding time (seconds)
EKMAR 3 124
Nav221_HS 22 161
UMUPI 1 81
Table 28: Holdstack data – Final Design RWY 21
The average delay for late departures is 3.1 minutes in RWY 03 and 2.6 minutes in RWY 21.
80
9. Conclusions
This was an ambitious study, focusing on a theme that is not new but has increased relevance today,
which is the saturation of Lisbon Airport. The software used, RAMS Plus, is used worldwide but by a
very small community of users. Being very specific and restricted, it required a lot of dedication and
persistence to obtain realistic simulations and results.
The discovery, learning and adaptation to the software were done with the reference scenarios.
Because there were no task lists defined to Lisbon, they were also fundamental to build and calibrate
the list of tasks. At the end, the capacity obtained was equal to the declared one.
From the several new scenarios created, it was possible to verify what they would, or would not, add
to the reference scenario. The division in four sectors did not improve in any way the airspace nor did
the new SIDs since the few benefits obtained did not overcome the additional complexity created. The
use of the routes and holdstacks in the military areas could simplify and reduce the work complexity of
the ATCo but, alone, did not demonstrate capability to add capacity to the airspace. The reduction in
RWY occupancy times was useful to reduce delays but, the RWYs having a capacity higher than the
airspace, there was no logic (neither significant results) on applying it without improving the airspace.
The most successful change analysed was the division of the APP sector in two, applying an implicit
functional sectorization.
The final design gets together the different measures that were determined as advantageous to
improve the capacity of the airspace. Together with a reduction of the runway occupancy times, it
boosts the airspace capacity to 48 movements, 10 more than the current declared capacity and more
than the runway can deal with realistic occupancy times.
There is still work to be done in the context of this study. The validation of new ATM systems requires
several steps, as described in chapter 3, and it is essential to use a more reliable task list. Also,
because the capacity of an airport does not depend exclusively on airspace and runways, a simulation
including the ground operations, resorting on RAMS’ ground module, would be important to identify
eventual constraints, particularly on the capacity of the gates.
81
References
[1] Cook, A. (2007). European Air Traffic Management - Principles, Practice and Research. Ashgate.
[2] Airbus Group. (2012). Navigating the Future - Global Market Forecast 2012-2031. Airbus.
[3] ANA - Aeroportos de Portugal. (2013). Plano Estratégico 2013_2017. ANA.
[4] SOL Online. (14-10-2012). 'É inevitável um novo aeroporto' em Lisboa. Retrieved from
http://sol.sapo.pt/inicio/Sociedade/Interior.aspx?content_id=60963 a 20-02-2014
[5] Diário de Notícias. (31-05-2012). Aeroporto da Portela esgotado em 5 anos. Retrieved from
http://www.dn.pt/inicio/economia/interior.aspx?content_id=2562340 a 20-02-2014
[6] Expresso. (27-10-2012). Novo aeroporto só depois de 2022. Retrieved from
http://expresso.sapo.pt/novo-aeroporto-so-depois-de-2022=f762571 a 20-02-2014
[7] Publico. (17-07-2013). O novo aeroporto de Lisboa é na Portela. Retrieved from
http://www.publico.pt/economia/noticia/o-novo-aeroporto-de-lisboa-e-na-portela-diz-secretario-
de-estado-dos-transportes-1600497 a 20-02-1990
[8] HM Waterguard. (n.d.). Retrieved from http://www.hm-
waterguard.org.uk/Offices%20&%20Buildings-England.htm em 23-04-2014
[9] EUROCONTROL. (n.d.). Description of the CAPAN Method.
[10] ISA Software. (2013, December). RAMS Plus User Manual. Release 6.0.
[11] Diab, N. (2008). Airport Capacity and Delay - Airport Planning & Management. Florida:
Everglades University.
[12] Ashford, N., & Wright, P. H. (1992). Airport Engineering. New York: John Wiley & Sons.
[13] Majumdar, A., Ochieng, W. Y., McAuley, G., Lenzi, J. M., & Lepadatu, C. (2005). The factor
affecting airspace capacity in Europe: A Framework Methodology based on Cross Sectional
Time-Series Analysis using Simulated Controller Workload Data. 6th USA/Europe ATM
Research and Development Seminar. Baltimore.
[14] Völckers, U., & Böhme, D. (1997). Dynamic Control of Ground Movements: State-of-the-Art
Review and Prespectives. AGARD Report 825 (pp. 2-1 - 2-11). Budapest: NATO.
[15] Neufville, R. d., & Odoni, A. (2003). Airport Systems: Planning, Design and Management.
McGraw-Hill.
[16] Majumdar, A. (2004). Developments in airspace capacity and safety research. Synergising ATC-
Airports-Airlines to meet Contemporary Challenges in Civil Aviation. New Delhi. Retrieved 07
82
22, 2014, from
http://www.atcguild.com/MEMBERS/Sem04/Slides/Dr.%20Arnab%20Majumdar.pdf
[17] Juridic, B., Babic, R. S., & Francetir, I. 2011). Zagreb Terminal Airspace Analysis. Zagreb.
Retrieved 02 24, 2014, from hrcak.srce.hr/file/12205
[18] Christien, R., & Benkouar, A. (2003). Air Traffic Complexity Indicators & ATC Sectors
Classification. Budapest. Retrieved 05 04, 2014, from
http://www.atmseminar.org/seminarContent/seminar5/papers/p_044_MPM.pdf
[19] Kopardekar, P., Scwartz, A., Magyarits, S., & Rhodes, J. (2009). Airspace Complexity
Measurement: An Air Traffic Control Simulation Analysis. International Journal of Industrial
Engineering, 61-70.
[20] Masalonis, A. J., Callaham, M. B., & Wanke, C. R. (2014, 05 04). Dynamic Density and
Complexity Metrics for Realtime Traffic Flow Management. Virginia, USA.
[21] Sherry, L. (2008). Runway Operations: Computing Runway Arrival Capacity. Virginia, George
Mason University. Retrieved 06 21, 2014, from
http://catsr.vse.gmu.edu/SYST460/RunwayCapacity_LectureNotes.pdf
[22] Sherry, L. (2009). Capacity of a Single Runway. In Runway Capacity Workbook. Virginia, George
Mason University. Retrieved 06 21, 2014, from
http://catsr.vse.gmu.edu/SYST460/RunwayCapacityWorkbook.pdf
[23] Stein, E. S. (1985). Air Traffic Controller Workload: An Examination of Workload Probe. Atlantic
City, FAA. Retrieved from http://hf.tc.faa.gov/technotes/dot-faa-ct-tn84-24.pdf
[24] Garcia-Avello, C., & Swierstra, S. (1997). Human Role in ATM: Suppor for Decision Making.
AGARD Report 825 (pp. 10-1 - 10-11). Budapest: NorthAtlantic Treaty Organization.
[25] Brooker, P. (2003). Control workload, airspace capacity and future systems. Human Factors and
Aerospace Safety, 3(1) (pp. 1-23). United Kingdom: Ashgate Publishing.
[26] Reichmoth, J. (1997). PHARE Demonstration: Arrivals Management. AGARD Report 825 (pp. 12-
1 - 12-12). Budapest: NorthAtlantic Treaty Organization.
[27] EUROCONTROL. (2013). European Airspace Design Methodology Guidelines - General
Principles and technical specifications for airspace design. European Route Network
Improvement Plan (part 1).
[28] Tobaruela, G., Majumdar, A., & Ochieng, W. Y. (2012). Identigying Airspace Capacity Factors in
the Air Traffic Management System. Proceedings of the 2nd International Conference on
Application and Theory of Automation in Command and Control Systems (pp. 219-224).
London: Irit Press.
83
[29] Wilson, I. A. (2000). ATM Support Tools in PHARE - The Importance of Matching the Concepts of
the Management or Control. 3rd USA Europe Air Traffic Management R&D Seminar. Napoli.
Retrieved 07 19, 2014, from
http://www.atmseminar.org/seminarContent/seminar3/papers/p_055_DSTCDM.pdf
[30] Crook, I., & Liang, D. (2003). FAA Operational Concept Validation - High Altitude Airspace
Analysis. ATM Seminar. Budapest. Retrieved 09 05, 2014, from
http://www.atmseminar.org/seminarContent/seminar5/papers/p_034_MPM.pdf
[31] Mirkovic, B. (2009). Building the Trust in the Model: RAMS Plus Case. 36th Symposium on
Operational Research. Belgrade. Retrieved 09 05, 2014, from
http://www.sf.bg.ac.rs/downloads/katedre/apatc/Symopis09_BojanaMirkovic.pdf
[32] Majumdar, A. (2003). A Framework For Modelling The Capacity Of Europe’s Airspace Using A
Model Of Air Traffic Controller Workload. Doctoral Thesis, University of London, London.
[33] Ackermann, I. (2009). Prague TMA and Lower Airpsace 2009+ - Model Based Simulation Report.
EUROCONTROL.
[34] EUROCONTROL. (2011). Fast-time Simulation Tools. Retrieved 09 12, 2014, from
EUROCONTROL Website:
http://www.eurocontrol.int/eec/public/standard_page/WP_Fast_Time_Simulation_Tools.html
[35] Airtopsoft SA. (2007). AirTOp. Retrieved 09 12, 2014, from AirTOpsoft - Air Traffic Optimization
website: http://www.airtopsoft.com/products.html
[36] ISA Software. (2013). RAMS Plus. Retrieved 09 12, 2014, from RAMS Plus website:
www.ramsplus.com
[37] Jeppesen. (2014). Total Airspace and Airport Modeler (TAAM). Retrieved 09 12, 2014, from
Jeppesen - A Boeing Company website: http://ww1.jeppesen.com/industry-
solutions/aviation/government/total-airspace-airport-modeler.jsp
[38] Airport Research Center. (2014). CAST Products. Retrieved 09 12, 2014, from Airport Research
Center website: http://www.airport-
consultants.com/index.php?option=com_content&view=section&id=6&Itemid=30
[39] The Watt Committee on Energy. (1981). Factors Determining Energy Costs and an Introduction to
the Influence of Electronics. London: The Watt Committee on Energy Ltd.
[40] EUROCONTROL. (2014). Real Time Simulations. Retrieved 09 12, 2014, from EUROCONTROL
website: http://www.eurocontrol.int/articles/real-time-simulations
[41] Whysall, P. (1998). Future Area Control Tools Support (FACTS). 2nd USA/Europe Air Traffic
Management R&D Seminar. Orlando.
84
[42] Universidad Politécnica de Catalunha. (2012). Software Licenses and Computational Facilites.
Retrieved 09 13, 2014, from Grupo de investigation ICARUS website:
http://www.icarus.upc.edu/facilities/computational-facilities?set_language=es
[43] EUROCONTROL. (2014). Lisboa LIS / LPPT airport information. Retrieved 04 10, 2014, from
EUROCONTROL Public Airport Corner:
https://www.eurocontrol.int/airport_corner_public/LPPT
[44] Aeroportos do Mundo. (2014). Aeroporto da Portela - Lisboa. Retrieved 09 18, 2014, from
Aeroportos do Mundo:
http://www.aeroportosdomundo.com/europa/portugal/aeroportos/portela.php
[45] ANA Aeroportos de Portugal SA. (2014). Aeroporto de Lisboa - Conheça o Aeroporto. Retrieved
09 18, 2014, from ANA Aeroportos de Portugal: http://www.ana.pt/pt-
PT/Aeroportos/Lisboa/Lisboa/OAeroporto/ConhecaoAeroporto/Paginas/Conheca-o-
Aeroporto.aspx
[46] ANA Aeroportos de Portugal SA. (2008). Plano de Expansão - O Aeroporto de Lisboa está mais
aeroporto. Retrieved 09 18, 2014, from
http://www.lisbonairportdevelopmentplan.com/home_v.html
[47] NAV Portugal, E.P.E. (2014). eAIP Portugal. eAIS Package. Retrieved from
http://www.nav.pt/ais/cd/2014-07-24-AIRAC/html/index.html
[48] Piavani, S. (2010). Aircraft Pictures. Retrieved 09 18, 2014, from Airliners.net:
http://www.airliners.net/photo//1867899/L/&sid=79e8cb02f108aafb85a6014d0647dcab
[49] NAV Portugal E.P.E. (2011). Lisbon Capacity Enhancement Exercise 2010-2011.
[50] Pavliv, S., Zuzic, M., & Pavicic, S. (2006). Runway Occupancy Time as Element of Runway
Capacity. Promet - Traffic & Transportation, Vol. 18, 2006, No.4, 293-299.
[51] Rocha, L. G. (1982). Air Traffic Services in Portugal: Civil-Military Coordination Aspects. AGARD
Conference Proceedings No.340. Lisbon: NATO.
[52] International Civil Aviation Organization. (2007). Air Traffic Management - Procedures for Air
Navigation Services (Doc 4444) (Fifteen Edition ed.). ICAO.
[53] MeteoLoures. (2011). Porque é que, normalmente, quando está Sol os aviões descolam/aterram
dum lado, e quando está chuva do outro? Retrieved 09 18, 2014, from MeteoLoures:
http://meteoloures.webnode.pt/news/porque%20e%20que,%20normalmente,%20quando%20
esta%20sol%20os%20avi%C3%B5es%20descolam-
aterram%20dum%20lado,%20e%20quando%20esta%20chuva%20do%20outro-/
[54] Instituto Nacional de Aviação Civil, I.P. (2013). Boletim Estatístico Trimestral N.º 17 - 1º Trimestre
2013. Lisboa: INAC.
85
[55] Instituto Nacional de Aviação Civil, I.P. (2013). Boletim Estatístico Trimestral N.º 18 - 2º Trimestre
2013. Lisboa: INAC.
[56] Instituto Nacional de Aviação Civil, I.P. (2013). Boletim Estatístico Trimestral N.º 19 - 3º Trimestre
2013. Lisboa: INAC.
[57] Instituto Nacional de Aviação Civil, I.P. (2014). Boletim Estatístico Trimestral N.º 20 - 4º Trimestre
2013. Lisboa: INAC.
[58] FlightRadar24. (2014). flightradar24 - Live Air Traffic. Retrieved from www.flightradar24.com
[59] Cujic, M. (2004). An Analysis of the Task List Impact Upon RAMS Workload Calculations.
Retrieved 10 06, 2014, from
http://www.sf.bg.ac.rs/downloads/katedre/apatc/ICRAT2004_Cujic.pdf
[60] Hong Kong International Airport. (n.d.). Hong Kong International Airport Master Plan 2030 -
Technical Report. Retrieved 10 08, 2014, from
http://hkia3way.blob.core.windows.net/pdf/en/TR_24May_Eng_Full.pdf
86
Appendix A – Conflict Resolution Flowchart
87
Appendix B – Task List
Object Trigger Task Offset Sector Weight
ALLALTERNATESIDS ATCAlternateSID RadarInterventionForDep 0 SectorCurrent 3
ALLSTARS ATCStarEntry IntermediateApproachSequence 0 SectorCurrent 2
ALLSTARS ATCStarEntry TxCoordination 0 SectorCurrent 2
ALLSTARS ATCStarEntry TxCoordination 0 SectorNext 2
ALLCENTERS ATCSectorPierce CFSToEstablishSectorEntryClearance 0 SectorCurrent 4
ALLCENTERS ATCSectorPierce Rx1stCall -20 SectorCurrent 3
ALLCENTERS ATCSectorPierce Acknowledge1stCall -15 SectorCurrent 3
ALLCENTERS ATCSectorPierce AssumeFlight 0 SectorCurrent 2
ALLCENTERS ATCSectorPierce CoorditationForLocalArrival 0 SectorCurrent 5
ALLCENTERS ATCSectorPierce CoordiantionForLocalDeparture 0 SectorCurrent 5
ALLCENTERS ATCSectorPierce RxHandoff -120 SectorCurrent 3
ALLCENTERS ATCSectorPierce CFSToEstablishInitialFLClearance -170 SectorCurrent 8
ALLCENTERS ATCSectorPierce CFSToEstablishSectorExitClearance -160 SectorCurrent 4
ALLCENTERS ATCSectorClipEntry CFSToEstablishSectorEntryClearance -60 SectorCurrent 5
ALLCENTERS ATCSectorSkipEntry CFSToEstablishSectorEntryClearance -60 SectorCurrent 5
ALLCENTERS ATCCruiseStart RadarSupervision 0 SectorControl 3
ALLCENTERS ATCSectorExit TxNewFreq -70 SectorCurrent 4
ALLCENTERS ATCSectorExit RxReadbackCOF -60 SectorCurrent 4
ALLCENTERS ATCSectorExit TxHandoff -22 SectorCurrent 4
ALLCENTERS ATCSectorExit RxReadbackHandoff -15 SectorCurrent 4
ALLCENTERS ATCSectorExit CoordinateForSpeedControl -60 SectorCurrent 4
ALLCENTERS ATCSectorPierce SpeedControlForArrSeq 0 SectorCurrent 4
ALLCENTERS ATCSectorPierce TxNewFL 0 SectorCurrent 6
ALLCENTERS ATCSectorPierce CoordinateForSpeedControl -55 SectorCurrent 3
ALLCENTERS ATCSectorPierce InitialSequencingForArrivals 60 SectorCurrent 3
ALLCENTERS ATCTacticalConflictFound UseExtPredVector 0 SectorCurrent 3
ALLCENTERS ATCTacticalConflictFound RadarInterventionForCF 0 SectorCurrent 10
ALLCENTERS ATCTacticalConflictResolution RadarInterventionForCF 0 SectorCurrent 60
ALLHOLDSTACKS ATCHoldStackEntry TxInstructionToEnterHold -180 SectorCurrent 5
ALLHOLDSTACKS ATCHoldStackEntry ManageHoldStack -60 SectorCurrent 10
ALLHOLDSTACKS ATCHoldStackLevelChange ManageHoldStack -10 SectorCurrent 10
ALLHOLDSTACKS ATCHoldStackLevelReached RxFlightLevelReachedReport 60 SectorCurrent 3
ALLHOLDSTACKS ATCHoldStackExit TxInstructionToLeaveHold 0 SectorCurrent 5
ADSAD ATCNavAid RadarInterventionForArrSeq 0 SectorCurrent 5
EKMAR ATCNavAid RadarInterventionForArrSeq 0 SectorCurrent 5
Nav272 ATCNavAid RadarVectoringForBaseTurn -5 SectorCurrent 7
Nav273 ATCNavAid ClearanceILSForFinal 0 SectorCurrent 7
Nav275 ATCNavAid ClearanceILSForFinal 60 SectorCurrent 7
Nav275 ATCNavAid RadarVectoringForBaseTurn -5 SectorCurrent 7
Nav276 ATCNavAid ClearanceILSForFinal 60 SectorCurrent 7
Nav276 ATCNavAid RadarVectoringForBaseTurn -5 SectorCurrent 7
Nav277 ATCNavAid ClearanceILSForFinal 60 SectorCurrent 7
Nav277 ATCNavAid RadarVectoringForBaseTurn -5 SectorCurrent 7
88
Nav278 ATCNavAid ClearanceILSForFinal 60 SectorCurrent 7
Nav278 ATCNavAid RadarVectoringForBaseTurn -5 SectorCurrent 7
Nav279 ATCNavAid ClearanceILSForFinal 60 SectorCurrent 7
Nav279 ATCNavAid RadarVectoringForBaseTurn -5 SectorCurrent 7
Nav280 ATCNavAid ClearanceILSForFinal 60 SectorCurrent 7
Nav280 ATCNavAid RadarVectoringForBaseTurn -5 SectorCurrent 7
Nav281 ATCNavAid ClearanceILSForFinal 60 SectorCurrent 7
Nav281 ATCNavAid RadarVectoringForBaseTurn -5 SectorCurrent 7
Nav282 ATCNavAid ClearanceILSForFinal 60 SectorCurrent 7
Nav282 ATCNavAid RadarVectoringForBaseTurn -5 SectorCurrent 7
Nav283 ATCNavAid ClearanceILSForFinal 60 SectorCurrent 7
Nav283 ATCNavAid RadarVectoringForBaseTurn -5 SectorCurrent 7
Nav284 ATCNavAid ClearanceILSForFinal 0 SectorCurrent 7
Nav285 ATCNavAid ClearanceILSForFinal 0 SectorCurrent 7
Nav286 ATCNavAid ClearanceILSForFinal 0 SectorCurrent 7
Nav287 ATCNavAid ClearanceILSForFinal 0 SectorCurrent 7
Nav288 ATCNavAid ClearanceILSForFinal 0 SectorCurrent 7
Nav289 ATCNavAid ClearanceILSForFinal 0 SectorCurrent 7
Nav290 ATCNavAid ClearanceILSForFinal 0 SectorCurrent 7
Nav291 ATCNavAid ClearanceILSForFinal 0 SectorCurrent 7
Nav292 ATCNavAid ClearanceILSForFinal 0 SectorCurrent 7
