Airport Capacity Constraints and Strategies for Mitigation: A Global Perspective 0128126574, 9780128126578

Table of contents :
Cover
Airport Capacity Constraints and Strategies for Mitigation: A Global Perspective
Copyright
Contents
Foreword
Part I: Basic concepts
1 Introduction
References
2 Concepts of capacity and methods of estimation
2.1 Overview: what do we mean by capacity?
2.2 Main factors influencing capacity
2.3 Approaches to estimate runway capacity
2.3.1 Queuing models
2.3.2 Analytical approaches
2.3.3 Empirical approaches
2.3.4 Simulation models
2.4 Conclusion
References
3 Capacity utilisation at airports worldwide
3.1 Overview: how to measure capacity utilisation at airports?
3.2 Air traffic ranking curves
3.3 Peak hour volume and annual service volume
3.4 The capacity utilisation index
3.5 Case study: development of capacity utilisation at three example airports: San Diego (SAN), London Heathrow (LHR) and B...
3.6 Conclusion
References
4 Constrained and under-utilised airports
4.1 Overview: point-to-point versus hub networks
4.2 Development of global air traffic distribution over time
4.3 Regional characteristics of air traffic distribution
4.4 How to describe levels of airport capacity constraint
4.5 Capacity constrained airports and under-utilised airports
4.6 Case study: development of capacity utilisation by capacity utilisation class at three example airports: San Diego (SAN...
4.7 Conclusion
References
Part II: Models for assessing mitigation strategies
5 General strategies for mitigating airport capacity constraints
5.1 Overview: typology of mitigation measures
5.2 Investment option: new runways
5.3 Rerouting traffic to under-utilised airports and/or using more off-peak hours
5.4 Raising seat capacity and load factor per flight
5.5 Case study: development of mitigation measures at three example airports: San Diego (SAN), London Heathrow (LHR) and Be...
5.6 Conclusion
References
6 Modelling future air passenger demand
6.1 Background
6.2 Model theory
6.3 Model estimation and testing
6.4 Model application: comparing different forecasts
References
7 Modelling future airport capacity and capacity utilisation
7.1 Background
7.2 Model theory and parameter estimation
7.3 Model application: comparison of model results with actual traffic data
7.4 Conclusion
References
8 Modelling future airport capacity enlargements and limits
8.1 Background
8.2 Model theory and parameter estimation
8.3 What is the potential impact of limited airport capacity on future growth of flight volume?
8.4 Conclusion
References
9 Modelling future development of the average number of passengers per flight
9.1 Background
9.2 Model theory and parameter estimation
9.3 Model application: 2008–16 projection of aircraft size (passengers per flight) on 40 sample routes
9.4 Conclusion
References
Part III: Forecasting future air traffic development up to 2040 and assessing mitigation strategies
10 Traffic forecast and mitigation strategies
10.1 Traffic situation in 2016
10.1.1 Passenger volume
10.1.2 Flight volume
10.1.3 Aircraft size
10.2 Forecast assumptions for 2030 and 2040
10.3 Traffic forecasts for 2016–30
10.3.1 Passenger volume
10.3.2 Flight volume
10.3.3 Aircraft size
10.3.4 Top 20 airports in terms of unaccommodated demand in 2030
10.4 Traffic forecasts for 2030–40
10.4.1 Passenger volume
10.4.2 Flight volume
10.4.3 Aircraft size
10.4.4 Top 20 airports in terms of unaccommodated demand in 2040
10.5 Assessing mitigation strategies
10.5.1 How do increasing aircraft size and adding new runways contribute to airport capacity improvements up to 2030 and 2040?
10.5.1.1 Up to 2030
10.5.1.2 Between 2030 and 2040
10.5.2 General mitigation measures in world regions
10.6 Case study: traffic forecast for San Diego (SAN), London Heathrow (LHR), Beijing Capital City (PEK) and Beijing Daxing...
10.6.1 Passenger volume
10.6.2 Number of aircraft movements
10.6.3 Aircraft size
10.6.4 Capacity analyses
10.7 Conclusion
References
11 Summary and conclusion
Reference
Appendix
List of abbreviations
Definition of world regions
Index
Back Cover

Citation preview

Airport Capacity Constraints and Strategies for Mitigation A Global Perspective

Airport Capacity Constraints and Strategies for Mitigation A Global Perspective

Marc C. Gelhausen German Aerospace Center (DLR), Institute of Air Transport and Airport Research, Cologne, Germany

Peter Berster German Aerospace Center (DLR), Institute of Air Transport and Airport Research, Cologne, Germany

Dieter Wilken German Aerospace Center (DLR), Institute of Air Transport and Airport Research, Cologne, Germany

Academic Press is an imprint of Elsevier 125 London Wall, London EC2Y 5AS, United Kingdom 525 B Street, Suite 1650, San Diego, CA 92101, United States 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom Copyright © 2020 Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN: 978-0-12-812657-8 For Information on all Academic Press publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: Joe Hayton Acquisition Editor: Brian Rommer Editorial Project Manager: Aleksandra Packowska Production Project Manager: Kamesh Ramajogi Cover Designer: Mark Rogers Typeset by MPS Limited, Chennai, India

Contents Foreword

ix

Part I Basic concepts

1

1.

Introduction

3

References

5

Concepts of capacity and methods of estimation

7

2.

3.

2.1 Overview: what do we mean by capacity? 2.2 Main factors influencing capacity 2.3 Approaches to estimate runway capacity 2.3.1 Queuing models 2.3.2 Analytical approaches 2.3.3 Empirical approaches 2.3.4 Simulation models 2.4 Conclusion References

8 12 16 18 20 21 27 27 29

Capacity utilisation at airports worldwide

31

3.1 3.2 3.3 3.4 3.5

31 34 43 48

Overview: how to measure capacity utilisation at airports? Air traffic ranking curves Peak hour volume and annual service volume The capacity utilisation index Case study: development of capacity utilisation at three example airports: San Diego (SAN), London Heathrow (LHR) and Beijing (PEK) 3.6 Conclusion References

55 61 63

v

vi

4.

Contents

Constrained and under-utilised airports

65

4.1 4.2 4.3 4.4 4.5 4.6

68 70 74 87 92

Overview: point-to-point versus hub networks Development of global air traffic distribution over time Regional characteristics of air traffic distribution How to describe levels of airport capacity constraint Capacity constrained airports and under-utilised airports Case study: development of capacity utilisation by capacity utilisation class at three example airports: San Diego (SAN), London Heathrow (LHR) and Beijing (PEK) 4.7 Conclusion References

Part II Models for assessing mitigation strategies 5.

6.

7.

General strategies for mitigating airport capacity constraints

97 101 103

105 107 107 110

5.1 Overview: typology of mitigation measures 5.2 Investment option: new runways 5.3 Rerouting traffic to under-utilised airports and/or using more off-peak hours 5.4 Raising seat capacity and load factor per flight 5.5 Case study: development of mitigation measures at three example airports: San Diego (SAN), London Heathrow (LHR) and Beijing (PEK) 5.6 Conclusion References

126 130 132

Modelling future air passenger demand

133

6.1 Background 6.2 Model theory 6.3 Model estimation and testing 6.4 Model application: comparing different forecasts References

133 135 141 155 159

Modelling future airport capacity and capacity utilisation

161

7.1 Background 7.2 Model theory and parameter estimation 7.3 Model application: comparison of model results with actual traffic data

116 119

162 163 178

Contents

8.

7.4 Conclusion References

187 188

Modelling future airport capacity enlargements and limits

189

8.1 Background 8.2 Model theory and parameter estimation 8.3 What is the potential impact of limited airport capacity on future growth of flight volume? 8.4 Conclusion References

9.

vii

Modelling future development of the average number of passengers per flight 9.1 Background 9.2 Model theory and parameter estimation 9.3 Model application: 2008 16 projection of aircraft size (passengers per flight) on 40 sample routes 9.4 Conclusion References

190 190 197 203 204

205 205 206 215 221 221

Part III Forecasting future air traffic development up to 2040 and assessing mitigation strategies

223

10. Traffic forecast and mitigation strategies

225

10.1 Traffic situation in 2016 10.1.1 Passenger volume 10.1.2 Flight volume 10.1.3 Aircraft size 10.2 Forecast assumptions for 2030 and 2040 10.3 Traffic forecasts for 2016 30 10.3.1 Passenger volume 10.3.2 Flight volume 10.3.3 Aircraft size 10.3.4 Top 20 airports in terms of unaccommodated demand in 2030 10.4 Traffic forecasts for 2030 40 10.4.1 Passenger volume 10.4.2 Flight volume 10.4.3 Aircraft size

225 225 226 227 229 232 232 236 240 240 245 246 251 252

viii

Contents

10.4.4 Top 20 airports in terms of unaccommodated demand in 2040 10.5 Assessing mitigation strategies 10.5.1 How do increasing aircraft size and adding new runways contribute to airport capacity improvements up to 2030 and 2040? 10.5.2 General mitigation measures in world regions 10.6 Case study: traffic forecast for San Diego (SAN), London Heathrow (LHR), Beijing Capital City (PEK) and Beijing Daxing (PKX) 10.6.1 Passenger volume 10.6.2 Number of aircraft movements 10.6.3 Aircraft size 10.6.4 Capacity analyses 10.7 Conclusion References

11. Summary and conclusion Reference Appendix Index

256 259

260 266

282 283 285 287 288 290 293

295 309 311 321

Foreword How did this book originate? In the years following the turn of this century, discussions in working groups within the Committee on Aviation Environmental Protection (CAEP) of the International Civil Aviation Organisation (ICAO) around the broad topic of future development paths of global air transport increasingly focused on the capacity issue of the air traffic system. Information on future demand for air travel and the number of passenger aircraft was readily available through market outlook reports of the aircraft industry and other semi-private and institutional sources; however, questions remained as to the underlying hypotheses of the forecasts. Of particular interest was the question as to what degree are the demand and traffic forecasts influenced by present and future system capacity shortfalls? Do existing airport capacity problems, like those at London Heathrow, hinder airlines from developing their networks in accordance with the demand for air journeys? And, if so, must we expect stronger effects, if capacity problems in the global airport network will propagate with growing demand? The European Organisation for the Safety of Air Navigation (Eurocontrol) took up this topic in the Challenges of Growth project and concluded that depending on scenario assumptions the European network will be unable to accommodate nearly two million flights in the year 2035 due to airport capacity constraints. The institutional interest in the capacity issue concurred with our research interests. We had already commenced analysis of the prevailing airport capacity situation and were developing methodological tools for forecasting demand and traffic under conditions of lack of airport capacity. We were only a stone’s throw away from collating and documenting our material when the publisher Elsevier approached and invited us to document our work in a text book. Now, after having completed this task, we are pleased to provide interested parties in science, institutions, and industry with our research results and recommendations. We would like to thank the German Aerospace Center (DLR) and our Institute of Air Transport and Airport Research for having given us the chance to devote ourselves entirely to writing this book for quite some time. We would like to thank our colleagues who have contributed to our knowledge, in particular about the issue of airport capacity. We hope that the efforts were worthwhile and that the benefits of publishing the contents will

ix

x

Foreword

outweigh the inadequacies of the book; they are entirely our responsibility. Last but not least, we want to thank our families who have had to put up with the disruption to their lives while we have been working on this book.

Part I

Basic concepts Part I of this book describes the fundamental concepts and empirical findings which form the basis for the models presented in Part II and the forecasts for 2030 and 2040 in Part III. The first part starts with a short introduction and then presents different methods of airport capacity estimation. Here, we discuss, among other issues, whether to use the hourly or annual airport capacity of an airport. Key topics of Part I are air traffic ranking curves and the capacity utilisation index (CUI), which are based upon air traffic ranking curves. This part closes with a global analysis of capacity constrained and underutilised airports.

1

Chapter 1

Introduction Part I of this book describes the fundamental concepts and empirical findings which form the basis for the models presented in Part II and the forecasts for 2030 and 2040 in Part III of this book. Part I starts with a short introduction and then presents different methods of airport capacity estimation. Here, we discuss, among other issues, whether to use the hourly or annual airport capacity of an airport. Key topics of Part I are air traffic ranking curves and the capacity utilisation index, which is based upon air traffic ranking curves. Part I closes with a global analysis of capacity constrained and underutilised airports. Global air traffic has grown substantially in the past; the pace of growth has only been interrupted by oil and financial crises, terrorism and wars. The number of passengers transported worldwide by air transportation reached a volume of almost 3796 million in 2016. Since 2000, this volume has more than doubled with an average annual growth rate of 5.3% as Fig. 1.1 illustrates (ICAO, 2017). While we have seen strong growth of air traffic worldwide in the past and can expect a continuation of growth for the long-term future, we have to take a note of the fact that some important airports are faced with capacity constraints so that airlines have problems in planning and scheduling flights in the preferred way. There are many airports with traffic volumes that reach capacity only at certain peak times, for instance, during some morning and evening hours. However, there are also airports with high traffic loadings which experience near capacity utilisation during many hours of the day, such as London Heathrow, Beijing and New York LaGuardia. On the other hand, the majority of airports have low traffic volumes and no capacity problems. The question is whether or not airport capacity constraints will become such a problem in the global air transport network to form a barrier to the future growth of demand. The objective of this book is to present a detailed empirical and model-based analysis of the impact of limited airport capacity on the future development of air traffic. So far, air traffic has continuously grown for the last 67 years, interrupted only by a small number of large global crises as Fig. 1.1 illustrates. After the 9/11 crisis in 2001, which lasted more or less until 2003, the long-term growth accelerated again and arrived at an average growth rate of about 8% in the period 2003 08. This growth was supported by further deregulation Airport Capacity Constraints and Strategies for Mitigation. DOI: https://doi.org/10.1016/B978-0-12-812657-8.00001-4 © 2020 Elsevier Inc. All rights reserved.

3

4

million

PART | I Basic concepts Oil crisis

4500

1971–79 11.5% p.a.

4000

9/11

1998–2000 7.1% p.a.

Financial crisis

2003–07 7.8% p.a.

2009–17 6.4% p.a.

1991–97 5.8% p.a.

3500 Passengers

Gulf war 1980–90 5.8% p.a.

3000 2500 2000 1500 1000

Excl. USSR

Incl. USSR

500

19 1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 2098 2000 2002 2004 2006 2008 2010 2012 2014 16

0

FIGURE 1.1 Development of global air traffic since 1950 [International Civil Aviation Organization (ICAO), 2017]. p.a., per annum.

of air transport markets, especially in Europe, pushing the development of the low-cost carrier segment. Nevertheless, due to the global economic and financial crisis in 2008 09, worldwide air traffic experienced a severe decline again. Thereafter, air traffic regained ground by an increasing global air passenger volume growth rate of 6.4% per year between 2009 and 2017 [International Civil Aviation Organization (ICAO), 2017]. Despite some interruptions, air traffic has followed a stable, long-term growth trend, which has accelerated since the late 1990s; however, capacity constraints at airports are becoming increasingly more important especially at major hubs. Capacity constraints include limited physical infrastructure, such as runways and terminal capacity, as well as administrative restrictions, such as night curfews, noise and emission limits. Limited airport capacity may reduce the negative effects of air transport on the surrounding environment of the airport. However, these constraints reduce the available capacity and, thus, form a barrier to further growth at these airports (Gelhausen et al., 2013). The approach we present in this book is an empirical and model-based one with a rather low level of detail with regard to individual airport characteristics, such as detailed runway system layout and passenger terminal configuration. It is data- and model-driven for two important reasons. First, processing a large number of airports in an acceptable time frame needs some reduction of details of individual airports. Furthermore, available data and its accuracy differ for airports in different world regions. Second, the approach is mainly intended to be employed within long-term forecasts with a horizon of about ten years or longer in large networks or on a global level, which further limits the level of detail that is realisable. Therefore the

Introduction Chapter | 1

5

method presented is not intended to and cannot be a substitute for a more detailed analysis on the single airport level. It rather serves as an instrument to describe the overall level of airport capacity utilisation on a rather highly aggregated level, for example, a long-term forecast of annual aircraft movements and to which degree an airport is able to handle the forecast traffic volume.

References Gelhausen, M.C., Berster, P., Wilken, D., 2013. Do airport capacity constraints have a serious impact on the future development of air traffic? J. Air Transp. Manage. 28, 3 13. International Civil Aviation Organization (ICAO), 2017. ICAO Traffic Statistics. ICAO, Montreal.

Chapter 2

Concepts of capacity and methods of estimation Airports represent the access points of air transport users to the global air traffic network. As part of the public air transport system, with scheduled airline services for passengers and goods, airports have become complex multifunctional entities with both airside and landside facilities. The airside comprises the manoeuvring area for aircraft on the runway and taxiway system, parking areas either at the terminal or in reserved ramp areas and infrastructure for other uses such as vehicle parking and cargo storage. The main landside components are the terminal buildings for passengers and cargo, parking areas for road vehicles and public transportation facilities such as rail terminals. There are typically also airport facilities such as hangars, fuel farms, power plants and other support facilities, and non-aeronautical land use such as for hotels, office buildings and car rental locations, which are common components of modern airports. However, in many cases, these are less critical with respect to capacity planning. In addition the airspace surrounding the airport is also regarded as being part of the airport’s airside. Given these diverse capacity relevant components of airports, it becomes evident that the term airport capacity often does not refer to the airport as a whole but most likely to one of the functional components. Therefore airport capacity could mean runway or runway system capacity, terminal capacity, apron/stand capacity, airspace capacity and surface access capacity. Airport capacity is often used to describe the capacity of the component with the lowest capacity of all airport components. In major airports, this is often the runway system. In a situation of growing demand for air transport services, as has been the case for many years, airport operators normally have the means to adapt ground traffic infrastructure, terminals and parts of the airside facilities to the growing demand as part of their continuous planning and enhancement process, because no lengthy plan approval procedure is required. However, increasing the runway capacity more than marginally means adding a new runway or even building a new airport, for which substantial land surface is needed. In Western states, this requires a planning approval procedure, which involves the administration, politics and the public. Getting the approval of Airport Capacity Constraints and Strategies for Mitigation. DOI: https://doi.org/10.1016/B978-0-12-812657-8.00002-6 © 2020 Elsevier Inc. All rights reserved.

7

8

PART | I Basic concepts

public authorities for building new runways has been almost an insurmountable task for many airports as the neighbouring population may be affected negatively by airport noise and gaseous emissions and, therefore, oppose plans to enhance airport capacity. Environmental factors are most likely the single biggest reason why communities object to the development and expansion of airports. London Heathrow is a predominant example of a major airport in great need of a new runway. However, it has failed for many years to obtain a decision from the British government to pursue this plan. Consequently, the capacity of the runway system of major airports can be typically regarded as the essential element of determining the overall capacity of the airport.

2.1

Overview: what do we mean by capacity?

Regardless of the term airport or runway capacity, we have to first describe what we mean by capacity. The term capacity refers to the capability of a facility to handle people, freight, vehicles, etc. Capacity is often regarded as the maximum number of traffic units, that is vehicles, that can pass through a traffic facility within a given time span under specified conditions regarding safety regulations, operating conditions, standards of expediency and comfort and possibly other conditions (Urbatzka and Wilken, 1997). In order to reach the highest number of traffic units possible, we have to assume that there is a continuous demand for using the traffic facility in such a time interval. This means that the time span selected must not be too long, for instance, not including night hours, since the demand typically drops during the night and only exceptionally reaches peak time levels. Nevertheless, in strategic airport planning, for example within BAA (Matthews, 1995), annual capacities are often used as an indicator of the ability of the runway to accommodate a certain service volume, rather than as a true measure of the maximum throughput capacity. In comparison with a forecast traffic volume, annual capacities serve to estimate the future capacity reserve. Especially in long-term airport planning, this capacity reserve estimate draws upon the annual capacity since the time unit of the demand and aircraft movement forecast is often the year. Complex procedures exist to calculate the annual capacity. In principle the capacity is deducted from hourly capacities multiplied by factors describing the duration of runway utilisation within a year. The Federal Aviation Administration (FAA) of the US Department of Transport has developed the concept of an annual service volume (ASV) for medium- and long-term infrastructure planning. In 1983 it proposed a method of estimating the ASV, published in the FAA Advisory Circular 150/ 5060-5 [Federal Aviation Administration (FAA), 1983] and referred to as the ‘handbook method’. The ASV is defined in the circular ‘as a reasonable

Concepts of capacity and methods of estimation Chapter | 2

9

estimate of an airport’s annual capacity, and accounts for differences in runway use configuration, aircraft mix, weather conditions that will be encountered over a year’s time’. The estimate of ASV is based on an average weighted hourly capacity which depends on different classes of runway use configurations, variations in traffic mix and the distinction of flights under instrument flight rules (IFR) and visual flight rules (VFR). ASV is then calculated as the product of the average weighted hourly capacity, a ratio of annual demand to average daily demand in peak month and the ratio of average daily demand to the average peak hour demand in peak month. The handbook gives values of the ASV for a single runway ranging from 195,000 aircraft movements per year to 240,000 movements, depending on the above-stated conditions. For detailed planning and dimensioning of infrastructure facilities the hourly capacity is normally retained, since the significance of the annual capacity regarding the throughput capability is lower because of seasonal and daily variations of traffic, which the annual capacity has to account for by applying reduction factors. Night hours, Sundays, some holidays and other off-peak periods are typically times of low traffic demand, which are therefore not well suited for inclusion in the time span which serves as a base for capacity calculation. As already stated, the time unit of measuring throughput capacity should be defined in such a way as to allow for a continuous utilisation of the runway by the demand for aircraft movements. In practical terms, that means a period of not more than one hour or two. In slot coordination the basic time unit of the so-called declared capacity (see Section 2.3.3) is typically an interval of ten minutes. If annual capacities are retained, the assumptions regarding the varying levels of service at peak and non-peak hours and the temporal variation of demand should clearly be stated; since these factors vary from airport to airport, annual capacities are not directly comparable between airports. Since the runway system is crucial for the capability of the airport to handle current and future traffic, the throughput of this airport component is of fundamental importance in airport planning. A great amount of research has been devoted to the subject of estimating runway capacity, especially in the United States. The FAA has initiated and carried out many capacity-related studies and issued Advisory Circulars on this subject. Three prominent examples of textbooks describing methodological approaches to estimating runway capacity are ‘Airport Systems Planning, Design, and Management’ by de Neufville and Odoni (2013), ‘Planning & Design of Airports’ by Horonjeff et al. (2010) and ‘Airport Engineering Planning, Design, and Development of 21st Century Airports’ by Ashford et al. (2011). Readers interested in the methodological detail of capacity derivation are advised to consult these sources. This book will rather concentrate on applying capacities in questions of capacity utilisation and constraints.

10

PART | I Basic concepts

Basically, we can distinguish between two concepts of capacity: the one aiming to be a measure of the maximum number of aircraft being served on a runway without any regard to the level of service, that is a measure of aircraft delays in the take-off or landing phase, and the other reflecting a specified level of service. The first concept mirrors a so-called theoretical, ultimate or saturation capacity, which indicates the maximum number of aircraft movements within one hour regardless of the delay single aircraft may encounter when they are ready for take-off or landing in conditions of continuous demand. As demand approaches ultimate capacity, delays of aircraft are likely to reach intolerably high levels. The phenomenon of sharply increasing delays in traffic situations approaching capacity without the possibility of reducing them after a short while can be observed in road and rail traffic as well. To account for the delay problem the second concept of practical or sustained capacity has found wide application. Movement rates are determined in relation to average delay levels [Federal Aviation Administration (FAA), 1969]. A practical capacity was devised primarily for planning purposes, whereby a tolerable average delay per aircraft movement was the criterion for setting the capacity as a limit to the number of movements per hour for the runway system under day-to-day operating conditions. In contrast to the concept of saturation capacity the concept of practical or sustained capacity incorporates, therefore, the quantitative measure (e.g. aircraft movements per hour) with a qualitative criterion. As early as the 1960s, the FAA introduced the ‘practical hourly capacity’ (PHCAP) as a planning value of capacity. It is the number of aircraft movements that can be handled in one hour on a runway with an average delay of four minutes. For many airports the practical capacity is still seen to be reached if the average delay of aircraft in the take-off queue will not exceed four minutes. At this average delay, single aircraft delays, which may exceed the average substantially, are not likely to increase further but rather stabilise or eventually diminish. In addition, single aircraft delays may exceed the average; however, they are not likely to violate the punctuality criterion of 15-minute delay a great deal. Only if delays exceed 15 minutes are the corresponding flights recorded as delayed flights in traffic statistics. In general the PHCAP reaches movement rates which are in the range of 80% 90% of the level of the theoretical concept of saturation capacity, depending on the local conditions at hand (de Neufville and Odoni, 2013). The PHCAP is an essential parameter for short-term, medium-term as well as for long-term planning, such as assigning air traffic control (ATC) slots in short-term actions, slot coordination and resource allocation in medium-term planning and planning of infrastructure in strategic planning. It may well be that for each of these planning options, different concepts and values of capacity may be adequate. In short-term actions, specific capacity values may be retrieved that relate to foreseeable weather conditions, traffic

11

Concepts of capacity and methods of estimation Chapter | 2

mix structure and ATC-specific conditions such as VFR or IFR. In mediumterm scheduling, for instance, the declared capacity is retained for coordinated airports (see Section 2.3.3). In long-term infrastructure planning a capacity should be chosen which reflects a weighted average of prevailing weather and traffic situations. Capacity coverage charts (CCCs), which show the capacity values of all hours of the year in relation to their relative occurrence, are a good source of selecting appropriate capacity values for different planning tasks. In Fig. 2.1 a generic CCC is shown with different capacity levels for different operational configurations of runway use in relation to the time (as part of the total time) of the respective runway use. A necessary prerequisite for drawing a CCC is that the traffic volumes of an airport are reaching the capacity at least in some peak hours. A common hypothesis of a CCC is that the ratio of arrivals to departures in one hour is about one, meaning about 50% arrivals and 50% departures, under prevailing weather conditions. This condition is not typical in peak hours at hub airports, which are characterised by arrival and departure banks. This again implies that capacities in certain arrival and departure mix situations may differ from those shown in the CCC. In addition a traffic mix other than the ‘typical’ one will have an effect on the capacity. In the case of a single runway airport, there are just two configurations of runway use: one with operations in a situation with winds coming from one side and the other configuration with operations with winds coming from the opposite direction. As winds more often come from one direction, say from 40

35

Configuration 1

Capacity (flights per hour)

30

Configuration 2

25

20

15 Visual meteorological conditions

10

Instrument meteorological conditions CAT II/III

5

0

0

10

20

30

40 50 60 Time (% of total time)

FIGURE 2.1 A generic capacity coverage chart of an airport.

70

80

90

100

12

PART | I Basic concepts

the westerly direction, than from other directions, one would account for this situation and yield a capacity value according to the prevailing condition, that is wind direction more than 50% of the time.

2.2

Main factors influencing capacity

It has been stated that the capacity of a runway is the maximum number of aircraft that can use the runway within an hour or any other short time interval, such as 10 or 30 minutes. For estimating the maximum number of aircraft, it is essential to know that the capacity is not a constant value but rather a random variable depending on a number of influencing factors. Since these factors vary among airports, each runway has a specific capacity. The most relevant factors are (their effects are described in great detail in de Neufville and Odoni, 2013) G G G G G G

G G

G

number and geometric layout of the runways, separation requirements between aircraft imposed by ATC, visibility, cloud ceiling and precipitation, wind direction and strength, mix of aircraft using the airport, mix of movements (arrivals and departures) on each runway and sequencing of movements, type and location of taxiway exits from the runway(s), state and performance of the ATC system (including the number and shape of standard instrument departures and standard arrivals routes), and noise-related and other administrative regulations regarding the use of runways and surrounding airspace.

Obviously, the capacity of the runway system is directly related to the number and the layout of runways. While globally the vast majority of airports have only one runway, important airports such as national gateways and hubs are equipped with more runways. Airports with two parallel runways, which are separated by more than about 1500 m, are normally operated independently from each other (dependent on the runway instrumentation and ATC rules) and have thus a capacity of two runways. The capacity of airports with two parallel runways closer together is lower, since aircraft movements on each of the runways are not independent. It is necessary to maintain separation minima, which depends on the distance between runways. Likewise, runways that intersect or converge have a capacity higher than that of a single runway but lower than that of two runways. The airspace surrounding airports, also called terminal area airspace or terminal control airspace, is of direct relevance to the capacity of airports, because aircraft approaching the airport or taking off from that airport have to follow specific procedures and separation rules in order to ensure the

Concepts of capacity and methods of estimation Chapter | 2

13

safety of all airspace users. The longitudinal separation requirements determine the maximum number of aircraft that can be handled on a runway within a certain time span. One of the main reasons for specifying separation minima is the fact that jet turbines of flying aircraft create wake vortices which endanger trailing aircraft when following too close to the leading aircraft. Since wake vortices depend on the power and size of the engines, the size of the wings and the size and weight of the aircraft, each aircraft is assigned to one of three classes. The separation minima are specified for aircraft approaching an airport in units of distance (km, nautical miles) and for aircraft departing from an airport both in units of time (minutes) and distance. The US FAA has issued such separation standards for the US airspace, while the International Civil Aviation Organisation (ICAO) (1984) has developed similar standards applicable in member states. As an example, the separation specifications of ICAO for a single runway use are briefly described in the following. According to ICAO, the maximum take-off mass determines the wake turbulence category H (heavy), M (medium) and L (light): G G G

Class Heavy: all aircraft types of 136 t or more Class Medium: aircraft types less than 136 t but more than 7 t Class Light: aircraft types of 7 t and less

The minimum separation between departing aircraft is one minute, if aircraft are to fly on tracks diverging by at least 45 degrees immediately after take-off so that lateral separation is provided. For both arriving and departing aircraft that are not radar-separated, time-based wake turbulence separation minima have been specified. The following minima are applied for aircraft landing behind a heavy or medium aircraft: G G

Medium aircraft behind a heavy aircraft: two minutes Light aircraft behind a heavy or medium aircraft: three minutes

For departing aircraft that are not radar-separated a minimum separation of two minutes shall be applied between a light or medium aircraft taking off behind a heavy aircraft or light aircraft taking off behind a medium aircraft. In air traffic service surveillance systems, arriving aircraft, in particular, are radar controlled, and distance-based wake turbulence separation minima are applied. For both the approach and departure phases of flight the separation minima are specified in Table 2.1. Assuming a population of medium aircraft such as the B737 or A320 taking off following each other with a minimum separation time of 60 seconds at a single runway airport the theoretical capacity of the runway would be 60 departures per hour. The corresponding capacity of aircraft in the final approach would be lower due to the greater wake turbulence separation. In practice runway capacity is lower than the theoretical capacity, since aircraft are, for safety reasons, not typically spaced with exact minimum time

14

PART | I Basic concepts

TABLE 2.1 Separation minima in air traffic service surveillance systems [International Civil Aviation Organisation (ICAO), 1984]. Preceding aircraft

Succeeding aircraft

Distance-based wake turbulence separation minima

Heavy

Heavy

7.4 km (4.0 NM)

Heavy

Medium

9.3 km (5.0 NM)

Heavy

Light

11.1 km (6.0 NM)

Medium

Light

9.3 km (5.0 NM)

intervals, but with additional buffer times to cope with uncertainties in aircraft spacing procedures. Weather conditions directly affect runway capacity. In particular, visibility and cloud ceiling are the two characteristics which determine the two weather categories with instrument flight rules (IFR) under instrument meteorological conditions and correspondingly visual flight rules (VFR) under visual meteorological conditions. For VFR flights, weather conditions must be such that the visibility of pilots exceeds 8 km and cloud ceiling is higher than about 750 m; IFR flights are operated in weather conditions inferior to those under VFR. In Europe the standard weather category for air traffic management (ATM) is IFR, while in the United States, VFR conditions govern. In VFR weather pilots are asked by ATC while approaching an airport to visually maintain a separation from the preceding aircraft, while in IFR, the ATC is responsible for maintaining longitudinal separation corresponding to those as listed in Table 2.1. The latter is typically higher than the VFR distance so that the capacity of airports in the United States is greater than in Europe for most of the time. Finally, precipitation such as heavy rain or snowfall may reduce the runway capacity due to poorer braking capability of aircraft. In severe winter weather conditions with icy runways or heavy snowstorms runways may even be closed to avoid the risk of unsafe landings of aircraft. Winds may affect the capacity if the crosswind component exceeds a certain strength. The International Civil Aviation Organisation (ICAO) (2009) specifies the maximum crosswind at about 20 kt for aircraft requiring runway lengths of more than 1500 m. Tailwinds must also not exceed certain limits such as about 5 kt. However, since take-offs and landings are always planned to be performed against the wind, conditions of tailwinds during take-off or landing are rather rare occasions. Airports which are equipped with several runways allowing aircraft to take off and land in various directions are better protected against crosswinds than airports with parallel runways where

Concepts of capacity and methods of estimation Chapter | 2

15

aircraft can operate only in one direction. However, bigger and heavier aircraft using an airport, such as wide-body aircraft, are more resistant to crosswinds. As we have seen, the longitudinal separation minima of aircraft approaching an airport depend on their size and weight; therefore, the mix of aircraft in both the landing and take-off queue affects the runway capacity. The more non-homogeneous the population of aircraft, the lower is the capacity. Typically, most aircraft operating at an international airport are medium sized, thus the more important the airport, the more wide-body aircraft are likely to use it. However, at smaller airports, aircraft of the light category are more numerous. In order to reduce the negative effect of different separation minima, ATM tries to sequence aircraft both by weight category and by taking off and landing. To a certain degree, aircraft approaching the airport can be sequenced in such a way as to form groups of aircraft of equal weight category and of aircraft in the process of taking off. Another means of simplifying operations in near-capacity conditions at airports with several runways may be to segregate arriving and departing aircraft by runway and by weight class as far as possible. A slight disadvantage of the segregated mode of using two or more runways is a decline in the throughput of aircraft movements per time unit as compared to a mixed use of a runway by both departing and landing aircraft. Here, the runway used for landings has a lower capacity and is thus underutilised as compared to the runway with departures only. Thus operations with arrival-departure sequences maximise runway utilisation but increase the workload of ATM. Infrastructural characteristics of runways such as the number, type and location of exit ramps have an impact on runway capacity, since the time an aircraft needs for landing when the runway is blocked for other aircraft depends on whether high-speed exits are available. A longer runway may have a positive effect on the capacity if a landing aircraft uses the first part of a runway while an aircraft is waiting at an entrance ramp in the middle of the runway to take off. This, thus, reduces the part of the occupancy time which prevents the waiting aircraft beginning its take-off. In addition, obstacles such as hilly terrain or mountains located in a runway’s direction may reduce the utilisation of the runway by limiting the number of alternative departure routes. Air navigation service providers (ANSP) are responsible for ATC and ATM. For example in Europe, the supranational ATC organisation Eurocontrol and national ANSP organisations are responsible for setting standards and issuing rules and procedures as well as carrying out the ATC task of navigation, surveillance and communication. Technological advancement in most countries has led to high standards of automation systems (both software and hardware of computers and display equipment). However, the ATM service continues to depend on the good functioning of a complex man machine interaction. Skilled air traffic controllers determine to a great

16

PART | I Basic concepts

extent the service quality, and one can probably state that the high level of air traffic safety has been achieved by means of a sophisticated system of ATM. The influence of ATM on airspace and airport capacity depends both on the performance of air traffic controllers and their technical support and on international institutional improvements of ATM regulation, in Europe more than in the United States. We can assume that the progress of improving airspace and airport capacity and air traffic efficiency due to advances in international ATM will be rather slow in time. Finally, local rules regarding the emission of noise and exhaust fumes may influence the aircraft mix and number, in particular during the nighttime. In Europe, there are many airports with night curfews on air traffic which partly or completely ban aircraft movements during specified night hours, so that the capacity may be reduced to zero. Since the demand at night seldom reaches levels as high as during daytime hours, the effect of night curfews on capacity utilisation is limited. The traffic may be hindered more by night curfews in early morning and late evening hours when the demand is still high but capacity is already sharply reduced. Airports with hub operations may therefore be more affected by night curfews than airports with mainly origin-destination demand.

2.3

Approaches to estimate runway capacity

The objective of this book is not to deal with the derivation of methods of estimating runway capacity but rather to apply capacities in the context of capacity utilisation in order to demonstrate the problem of capacity constraints of airports and show ways to mitigate the capacity crunch. Nevertheless, it is useful to look back and briefly report on the various approaches that have been developed over time. Thereby, we describe functional relationships between capacity and influencing factors and present some results of capacity estimates of a sample of airports worldwide. A number of mathematical and empirical approaches have been developed over the years with the objective to provide a tool that allows the determination of runway capacity for various conditions and composition of air traffic. As early as 1959, when air traffic as a public transport mode played only a minor role in many parts of the world, Blumstein (1959) developed a functional model for estimating the landing capacity of a single runway in relation to factors such as the length of the common final approach path, the speed of aircraft following each other on final approach, the runway occupancy time of aircraft landing and the minimum separation required by ATC of aircraft on final approach. The model then estimates the landing capacity by calculating minimum possible time intervals between successive aircraft arrivals and converting that outcome into the maximum possible number of arrivals per hour. The same approach can be extended to runways used exclusively for take-offs to estimate the departure capacity.

17

Concepts of capacity and methods of estimation Chapter | 2

Based on the Blumstein formulations, computer-based mathematical models have been developed, taking the former constant values of approach speeds, runway occupancy times and minimum separations of successive aircraft as random variables (see Harris, 1972 and Odoni, 1972) and thereby extending the applicability of the models. Combining all possible runway utilisation situations of exclusive arrivals and departures and mixed operations of arrivals and departures yields a so-called capacity envelope (Gilbo, 1993). The envelope denotes a boundary line below which observed and/or calculated capacity values of each utilisation case of possible departure and arrival mixes are located in a diagram with the number of arrivals on the horizontal axis and the number of departures on the vertical axis. Capacity values are found by the sum of arrivals and departures of each point on the envelope. As an example, the capacity envelope for Munich airport is shown in Fig. 2.2. Each point on the envelope represents the combined departure and arrival capacity per hour for the parallel runway system of Munich airport in the year 2016; the departure capacity is thereby around 60 take-offs, while the arrival capacity of 58 landings is about two movements lower than the departure capacity. The departure capacity can be maintained up to a volume of about 38 landings. If more landings occur, the maximum number of departures decreases to about 34 take-offs, when the arrival capacity is reached. Nevertheless, it is not the intention of this book to describe and discuss the various approaches of estimating runway capacity in detail (see, for example de Neufville and Odoni, 2013) but rather to give a short overview

Note: Scheduled arrivals and departures from 6 a.m. to 23 p.m.

70 60

Departures

50 40

≤5 5–10 10–15

30

15–20 >20

20 10 0

0

10

20

30 Arrivals

40

50

60

FIGURE 2.2 Capacity envelope of Munich airport 2016 [Official Airline Guide (OAG), 2016].

18

PART | I Basic concepts

of the different methodological approaches. Basically, there are five lines of thought that have developed over time (Ashford et al., 2011): G G G G G

Queuing models Analytical approaches Empirical approaches Benchmarking Simulation models

2.3.1

Queuing models

Capacity models, relying on queuing theory, were developed in the 1960s when capacity problems were encountered only at a few busy airports, primarily in the United States. Queuing theory deals with the relationship between demand for using a runway and congestion caused by queues of aircraft waiting to be served for landing or taking off. The variable selected for best describing the intensity of congestion and at the same time the (lack of) quality of system performance was the mean delay of aircraft in the arrival or departure queue. A basic hypothesis of the models was that aircraft arrived randomly and the demand process for arrivals and departures was characterised by a Poisson-type distribution with a first come, first served priority. This implies that for the model specification, only the mean and the variance of the service time have to be known. Models relying on queuing theory are described and discussed in detail, for instance, in Ashford et al. (2011) and Horonjeff et al. (2010). Although they have been extensively applied in the United States, they have been more or less replaced in recent years by other approaches such as simulation techniques and benchmarking. One reason is that they only account for some of the many factors influencing capacity. Another reason is that the hypothesis of Poisson-type distributed arrivals and departures has become less and less stringent today. When the queuing approaches were developed in the 1960s, the relationship between demand for runway utilisation (aircraft in the arrival and departure queue) and the average delay per aircraft could be described by a function which is depicted in Fig. 2.3. At rather low levels of demand, equivalent to a low ratio of volume to capacity, average aircraft delay is almost negligible, and the delay grows very slowly with demand until the number of aircraft per hour comes closer to the runway capacity. When flight volumes reach values near capacity, the delay grows much faster and overproportionally in relation to further increasing demand, until at least in theory average delay grows infinitely with one additional aircraft when the total number of aircraft per hour reaches the saturation capacity. To avoid such delays, the US FAA and other institutions responsible for airport planning have introduced a practical capacity which is associated with a certain average aircraft delay, being somewhat lower than

Concepts of capacity and methods of estimation Chapter | 2

19

120

100

Saturation capacity

Average delay (min)

80

60

40

20

4 min

0 0

20

40

60 PH cap

80

100

120

Traffic (flights per hour)

FIGURE 2.3 General relationship between average aircraft delay per hour and hourly demand for using a runway.

the saturation capacity, which is not related to any value of delay. As has been mentioned, the FAA has established the four-minute-delay capacity called PHCAP, which has been derived from graphs like that in Fig. 2.3 and which has been widely used in airport infrastructure planning. Traffic volumes have grown substantially since the time when the queuing approaches were developed. In the time span of 45 years between 1971 and 2016, global air traffic has increased by a factor of 13, that is from 0.5 trillion passenger kilometres (or revenue passenger kilometres) to about 6.5 trillion passenger kilometres. Flight schedules of today’s traffic levels have become more and more uniform and less peaky over time due to slot coordination (see Section 2.3.3 on declared capacity) being practised in many countries except in the United States. The main objective of these measures is to avoid high levels of delay at airports. However, since airport infrastructure could not keep up with the traffic growth, delays are still a daily problem of high-volume and hub airports, but under normal conditions with delays less pronounced than they would be under conditions without slot coordination. Therefore peak demand associated with peak delays, as shown in Fig. 2.3, is typically avoided today by determining practical capacities called ‘declared capacity’, which are used as slot capacity in planning future flight schedules and allocating flights to free slots by slot coordination. The delay-demand function of many airports thus no longer extends to the far right part of the function but remains in the segment of lower than peak delays and volumes.

20

PART | I Basic concepts

2.3.2

Analytical approaches

Applications and further developments of the queuing model have led to mathematical functions, in which minimum separations between aircraft in landing and departing sequences were calculated based on extensive analyses of time intervals of the landing and departing process. The basic type is a function for landings which depends on three variables: the length of the common approach path, aircraft speeds in the approach zone and the minimum separation of two aircraft following each other in approaching the runway (for a detailed description see, for example Ashford et al., 2011). The landing case can be subdivided into an overtaking case, with the speed of the trailing aircraft equal to or greater than the speed of the leading aircraft, and an opening case, in which the speed of the leading aircraft exceeds the speed of the trailing aircraft. In 1972 Harris developed a model for each of these cases, in which the minimum time separation for an aircraft following another aircraft depended on the minimum safety separation distance, the speed of each of the two aircraft and the length of the common approach path. For estimating the runway capacity, all aircraft being served within one hour were grouped into speed classes, and for all aircraft pairs, the minimum intervals were calculated. With the occurrence of aircraft pairs for each speed class being known, an expected weighted minimum separation interval could be approximated by the sum of the single minimum separations, multiplied by the percentage of the occurrence. The runway capacity is then given by the inverse of the weighted minimum separation interval. The basic landing case model has been improved in subsequent years to take into account more complex traffic situations and, in particular, both arrival and departure cases on runways used exclusively for arrivals or departures and for mixed operations. Based on these model improvements, a capacity handbook was published by Federal Aviation Administration (FAA) (1976). The main feature of the handbook is a great number of graphs by which the runway capacity can be calculated for a series of runway configurations by taking into account variables such as aircraft mix, landing and take-off utilisation of runways, geometric layout of runways, and VFR or IFR operating conditions. For preliminary capacity analysis the FAA has also published tables in various documents with capacity values for different runway configurations, different cases of aircraft mix and for VFR and IFR conditions [see, for example Federal Aviation Administration (FAA), 1983]. As an example, the capacity of a single runway is stated to vary in relation to the aircraft mix between 51 and 98 aircraft movements per hour in VFR conditions and between 50 and 59 movements in IFR conditions. In comparison with more recent estimates of runway capacity, these values seem to be somewhat optimistic, although realistic. As we will see in the following chapter, benchmark studies of the FAA have resulted in capacity values for San Diego airport, the only single runway airport among the 30 core airports

Concepts of capacity and methods of estimation Chapter | 2

21

of the United States, of up to 57 aircraft movements per hour under VFR conditions and of 48 movements under IFR conditions.

2.3.3

Empirical approaches

Empirical analyses comprise a wide variety of individual approaches; they have in common some kind of both qualitative and quantitative analyses, often performed within a group of experts. They are based on surveys, observations and statistical and other data which serve as inputs for developing analytical and/or simulation models. Because of their relevance for airport capacity, and accessibility studies and airport planning, in many parts of the world, we concentrate on two main approaches, which are G G

benchmark studies, and consensus approach, in particular, for estimating ‘declared capacity’.

While the Federal Aviation Administration (FAA) (2004, 2014) has developed and applied queuing models and analytical approaches for a long time, in recent years, benchmark studies have increasingly served as a tool for estimating airport capacity profiles for the 30 core US airports. Capacity benchmarks have been defined as the maximum number of flights (arrivals and departures in balanced operations) an airport can routinely handle in an hour, for the runway configuration that provides the highest sustainable throughput during periods of strong demand, in each specified weather condition. They serve as a reference point on the state of capacity of these principal airports. However, as the FAA states, they are not a substitute for more detailed analyses that should precede major investment and policy decisions. The method used by the Federal Aviation Administration (FAA) (2014) incorporated both qualitative and quantitative approaches. A first estimate (called Facility Reported Rate) was given by the ATC team at each airport; the estimate was primarily based on the comprehensive experience of the control tower personnel. This rate has been used by ATC in traffic flow and metering initiatives. The FAA applied, in parallel, a simulation model (runwaySimulator) developed by the MITRE Corporation to simulate the airport’s operations in near-capacity situations. As the FAA report Airport Capacity Profiles [Federal Aviation Administration (FAA), 2014] states, the simulation model was used to estimate the capacity because it offered significant improvements over the analytical tool that was used for previous benchmark studies. In addition the ATC estimates and model results were supplemented by analyses based on operational data the information of which yielded preliminary capacity rates, which were then confirmed or adjusted by ATC facility personnel. As weather has a great influence on the throughput of an airport’s facility, capacity rates were derived for three weather categories: G

Visual: ceiling and visibility allow for visual approaches of aircraft, specific to each airport.

22 G

G

PART | I Basic concepts

Marginal: ceiling and visibility are below visual approach minima, but better than instrument conditions. Instrument: IFR conditions apply, ceiling with less than 1000 ft. or visibility less than three statute miles.

Table 2.2 shows the capacity rates for operations at the 30 core US airports in the year 2014. The rates of each airport and weather condition are presented as a range between the hourly rates reported by ATC and those estimated by the simulation model. As can be seen, there is a wide range of capacity values, depending on the number and configuration of runways. San Diego airport has the lowest hourly capacity with 48 aircraft movements in IFR conditions and up to 57 movements in visual weather conditions. These capacity values mirror the throughput of a single runway; indeed, San Diego is the only airport among the 30 core airports with only one runway in operation. Many of the core airports have significantly higher throughput values, ranging up to almost 300 movements (in VFR conditions) at Denver airport. These airports provide runway systems consisting of several parallel, intersecting or crossing runways, and hence capacities representing a multiple of single runway airports. Another example of empirical analysis is the consensus approach, where a group of experts and/or interested people analyse the traffic volume of an airport in near-capacity situations and discuss the potential size of traffic assuming a utilisation level at the capacity limit. Typically, there is no standard procedure for estimating this value, various qualitative and quantitative approaches relying more or less on empirical studies and data may be followed. Personal experience of traffic and delay under different situations of traffic structure and size play an important role in the consensus approach. An often practised example of the consensus approach is the estimation of the declared capacity in ‘coordination committees’ at slot coordinated airports. The distribution of air traffic in the global network of around 4054 airports is characterised by a high degree of concentration on a rather limited number of important airports, while, on the other hand, the great majority of airports handle relatively low traffic volumes well below capacity limits (Gelhausen et al., 2013). The biggest 120 airports, corresponding to just 3% of all airports, handled half of the total traffic volume of about 35.5 million flights in 2016, while the other 3934 airports handled the same amount of traffic, on average each one handling 4500 flights per year. Many of the topranked airports are identical to those which have capacity problems during peak hours or over longer time periods, as will be shown later in the book (Chapter 4: Constrained and under-utilised airports). In order to manage capacity at constrained airports an administrative type of schedule or slot coordination has existed in European countries for over 40 years and has been applied in EU member states as an official procedure since 1993, following the rules set forth in the European Council Regulation

Concepts of capacity and methods of estimation Chapter | 2

23

TABLE 2.2 Hourly capacity rate ranges for operations at the 30 core airports of the United States [Federal Aviation Administration (FAA), 2014]. Airport identifier and name

ATL

Atlanta Hartsfield Jackson

Aircraft operations (arrivals and departures) per hour Visual

Marginal

Instrument

216 226 (AP)

201 208 (AP)

175 190 (AP)

206 (DP)

166 169 (LIMC AP)

219 222 (DP)

183 186 (DP)

168 179 (LIMC DP) BOS

Boston Logan

116 125

109 112

84 86

BWI

Baltimore Washington

68 80

64 80

62 64

CLT

Charlotte Douglas

176 182

161 162

138 147

DCA

Ronald Reagan Washington

69 72

69 72

54 64

DEN

Denver

262 266 (AP)

224 279

224 243

266 298 (DP) DFW

Dallas/Fort Worth

226 264

194 245

170

DTW

Detroit Metropolitan Wayne County

178 184

163 164

136

EWR

Newark Liberty

94 99 (AP)

76 84

68 70

66 72

56 66

94 100 (DP) FLL

Fort Lauderdale Hollywood

74 82

HNL

Honolulu

117 120

91 105

60 77

IAD

Washington Dulles

150 159 (AP)

112 120 (AP)

108 111 (AP)

156 164 (DP)

136 145 (DP)

125 132 (DP)

IAH

Houston George Bush

172 199

152 180

144 151

JFK

New York John F. Kennedy

84 87 (AP)

85 86

74 84

90 93 (DP)

LAS

Las Vegas McCarran

122 128

106 111

78 83

LAX

Los Angeles

167 176

147 153

133 143 (Continued )

24

PART | I Basic concepts

TABLE 2.2 (Continued) Airport identifier and name

Aircraft operations (arrivals and departures) per hour Visual

Marginal

Instrument

LGA

New York LaGuardia

80 86

76 77

74 76

MCO

Orlando

160 171

148 161

144

MDW

Chicago Midway

64 84

64 74

52 70

MEM

Memphis

144 160

133 150

111 134

MIA

Miami

132 150

132 148

100 104

MSP

Minneapolis Saint Paul

156 167

142 151

114 141

ORD

Chicago O’Hare

214 225

194 200

168 178

PHL

Philadelphia

120 126

94 96

84 88

PHX

Phoenix Sky Harbor

138 145

108 109

96 101

SAN

San Diego

48 57

48 52

48

SEA

Seattle Tacoma

100 112

86 100

76 78

SFO

San Francisco

100 110

90 93

70 72

SLC

Salt Lake City

148 150

138 140

114 120

TPA

Tampa

113 150

95 115

90 95

AP, arrival priority configuration; DP, departure priority configuration; LIMC, low instrument.

No 95/93 (EU, 2004). The European rules have been based on the IATA Worldwide Slot Guidelines (IATA, 2017), which form a procedural framework of slot coordination in many countries of the world, although not in the United States. In 2017 there were around 325 airports worldwide which were partly or fully slot coordinated by means of the IATA coordination framework, of which 180 airports in the European Economic Area plus Switzerland were coordinated on the basis of the EU Regulation No 95/93. Nearly 90 slot coordinating institutions in 70 countries were responsible for carrying out the global slot allocation of flight schedules of around 250 airlines, including those of the United States. The basic principle of the EU/IATA slot coordination is not a marketbased idea of getting access to scarce airport capacity but an administrative right for an airline to use a slot, that is the airport infrastructure at a given point in time for a take-off or landing in the next season, if the slot has been used in at least 80% of all possible cases in the previous equivalent season (‘use-it-or-lose-it’ rule). This so-called grandfather right is the essential

Concepts of capacity and methods of estimation Chapter | 2

25

element of the slot coordination process and guarantees free market access to incumbent airlines at least in the same way as in the previous season, whereas new entrants have only limited chances, depending on the scarcity of slots. The maximum number of slots that can be allocated by the coordinating institution, commonly called the slot coordinator, is equal with the declared capacity, which is determined twice per year by an administrative authority of the state. Capacity values of slot coordinated airports are documented as declared capacity and used by slot coordinators as stipulation for allocating slot requests of airlines at these airports. Details of the EU slot regulation can be found in the Council Regulation No 95/93 (EU, 2004) and are well documented in the IATA World Slot Guidelines (IATA, 2017). The declared capacity specifies the number of aircraft movements that can be scheduled per unit of time (e.g. 10, 30 minutes or one hour) at an airport as estimated by the local airport stakeholders in the coordination committee. The main task of the committee is to determine the capacity of all functional elements of an airport, such as runways, taxiways, ramps, aircraft parking stands, gates, terminals or groundside facilities called coordination parameters that may have an influence on the total capacity of that airport, ‘taking account of all relevant technical, operational and environmental constraints as well as any change thereto’ (EU, 2004). The coordination committee is asked ‘to base the estimation on an objective analysis of the possibilities of accommodating the air traffic’ (EU, 2004). However, the members of the committee are free to choose the methodological approach of how to estimate the capacities of the infrastructural elements of the airport. In the case of diverging opinions the committee has to come to a consensus when proposing capacities to the administration to become the official coordination parameter values. The declared capacity is then the most critical coordination parameter. In high-volume airports, this is typically the runway system. This approach reflects an attempt to find a compromise between high and low capacity conditions at an airport, which cannot be predicted a season in advance. Table 2.3 gives declared capacity values for the year 2016 by runway capacity class for a selection of level three airports, where runway capacities constrain the number of aircraft movements. As can be seen, declared capacities vary between airports, depending among other things on local conditions such as weather, traffic and airspace structure, noise regulations. Typically, single runway airports in Europe have a practical capacity under IFR conditions of about 35 50 aircraft movements, if there are no severe obstacles as mentioned earlier, such as hills or special weather conditions. Two parallel runways that operated independently according to ICAO rules have a capacity of up to 90 aircraft movements. While in Europe take-offs and landings are normally planned for IFR weather conditions, VFR conditions typically govern in the United States,

26

PART | I Basic concepts

TABLE 2.3 Declared capacities of selected runway capacity constrained airports by runway capacity class in 2016 (own research of national and airport-specific slot coordinator organisations). Airport (IATA code)

Declared capacity (aircraft movements per hour)

Single runway airports London Gatwick (LGW)

38 50

London Stansted (STN)

36 50

Dublin (DUB)

38 46

Stuttgart (STR)

42

Geneva (GVA)

36

Airports with two intersecting runways Hamburg (HAM)

48

Warsaw (WAW)

38 42

Perth (PER)

38

Lisbon (LIS)

35 40

Airports with two close parallel runways Dubai (DXB)

57 65

Manchester (MAN)

44 56

Berlin Tegel (TXL)

52

Nice (NCE)

40 50 a

Du¨sseldorf (DUS)

43 45

Airports with two independent parallel runways Munich (MUC)

90

London Heathrow (LHR)

79 89

Oslo (OSL)

76 80

Palma de Mallorca (PMI)

66

a

Du¨sseldorf capacity is limited by administrative rules to a single runway capacity.

although in poor weather conditions, IFR flight rules will apply as well. Capacities are higher in VFR conditions, often by 10% 20%. San Diego (SAN) airport, as a single runway airport, has a capacity of 48 movements per hour in IFR conditions and up to 57 in VFR conditions. These values, together with the London Gatwick capacity, may be regarded as benchmarks of single runway capacity. Proxy estimates of practical runway capacity

Concepts of capacity and methods of estimation Chapter | 2

27

seem to be in the order of 35 45 aircraft movements in typical European weather and operating conditions and between 45 and 55 movements in good weather conditions as in the United States.

2.3.4

Simulation models

Simulation models represent a flexible tool often applied for estimating the throughput of airports under well-defined conditions. A typical question to be answered by applying simulation models is: Given the actual schedule of flights what are the possibilities of adding more flights in the present runway system or by adding infrastructure, for example, in the form of fast exit ramps or a new runway?

By simulating single aircraft movements in space and time according to prevailing operating conditions and rules simulation models can help to find measures to enhance the runway capacity by adding movements to an existing or postulated schedule of flights. Capacity limits are indicated by the increase of flight delays as a consequence of increasing aircraft movements. Event-based simulations are driven by complex software, in which the infrastructural and operational configurations are represented by a network of links and nodes, rules are defined for movements of aircraft from node to node, and discrete events are described for the arrival and departure times of aircraft. For this, geometric data (i.e. coordinates) of flight routes and the runway and taxiway system, of ramp and terminal areas, functional data of the runway layout, especially priority rules, movement data of aircraft, and structural data of the aircraft population and, last but not least, a flight plan need to be defined, because they form essential model input for the simulation to work. As a result of a simulation run, detailed reports of movement rates together with delay and other characteristics are produced for critical sections of the airport. Some of the better known simulation models are the Airport and Airspace Simulation Model, the Airport Delay Simulation Model (ADSIM) and the Runway Delay Simulation Model (RDSIM) of the US FAA, Airport Machine, SIMMOD and the Total Airspace and Airport Modeller of the Preston Group. These models are primarily used for optimising measures that are thought to enhance capacity, by simulating each flight movement step-by-step in time and space, and varying the configuration of runways, taxiways ramp areas and gate positions.

2.4

Conclusion

When speaking of airport capacity, it is often not clear what is meant by that term. As a modern airport with scheduled traffic consists of many capacity relevant components with different partial capacities, airport capacity

28

PART | I Basic concepts

typically refers to the capacity of the weakest element. In high-volume airports, this is, in many cases, the runway system, so that we typically speak of airport capacity while we mean the runway capacity of that airport. Furthermore, when speaking of capacity, we typically mean some kind of practical capacity rather than a theoretical capacity, which is, in contrast to the practical capacity, not related to a measure of the quality of operations. A rather common definition of practical capacity is the PHCAP introduced by the FAA, which specifies the number of aircraft movements that can be handled in one hour on a runway with an average aircraft delay of four minutes. Experience with the propagation of aircraft delays has shown that with an average delay of four minutes, single aircraft delays, which may exceed the average delay substantially, are not likely to increase further but rather stabilise or eventually diminish. When we say that an airport has a capacity of a certain number of aircraft movements, we may give the impression that the capacity has a more or less fixed value. In fact, the capacity is a variable depending on many influencing factors, such as weather conditions, separation requirements between landing and departing aircraft, traffic mix, local infrastructure conditions, and noiserelated and other environmental regulations. Depending on what value or condition is assumed for each influencing variable, the capacity will vary and thus form different levels of throughput limits. And, depending on short, medium or long-term tasks like assigning ATC slots in short-term actions or setting an official value of the declared capacity or planning for a new infrastructure component of an airport, the resultant capacity may be different for each task because the underlying assumptions regarding the values and composition of influencing factors may vary with the task at hand. As we have seen, there is a wealth of approaches to estimating runway capacity. The earlier approaches concentrated on mathematical functions, based partly on queuing theory. Traffic growth along with more complex traffic structures handled on multi-runway systems at airports have led to further development of analytical approaches, a highlight of which was the capacity handbook of the FAA in 1976. The main feature of the handbook was a great number of graphs by which the runway capacity could be calculated for a series of runway configurations, traffic structures and weather conditions. More recently, empirical approaches and simulation models have experienced a wider application. Two prominent examples of empirical analyses are benchmark studies, carried out primarily in the United States, and the consensus approach in the form of establishing a value of the declared capacity in the annual process of slot coordination at capacity constrained airports. Both the benchmark and consensus approach rely on qualitative and quantitative analyses, the relative importance of the main method depending on the group of experts charged with the estimation task. In preparing the airport capacity benchmark report of the core US airports, the FAA used two ways

Concepts of capacity and methods of estimation Chapter | 2

29

of approaching the capacity estimate. The first one was given by the local ATC team at each airport, which was based on comprehensive experience, and a second estimate was based on results of a simulation model (runwaySimulator). Since computers have become increasingly powerful instruments, they are used for simulating traffic and aircraft delay at airports in near-capacity conditions by adding single aircraft movements under varying options, for instance, of the runway system and other infrastructure elements. The declared capacity of slot coordinated airports worldwide is estimated by the coordination committee of each airport involved (level two and three airports), whereby each committee has the freedom to analyse the capacity by applying methods of their choice. It is likely that qualitative approaches in the form of expert estimates are applied, and different quantitative estimation methods, based, for example on simulation models, may play a role as well. This non-uniform procedure is one of the main reasons why declared capacities vary a great deal and are not comparable among airports. Nevertheless, declared capacity is estimated twice annually for over 300 level two and three airports, many of which are the major airports worldwide. Applying mathematical and simulation models requires the input of various data which is typically not part of the statistical data that each airport collects; for example time separation or occupancy times of aircraft, combined with weather data. Thus they have to be retrieved through special surveys or analyses. Due to the data requirements, these models have been mainly applied to single airports or a group of airports in a country but not to a great number of airports, say in Europe. In this book, our objective is to analyse the global capacity situation of airports, which means that the empirical approach has to be applied for all airports in a consistent way. We have therefore developed a new approach which includes empirical and benchmark elements. The methodological nucleus is the traffic volume ranking functions of congested airports, based on detailed flight data by airport. The model calculates hourly proxy capacities and capacity utilisation rates from these functions and relates them to ASVs. This approach enables us to analyse the capacity situation of a large number of airports in all parts of the world. Combined with a forecast of annual traffic volume, we estimate ASVs and calculate the capacity surplus or shortcoming of each constrained airport and propose mitigation measures.

References Ashford, N., Mumayiz, S., Wright, P., 2011. Airport Engineering, Planning, Design, and Development of 21st Century Airports. John Wiley & Sons, Hoboken, NJ. Blumstein, A., 1959. The landing capacity of a runway. Oper. Res. 7 (6), 752 763. de Neufville, R., Odoni, A., 2013. Airport Systems, Planning, Design, and Management. McGraw Hill, New York. European Union (EU), 2004. Council Regulation (EEC) No 95/93 of 18 January 1993 on Common Rules for the Allocation of Slots at Community Airports, revised as per Regulation 793/2004 of 21 April 2004, Brussels.

30

PART | I Basic concepts

Federal Aviation Administration (FAA), 1969. Airport Capacity Criteria used in Long-Range Planning, FAA Advisory Circular AC 150/5060-3A. Washington, DC. Federal Aviation Administration (FAA), 1976. Supporting Documentation for Technical Report on Airport Capacity and Delay Studies, Report No. FAA-RD-76-162, prepared for the FAA by Douglas Aircraft Company et al. Washington, DC. Federal Aviation Administration (FAA), 1983. Airport Capacity and Delay, FAA Advisory Circular AC 150/5060-5. Washington, DC. Federal Aviation Administration (FAA), 2004. Airport Capacity Benchmark Report 2004. Washington, DC. Federal Aviation Administration (FAA), 2014. Airport Capacity Profiles. Washington, DC. Gelhausen, M.C., Berster, P., Wilken, D., 2013. Do airport capacity constraints have a serious impact on the future development of air traffic? J. Air Transp. Manage. 28, 3 13. Gilbo, E., 1993. Airport capacity: representation, estimation, optimisation. IEEE Trans. Contr. Syst. Technol. 1 (3), 144 154. Harris, R.M., 1972. Models for Runway Capacity. Report FAA-EM-73-5. The MITRE Corporation, McLean. Horonjeff, R., McKelvey, F., Sproule, W., Young, S., 2010. Planning & Design of Airports. McGraw Hill, New York. International Air Transport Association (IATA), 2017. Worldwide Slot Guidelines. Montreal. International Civil Aviation Organisation (ICAO), 1984. Air Traffic Services Planning Manual. Montreal. International Civil Aviation Organisation (ICAO), 2009. Aerodromes, Annex 14 to the Convention on International Civil Aviation, Volume I: Aerodrome Design and Operations. Montreal. Matthews, L., 1995. Forecasting peak passenger flows at airports. Transportation 22 (1), 55 72. Odoni, A., 1972. Efficient operation of runways. In: Drake, A., Keeney, R., Morse, P. (Eds.), Analysis of Public Systems. MIT Press, Cambridge. Official Airline Guide (OAG), 2016. Market Analysis. Reed Travel Group, Dunstable. Urbatzka, E., Wilken, D., 1997. Estimating runway capacities of German airports. J. Transp. Plann. Technol. 20 (2), 103 129.

Chapter 3

Capacity utilisation at airports worldwide Following the description of concepts of capacity, in particular runway capacity, in this chapter, we focus on capacity utilisation. Capacity utilisation is meant to be a neutral description of the capacity constraint situation at airports. The term capacity utilisation relates the demand for runway usage, which is the number of aircraft movements, to the capacity of a runway or runway system, whereby the capacity equals the maximum number of aircraft movements per time unit at a specified or non-specified level of service (maximum throughput). As a mathematical function, the term can be written as Capacity utilisation 5

Trafficðnumber of aircraft movementsÞ Capacityðnumber of aircraft movementsÞ

As such, the function encompasses the whole range of capacity utilisation and not only values near the capacity limit, which stand for a capacity constraint situation. The prime interest of this book is, however, to analyse and identify capacity constrained airports with the objective of quantifying the capacity problem and showing mitigation measures. This chapter introduces the concepts of air traffic ranking curves and the capacity utilisation index (CUI) to measure capacity utilisation for identifying airport capacity bottlenecks worldwide. The CUI is defined as the ratio between the traffic volumes in the average daytime hour and the 5% peak hour. For estimating the CUI in future situations, we have developed a function relating the 5% peak hour volume with the annual traffic volume, since the latter is usually the subject of airport traffic forecasts. We categorised airports by traffic volume and CUI to identify airports according to the degree of capacity constraint. The chapter closes with a case study of how capacity utilisation has evolved over time at some example airports by comparing traffic volumes and patterns, traffic ranking curves and capacity utilisation at major airports in Europe, the United States and in Asia.

Airport Capacity Constraints and Strategies for Mitigation. DOI: https://doi.org/10.1016/B978-0-12-812657-8.00003-8 © 2020 Elsevier Inc. All rights reserved.

31

32

PART | I Basic concepts

3.1 Overview: how to measure capacity utilisation at airports? In order to assess the constraint situation at airports, we have to compare traffic volumes, as given by the number of aircraft movements, with the runway capacity. For the comparison, both the volume and capacity must have the same definition, that is the number of aircraft movements (take-offs and landings) per time unit. Demand and aircraft movements are normally given and forecast on an annual basis, while the capacity as a measure of the true throughput of the system is calculated or estimated for a short time period, typically for one hour or even shorter intervals. Annual capacities are used for long-term planning purposes as a measure of available service volume as well, rather than for measuring the maximum throughput of a runway in a comparable way. In answering the question of the conformity of traffic demand (aircraft movements) with capacity, (future) annual traffic volumes have to be converted into peak hour volumes and then compared with capacity. In this chapter, we discuss the problem of selecting a suitable peak hour and describe the distribution of hourly capacity utilisation of airports worldwide within a year by means of ‘traffic ranking curves’. Functional relationships between peak hour and annual traffic volumes, which have been derived for each type of airport capacity, allow the conversion of forecast annual volumes into peak hour volumes. Some hub airports, in particular London Heathrow and Beijing, are capacity constrained during most hours of the day. At many other major airports, capacity constraints occur mainly at peak times of traffic. However, off-peak times are periods of low traffic demand, such as night hours, Sundays and some holidays. In analysing the traffic distribution at airports over time, we have seen that differences in traffic volume between peak and off-peak periods are typically great, even at capacity constrained airports. Thus there seems to be only limited potential for shifting traffic from peak to off-peak periods and therefore we have to analyse peak and off-peak periods separately; talking about an average period would be misleading in terms of the capacity situation at an airport. Larger aircraft may be an option to release capacity constraints to some extent. Especially at hub airports, host carriers tend to increase their flight frequency to restrict competitors’ access to core markets (Pitfield et al., 2010; Givoni and Rietveld, 2009). However, there are limits to a strategy of operating larger aircraft; if an airport has a great deal of feeder traffic to a number of smaller airports, frequencies are typically higher (Brueckner and Zhang, 2001). Frankfurt is an example of such an airport. On the other hand, London Heathrow is an example of a hub airport with less feeder traffic to smaller airports. Consequently, average seat capacity per flight is a little lower in Frankfurt (174 seats per flight in 2016) than in London Heathrow

Capacity utilisation at airports worldwide Chapter | 3

33

(200 seats per flight in 2016) [Official Airline Guide (OAG), 2016]; however, the capacity situation is worse in London Heathrow than in Frankfurt. Given the two-fold approach, namely the annual demand forecast and the estimation of hourly capacity, a conversion rate is needed to harmonise the time units. Since an annual capacity value fails to describe the throughput capability of the runway system in a meaningful way  due to the fact that demand will never reach near-capacity levels in all 8760 hours of the year  the annual traffic volume has to be converted to an hourly volume. In this context, several methodological issues arise. They are as follows: G G

How to estimate the hourly capacity of a runway system? How to estimate the relevant hourly demand for runway usage, which is the traffic volume?

The latter issue requires, first, the selection of an hour of the relevant traffic year, which can be compared with capacity. Consequently, the following questions have to be answered: G G

Which hour of the year is to be used for comparing traffic with capacity? How to estimate the traffic volume for this hour based on the forecast annual demand?

This chapter focuses primarily on the two issues of selecting a relevant ‘peak hour’ and estimating the traffic volume in that hour. The issue of capacity concepts has been dealt with in Chapter 2, Concepts of capacity and methods of estimation, and will be discussed further in the context of the question of the peak hour selection, although not in great detail, since this research question has been treated extensively in the literature [Ashford et al., 2011; Federal Aviation Administration (FAA), 1969, 1983; Horonjeff et al., 2010]. There have been previous studies on this subject. Urbatzka (1991, 1993) developed a method for estimating the runway capacity of German airports with one runway and of Hamburg airport. Weigelt (1994) estimated a functional relationship between hourly and annual aircraft movements for six German airports. Matthews (1995) reviewed some concepts and issues relating to peak passenger flows at airports and the relationship between peak passenger flows and annual traffic volume. Urbatzka and Wilken (1997) developed an analytical approach for estimating the runway capacities of German airports with a single or multi-runway system and included parameters, such as aircraft weight mix, runway use, sequencing of arrivals and departures and arrivaldeparture ratio. Wang and Pitfield (1999) estimated a functional relationship between a 3.5% design peak hour traffic and annual throughput for departing passengers for 48 Brazilian airports. For selecting the relevant peak hour demand, which serves as a base for comparing traffic with capacity in order to give a quantitative measure of capacity utilisation, many options are conceivable; there is no standard method for determining this specific value. Theoretically, an hour with

34

PART | I Basic concepts

average traffic volume or, even better, one with some kind of peak traffic could be selected. In planning studies, typically a peak hour is used on grounds that the infrastructure should be able to handle traffic in high demand situations that occur regularly. Several concepts of ‘typical peak hours’ have been developed and applied in airport planning (Ashford et al., 2011). Two concepts, which are of special interest in this context and which have been applied in the past for instance by BAA (Wang and Pitfield, 1999) and for German airports (Urbatzka and Wilken, 1997), are as follows: G

G

The hour with the 30th highest traffic volume of the year of planning horizon, called the 30th peak hour. The hour with a traffic volume that is reached or exceeded in 5% of all operational hours of the planning year, called the 5% peak hour.

Other concepts basically differ in their choice of the peak hour and percentile, respectively. In order to find and identify these hours of a base year, the traffic by day and hour has to be known for all operational hours of the year and then ranked according to traffic volume. Traffic ranking curves do exactly that; by interpreting the slope of the function, one can identify and delimit a range with typical peak hour volumes, in contrast to those with very high peak hour volumes, on the one hand, and average hour and offpeak hour volumes, on the other hand. These traffic ranking curves will be introduced in the following section.

3.2

Air traffic ranking curves

Air traffic ranking curves are a tool for visualising the capacity situation at an airport. Aircraft movements per hour are displayed in descending order for all hours of the year, starting with the busiest hour. This functional relationship is called the ‘air traffic ranking curve’ (e.g. Matthews, 1995; Urbatzka and Wilken, 1997; Weigelt, 1994; Wilken et al., 2011). Data may be retrieved from airport statistics of flights at that airport or from airline schedule data, which is provided by the Official Airline Guide (OAG) for more than 4000 airports. If the latter is used, one has to be aware of the fact that these databases contain, first, all airline schedule data but not data of other flights, such as ad hoc charter cargo flights of freight integrators, transfer flights and other commercial or non-commercial flights. A comparison of traffic at all German airports in 2008, as provided by official German statistics and the OAG, has shown that the OAG traffic volume captured almost 90% of the total volume. The difference between airline traffic (based on OAG data) and total traffic becomes smaller with the growing importance of airports since they concentrate on airline traffic, whilst airports with less traffic are more frequented by non-scheduled and other commercial traffic. If we take only the five busiest airports in Germany, which handle about twothirds of total German air traffic, the difference between the OAG data and

Capacity utilisation at airports worldwide Chapter | 3

35

total traffic narrows down to 3%. In the following discussion, we base our analysis on OAG data, since they form the only available base which offers data from airports around the world (Wilken et al., 2011). Before exploring the ranking of traffic by hour, it is useful to study the traffic pattern by hour, day and month in order to demonstrate that air traffic is not a uniform phenomenon over time but rather one with a high degree of variability. In order to visualise traffic patterns at airports we selected  more or less arbitrarily  three busy airports with different volumes and traffic compositions are as follows: G

G

G

Frankfurt (FRA) in Germany, a major hub airport with 454,775 aircraft movements in 2016 (OAG), is used primarily by full-service network carriers, with peak hour capacity problems. London Stansted (STN) in the United Kingdom, a single runway airport with mainly origindestination traffic of 154,785 aircraft movements in 2016 (OAG), is used to a high degree by low-cost carriers. Palma de Mallorca (PMI) in Spain, an airport with two independent runways and mainly touristic traffic of 172,340 aircraft movements in 2016 (OAG), is used with a seasonal peaking during the summer season.

Fig. 3.1 displays the hourly variation of aircraft movements at Frankfurt, London Stansted and Palma de Mallorca airports during the peak day in 2016. Since we are comparing peak days, we have identified different dates for each airport, namely the peak day of the year 2016 of each airport. As can be seen, the traffic variation is rather pronounced at the three airports, even on peak days. At London Stansted the ratio between high and 120

Aircraft movements per hour

Frankfurt: 1394 aircraft movements (31 August 2016)

100

80 Palma de Mallorca: 895 aircraft movements (11 August 2016)

60

40

London Stansted: 489 aircraft movements (28 August 2016)

20

0

0

1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20 21 22 23

Hours of the day

FIGURE 3.1 Hourly variation of aircraft movements during peak day 2016 at Frankfurt, Palma de Mallorca and London Stansted airports [Official Airline Guide (OAG), 2016].

36

PART | I Basic concepts

low traffic is in the order of more than two. Variation is a little less at Frankfurt, due to the capacity constraint in peak hours limiting the number of aircraft movements there. Frankfurt had a declared capacity of 100 aircraft movements for scheduled flights per hour in 2016, which was reached in three hours of the day. The amount of night-time traffic at the airports, particularly between midnight and 5:00 a.m., was rather small. Such a low demand is typical for most airports but not for all, see for example Dubai (DXB) airport. Furthermore, cargo traffic, in particular integrator freight, is concentrated during night hours and only at some airports. There are only relatively few airports with intensive night operations, such as Memphis and Cologne/Bonn. Also, the variation of traffic by day during a week shows the low demand for night movements and, in addition, the weaker demand at weekends at many airports. Fig. 3.2 shows the hourly variation of aircraft movements at Frankfurt, London Stansted and Palma de Mallorca airports during the peak week of each airport in 2016. At London Stansted and Palma de Mallorca the traffic dropped significantly at weekends. On the other hand, in Frankfurt, the relative weekend traffic reduction was much smaller. Other airports with smaller traffic volumes are likely to have a more pronounced difference between weekend and weekday traffic. In addition to the hourly and daily variation of traffic, there is the seasonal variation during a year, as shown in Fig. 3.3. Typically, the summer months are characterised by high demand levels, whereas from November to February, traffic is relatively low. This is especially true for London Stansted airport that handles much low-cost traffic with passengers travelling for private reasons; for example sightseeing purposes during the summer 120

Aircraft movements per hour

Frankfurt: 9542 aircraft movements (5 to 11 September 2016)

100 80 Palma de Mallorca: 5508 aircraft movements (1 to 7 August 2016)

60 London Stansted: 3313 aircraft movements (22 to 28 August 2016)

40 20 0

0

6 12 18 0

6 12 18 0

Monday

Tuesday

6 12 18 0 Wednesday

6 12 18 0 Thursday

6 12 18 0 Friday

6 12 18 0 Saturday

6 12 18 Sunday

FIGURE 3.2 Hourly variation of aircraft movements during the peak week in 2016 at Frankfurt, Palma de Mallorca and London Stansted airports [Official Airline Guide (OAG), 2016].

Capacity utilisation at airports worldwide Chapter | 3 45,000

37

Frankfurt: 454,775 aircraft movements

Aircraft movements per month

40,000 35,000 30,000 25,000

London Stansted: 154,785 aircraft movements

20,000 15,000

Palma de Mallorca: 172,340 aircraft movements

10,000 5000 0

FIGURE 3.3 Monthly variation of aircraft movements in 2016 at Frankfurt, Palma de Mallorca and London Stansted airports [Official Airline Guide (OAG), 2016].

season. In Palma de Mallorca, traffic used to be relatively low in winter months in the past as well. However, Mallorca has become a prime holiday destination for Europeans, so that winter traffic is not substantially lower than traffic in the summer season. Even in Frankfurt, where airlines are hindered in peak hours from freely planning and offering flight schedules to match demand due to capacity restraints, the traffic volume is clearly lower in the winter months than in the summer season. Fig. 3.4 displays the hourly traffic ranking at Frankfurt, Palma de Mallorca and London Stansted airports during 2016. The curves show the number of hours at each volume level, starting with the highest hourly volume and ending with the lowest volumes of one and no movement per hour. The horizontal axis gives the traffic hours of the year in the order of the traffic volume of each of these hours, while the vertical axis shows the hourly traffic volume. As can be seen, in the middle range of the traffic volume curve, there are many hours with little variation in the level of traffic, whereas at the highest and at the lower end of traffic volumes in early morning and late evening hours, its change is more pronounced with each hour in the ranking sequence. In addition, there are many hours with very low or zero demand, which are usually the hours at night. The less inclined the slope of the curve over long periods of time, the closer is the airport to reaching capacity, if and only if the top-ranking hourly volumes reach levels near the capacity. Comparing the three curves, one can see that the Frankfurt and London Stansted curves are less inclined than the Palma de Mallorca curve, thus indicating a higher capacity utilisation at the former airports.

38

PART | I Basic concepts

Aircraft movements per hour

120 100 80 60

Frankfurt: 30th peak hour volume: 104 5% peak hour volume: 98

Palma de Mallorca: 30th peak hour volume: 62 5% peak hour volume: 56

40 20 0

London Stansted: 30th peak hour volume: 44 5% peak hour volume: 40

1 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75 78 81 84 87 Hours of the year

FIGURE 3.4 Traffic ranking by hours of operation of the year 2016 at Frankfurt, Palma de Mallorca and London Stansted airports [Official Airline Guide (OAG), 2016].

Since the chosen airports have different runway systems  from one runway at London Stansted to four at Frankfurt  with different capacities, the ranking curves commence at different levels of traffic volume. The highest hourly volume in Frankfurt is nearly 110 aircraft movements per hour, while London Stansted handles a maximum of nearly 50 movements. As already mentioned, the declared capacity of Frankfurt airport was 100 aircraft movements per hour in 2016. This value was reached and exceeded in about 5% of all hours of the year as can be seen in the ranking curve. The 5% peak hour volume was 98 movements, a value very close to the declared capacity. If traffic ranking curves are belly shaped like those of Frankfurt and London Stansted, then the 5% peak hour volumes are indicative of the practical capacity of the airport, as the analysis of airports with Frankfurt-like ranking curves has shown. This characteristic of ranking curves will be used to identify airports with high capacity utilisation as will be shown later in this chapter. The ranking curves demonstrate hours of ‘typical traffic’ where volume changes gradually as the hours change, and those with ‘atypical traffic’ where volume changes rapidly as the hours change. This is the case in the hours with the highest traffic volumes of the year and in the morning and evening hours when traffic picks up or goes down rapidly. Typical traffic during the day would be found in the range of the 20005000th hour, depending on the number of operating hours per year, because relative changes of volume between following hours in the ranking are rather small and smaller than in other hours. In the case of Frankfurt airport, the mean of a typical hourly traffic volume during a day in 2016 was about 77 air transport movements. The variation of this volume during the day is about 127%

Capacity utilisation at airports worldwide Chapter | 3

39

and 235%, that is between 50 and 98 movements if, on the one hand, the absolutely highest volumes of the first 400 hours or so and, on the other hand, the hours at night (beyond the 5400th hour of the year) are excluded. London Heathrow stands for an airport with high traffic volume and capacity utilisation (about 476,000 aircraft movements in 2016). The declared capacity reaches a maximum value of 88 movements per hour for some hours of the day, and the 5% peak hour volume in 2016 has been found to be 87 aircraft movements, which is close to capacity (see Fig. 3.5). Since the slope is such that the variation of hourly traffic is very small over many hours of daytime operations, the highest volume values of the curve are indicative of the capacity of the airport, which in such cases is typically the capacity of the runway system. It has been found that the London Heathrow ranking curve has for long been the curve with the lowest slope among all airports, demonstrating that London Heathrow has been the airport with the highest capacity utilisation. Fig. 3.6 displays the air traffic rankings for a sample of 60 airports. Here, ranking curves have been normalised relative to the hour with the highest traffic volume of the year to give a more general picture of the structure of such ranking curves. Of these 60 airports, the top five and the lowest five airports in terms of annual aircraft movements have been chosen at a time from each capacity class, which will be described in more detail later in this book. However, there is a lower bound of 70,000 aircraft movements per year for an airport to be included in the data sample; thus small airports are not represented in Fig. 3.6, because their peak hour volumes are typically far below the capacity of their runway system. The horizontal axis gives, as before, the traffic hours of the year in the order of the traffic volume of each 30th peak hour volume: 90

Aircraft movements per hour

100 90

5% peak hour volume: 87

80 70 60 50 40 30 20 10 0

1 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75 78 81 84 87

Hours of the year

FIGURE 3.5 Air traffic ranking curve of London Heathrow [Official Airline Guide (OAG), 2016].

Aircraft movements per hour to aircraft movements of the busiest hour of the year

40

PART | I Basic concepts 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

Hours of the year

FIGURE 3.6 Relative traffic ranking by hours of operation of the year 2016 for a sample of 60 airports worldwide [Official Airline Guide (OAG), 2016].

of these hours, while the vertical axis shows the ratio of the hourly traffic volume to the highest hourly volume. As can be seen, ranking curves vary widely in their form, depending on the level of capacity utilisation. The number of operating hours differs as well, as only some airports operate all 8760 hours of the year. Likewise, Fig. 3.7 shows the normalised air traffic ranking curves of the three airports Frankfurt, London Stansted and Palma de Mallorca  the absolute ranking curves of which were displayed in Fig. 3.4  during 2016 relative to the hour with the highest traffic volume of the year in more detail. As can be seen, in the middle range of traffic volume, there are many hours with little variation in the level of traffic, whereas at the very high traffic volumes (at the left end of the curve) this is less the case. These features are more pronounced at airports with higher capacity utilisation, see for example Frankfurt, than at airports with higher capacity reserves. The curves have the same general form but differ in the slope over the hours with high volumes. The higher the traffic in relation to capacity, meaning the higher the capacity utilisation, the less the slope is inclined, except for the starting part of the curve. Accordingly, the slope is very inclined in the hours with the highest hourly volumes. As can be seen, after about 300500 hours of operation and depending on the capacity utilisation, the slope changes and becomes less inclined. Over the following 5000 hours or so, the slope remains rather constant; slopes become more inclined or even very inclined again during the following 1000 hours or so and the slope levels off in the remaining hours. Assuming around 1517 hours of demand

Capacity utilisation at airports worldwide Chapter | 3

41

Aircraft movements per hour to aircraft movements of the busiest hours of the year

1.0 0.9 0.8

Frankfurt: 454,775 aircraft movements 30th peak hour volume: 104 360th hour or 5% peak hour volume: 98

0.7 0.6

London Stansted: 154,785 aircraft movements 30th peak hour volume: 44 366th hour or 5% peak hour volume: 40

0.5 0.4 0.3 0.2 0.1

Palma de Mallorca: 172,340 aircraft movements 30th peak hour volume: 62 363th hour or 5% peak hour volume: 56

0.0

Hours of the year

FIGURE 3.7 Relative traffic ranking by hours of operation of the year 2016 at Frankfurt, Palma de Mallorca and London Stansted airports [Official Airline Guide (OAG), 2016].

during daytime and correspondingly 79 hours in early morning and late evening and with rather low demand at night or even with a night curfew with zero demand, there are about 55006200 hours with relatively high demand, as demonstrated by the ranking curves. Accordingly, there are about 26003300 hours of low or zero demand at night, depending on the existence of night flight regulations. As has been shown in Figs 3.13.3, traffic varies in time, in particular between day and night, during the day between peak and off-peak hours, between weekdays and weekend and also between seasons of the year. The variation decreases not only with the traffic importance of the airport but also with the seriousness of the constraint on the airport’s capacity. Nevertheless, even in important hub airports and capacity restrained airports, the difference between traffic in low demand situations, such as at night, and in peak demand situations remains very high. The high degree of traffic variation is, therefore, a fundamental argument for an hourly capacity concept. Any kind of annual capacity would be subject to a discussion of how to account for the intervals of low demand during a year (see Chapter 2: Concepts of capacity and methods of estimation). As can be seen in Fig. 3.4, the hour with the 30th highest traffic is a rather ‘atypical’ peak hour, since the traffic varies sharply between neighbouring hours in the rankings. If the objective is therefore to derive a ‘typical’ peak hour, then we have to search for such an hour in that part of the ranking function, where the curve changes from steep into a less inclined and rather constant flat gradient, that is roughly between the 350th and 450th operating hour (although this is somewhat less the case in the London Stansted and Palma de Mallorca examples). The exact determination of the

42

PART | I Basic concepts

typical peak hour within this area remains arbitrary. Therefore we chose the 5% peak hour as analogous with the 5% busy hour rate, which is among others used as a base for infrastructure planning by BAA in the United Kingdom (Matthews, 1995; Urbatzka and Wilken, 1997; Poole, 1972). The 5% busy hour rate is defined as the value of passenger flow for which 5% of the passengers encounter a flow rate at this level or above (Matthews, 1995). Matthews (1995) and Poole (1972) argued that the 5% busy hour rate in the planning process of BAA is a sound choice, because it is less volatile than some other measures in use, for example the 30th peak hour. Similar to the 5% busy hour rate, we chose a 5% peak hour based on aircraft movements for two reasons. First, runway capacities are naturally a matter of aircraft movements rather than passenger flows, aside from the fact that both are highly correlated. Second, traffic data, that is aircraft movements, by hour is available for most airports worldwide, for example from OAG, thus allowing a global analysis and making the approach of airport utilisation applicable to a large number of airports worldwide. Moreover, the choice of the 5% peak hour is supported by the results of the regression analysis of functionally relating peak hour with annual traffic volume to be dealt with later in this book. Here, among other things, a high R2 serves as evidence of a sound choice of peak hour (Wang and Pitfield, 1999). As a result, airport planners, for example BAA, which owns and operates airports in the United Kingdom, have often chosen a 5% value as a basis for designing facilities and capacity (Matthews, 1995; Urbatzka and Wilken, 1997; Poole, 1972). After demonstrating the peak hour characteristics by means of the ranking curves of many airports, we have adopted this concept and in the following discussion deal with the 5% peak hour as the typical peak hour chosen for estimating capacity reserves and utilisation. In a network or global analysis the ranking curves, as developed based on OAG data, allow determining the 5% peak hour traffic volume (PHTV) of airports to be used in a consistent way. An additional advantage is given by the fact that peak hour volumes are comparable between airports. In addition, given that traffic volume of an airport exceeds a certain threshold, which is roughly about 70,000 aircraft movements, ranking curves are a tool for analysing and estimating hourly capacity of airports working under near-capacity conditions. However, these numbers are only rough guidelines and are based upon empirical analysis of about 4000 airports. If the traffic volume of an airport is too low compared to its runway capacity, the peak hour volume typically lies well below the runway capacity. As a result, analysis of capacity utilisation is not feasible by means of the traffic ranking curve. However, in the case of sufficient traffic volume, the question as to whether or not an airport has reached almost capacity in daily operation and to what extent the airport is constrained can be seen easily by looking at the slope of the ranking curve over all hours of the day (excluding night hours). The highest volume values of the Frankfurt curve are indicative of

Capacity utilisation at airports worldwide Chapter | 3

43

the capacity of the airport, which in such cases is typically the capacity of the runway system. Also, at London Stansted, which does not operate near capacity, the highest traffic volumes reach levels of 45 to almost 50 movements per hour, which correspond roughly with the instrument flight rules (IFR) capacity of a single runway. These values are also rather close to the declared capacity of the airport in 2016, which was set to 50 aircraft movements per hour. It should be noted in this context that the number of movements, according to OAG data, corresponds to the on-block and off-block occurrences, whereas the capacity refers to the number of movements on the runway. A measure of capacity utilisation is given by relating the demand of the 5% peak hour with the capacity. At London Stansted, in 2016, the corresponding volumecapacity ratio was about 40/50 and, therefore, there was a peak hour capacity utilisation of 80% in IFR conditions. In the case of Frankfurt the utilisation was about 98% (98/100), indicating that in typical peak hours occurring every day, the declared capacity (under IFR conditions) was almost reached. So far, we have introduced air traffic ranking curves and discussed the pros and cons of different peak hour concepts. The next section shows how we can derive systematically peak hour volume from a given annual service volume; for example from an air traffic forecast, such as the Airbus Global Market Forecast or Boeing Current Market Outlook.

3.3

Peak hour volume and annual service volume

The objective of estimating the relationship between peak hour and annual traffic volumes is to identify the peak hour traffic for a planned situation, when the annual volume of this situation is known, for example as a result of a forecast. The need to apply such functions typically arises in airport planning processes, when studies are made to analyse the need for extending runway capacity due to a lack of existing reserves. Usually, demand forecasts of annual traffic form the quantitative base for evaluating planning proposals. Since the need for planning new capacity arises primarily in situations when the existing traffic approaches capacity, we have concentrated our analysis on those 200 of the 4054 airports that are characterised at least by a reasonable high degree of capacity utilisation, which is explained in more detail later in this chapter. A further subdivision of the total airport sample into ‘capacity classes’ was indicated because the demand and, therefore, the peak hour volume cannot exceed the capacity of the runway system, the level of which is dependent on the number of runways and their configuration. However, we had to ensure that each capacity class has enough member airports for regression analysis to produce sound results, so that there is a trade-off between more detailed capacity classes and the number of their

44

PART | I Basic concepts

member airports. This relationship has to be balanced well, especially for airports with four runways and more. In the sample of 200 high-volume airports given, we have identified the following capacity classes, for which functions have been calibrated, and they explain a high degree of the variance in the data sample, thus supporting the chosen capacity classes: G G G G G G

Single runway Two runways, independent parallel Two runways, dependent parallel Two runways, crossing Three runways Four runways and more

For these 200 airports, we collected data on traffic schedules [Official Airline Guide (OAG), 2008] and runway configurations [Digital Aeronautical Flight Information Files (DAFIF), 2008] and calculated traffic ranking curves with 5% peak hour volumes and the number of hours of day and night operation. In order to verify the functional relationship between peak hour volume measured in hourly aircraft movements and annual traffic volume measured in annual aircraft movements, we studied several functional types in regression analysis. The overall best result was obtained by a function describing the 5% peak hour volume in relation to annual volume (annual aircraft movements, AACM) or the logarithm of annual volume LN (AACM), respectively, a factor of annual utilisation of the runway (greater five, GF) and a binary variable which describes whether an airport is located in Europe (EUR) and thus operating under conditions of slot coordination and IFR in air traffic control or not. We chose the number of hours with more than five aircraft movements per hour to describe annual runway utilisation, roughly corresponding to the number of day hours (in contrast to night hours) and which describe the time period with high demand. Therefore the variable GF is included to account for differences in opening hours, flight restrictions (especially night curfews), etc. Table 3.1 shows the calibration results for the selected categories of runway system. The variables are all highly significant at levels of 1% or less and, despite the many different airports in a given class of runway system, we have identified a surprisingly stable relationship between the 5% peak hour volume, annual aircraft movements, the number of hours with more than five aircraft movements per hour and whether the airport is located in Europe and thus slot coordinated and operating under IFR conditions or not. Depending on the capacity class of an airport, the model explains between 90% and 99% of the observed variance (R2) in the data sample. For a single runway airport, for instance, the function for the 5% PHTV is given by 5% PHTV 5 2 213:08535 1 22:90694 3 LNðAACMÞ 2 0:00409 3 GF

TABLE 3.1 Estimation results (dependent variable: 5% peak hour volume of an airport) (Wilken et al., 2011). RWY system Single RWY

Variable Constant LN(AACM)

Coefficient

Mean

Min/max

2213.08535



22.90694



87473

72,360/197,511

5831

4388/7481

GF

20.00409



Two RWYs

Constant

49.91377



independent

AACM

0.00021



180,545

75,668/430,154

20.00703



6966

5927/8581

34.99945



0.00020



155934

73,367/347,602

20.00479



6735

5124/8784

2353.48633



LN(AACM)

37.42267



139783

74,270/386,757

GF

20.00812



6613

5466/8783

EUR

23.14930



2500.05569



47.78217



210778

72,261/479,294

20.00453



6781

5990/8641

24.22136



GF Two RWYs dependent

Constant AACM GF

Two RWYs crossing

Three RWYs

Constant

Constant LN(AACM) GF EUR

R2

No. of obs.

89.41%

58

98.12%

23

96.29%

29

98.51%

21

93.49%

40

(Continued )

TABLE 3.1 (Continued) RWY system Four RWYs and more

Variable Constant AACM GF EUR

 

Coefficient

Mean

Min/max

77.05064



0.00021



377438

113,195/956,380

20.01065



6928

5782/8334

20.00104



R2

No. of obs.

99.15%

29

Significant at the # 1% level. EUR, Europe; GF, greater five; LN(AACM), logarithm of annual aircraft movements; RWYs, runways; AACM, annual aircraft movements.

Capacity utilisation at airports worldwide Chapter | 3

47

The functions of the other runway capacity classes are very similar. The higher the annual aircraft movements, the higher is the 5% peak hour volume. On the other hand, the higher the number of hours with more than five aircraft movements per hour, the lower is the 5% peak hour traffic. This conclusion may need some explanation. As mentioned, we have included only airports that have at least a reasonably high degree of capacity utilisation and have a minimum of about 70,000 aircraft movements per year. Therefore, as the number of unrestricted operating hours at an airport increases, flexibility in scheduling aircraft movements increases due to more available capacity and, consequently, peak load decreases. This leads to a reduction of the 5% peak hour volume. If, on the other hand, we look at small airports, which have ample capacity reserves, the relationship between the number of hours with more than five aircraft movements per hour and the 5% peak hour volume reverses, because of a low level of capacity utilisation. Our analysis therefore focused on airports with a high degree of capacity utilisation, which is roughly given at airports with annual traffic volumes of more than 70,000 aircraft movements. For these airports the number of hours with more than five aircraft movements per hour has a negative effect on the 5% peak hour volume. As can be seen, the EUR variable has a negative effect, too, although not for airports with one or two parallel runways. If an airport has two crossing runways or more, or is more complex, and less independent runway configurations, slot coordination and IFR conditions in air traffic control lower 5% peak hour volumes. For example, if there were two identical airports with three runways in Europe and the United States, 5% peak hour volumes would be on average about four movements lower at the European airport than at the US airport. First-come-first-served and visual flight rules conditions in the United States tend to favour the throughput of the runway and peak load volumes in hub operations. We would assume that the dampening effect of the EUR variable is about as strong in the lower as well as upper capacity classes of airports. A reason why this variable is not significant in the 5% peak hour functions of the lower capacity classes may be that the sample airports in these classes are to a lesser degree hub airports, thus requiring less capacity in peak hours than in inbound and outbound traffic waves at hub airports, and are not operating at or near capacity as often as airports in the upper capacity classes. Further research is needed to clarify the differences of effects of the EUR variable in each airport capacity class. Nevertheless, the explanatory power of the submodels is with R2 between 90% and 99% very high. We mainly focused on potentially capacity constrained airports. Therefore Table 3.1 shows the mean, minimum and maximum values of the observed data of the AACM and GF variable, since the estimated functions are reliable only within a certain range. However, this range is not limited

48

PART | I Basic concepts

strictly by the minimum and the maximum values of the independent variables in Table 3.1. The model includes only a small number of independent variables for several reasons. First, we are limited in the number of variables because of the small sample of relevant airports. Second, the spare number of independent variables makes the model easy to apply for forecasting purposes. As we are interested in the relationship between the 5% peak hour volume and the annual aircraft movements at an airport, the first independent variable is given by definition. The number of hours with more than five aircraft movements per hour at an airport is a variable, which may be defined reasonably well in a scenario analysis. Model fit is good since about 90%99% of the observed variance is explained by the model. Consequently, there is no need to forecast a large number of independent variables for model application.

3.4

The capacity utilisation index

In this chapter, we develop an index of airport capacity utilisation to assess the capacity reserve of an airport. This index is designed to be applied in airport networks up to the global level and is based on the concepts of the 5% peak hour and the average hour traffic volume of an airport, which are deduced from air traffic ranking curves, as described in Section 3.2. The basic idea behind the capacity utilisation index (CUI) is that the more inclined an air traffic ranking curve is, the greater is the capacity reserve of an airport. These ranking functions have been derived for around 4000 airports worldwide on the basis of OAG data. The traffic volume of an airport serves as a measure of its importance in the global air traffic network. In connection with the airport CUI, we can categorise airports according to their importance within the global air traffic system and of the level of capacity reserve or of capacity constraint, respectively. Thereby, we can identify those airports that are critical for future growth of air traffic, that is those airports which serve a large share of the global flight volume and, at the same time, have only a small capacity reserve. Fig. 3.8 exemplarily displays the number of aircraft departures in Europe as well as at the European hub airports of Frankfurt and London Heathrow. The year 2000 forms a reference basis (2000 5 100). Because Frankfurt had already reached capacity limits then, until a new runway was built in 2011, and London Heathrow has been operating at the capacity level for many years, these two main airports have only partially participated in the general market growth in the past (200617). Both airports are slot coordinated, and free slots are only available at unattractive times, such as night times or weekends. As compared with the year 2000, the number of departures increased by 32% at European airports, primarily since the year 2013. The number of departures increased by only 10% at Frankfurt and by just 3% at London Heathrow. From 2006 to 2008 the number of take-offs remained

Capacity utilisation at airports worldwide Chapter | 3

49

135 130

Year 2000 = 100

125 120 115 110 105 100

2006

2007

2008

2009

2010 Europe

2011

2012 Frankfurt

2013

2014

2015

2016

2017

Heathrow

FIGURE 3.8 Development of the number of take-offs in Europe and at Frankfurt and London Heathrow airports [Official Airline Guide (OAG), 2017].

almost constant at these airports, although departures increased by nearly 10% points at European airports. In 2009 the number of take-offs dropped less than in Europe overall and in 2010 they basically reached precrisis levels. This indicates that Frankfurt and London Heathrow already had severe capacity problems in these years; the latter is still operating at the limit. Frankfurt traffic did not grow from 2011 to 2016 for several reasons, in particular because of a lack of local demand growth, no favourable airport policy for low-cost carriers until 2016 and a Lufthansa strategy of diverting parts of the hub traffic to other Lufthansa hubs, such as Munich. Fig. 3.9 shows the cumulative distribution of airport traffic in the global network in 2016. The distribution reveals the strong concentration of global air traffic at relatively few and thus the largest airports. The top 100 airports as measured by the number of flights represent 2.5% of around 4000 airports, which have been subject to this analysis in 2016, but they handle almost half (45%) of the global flight volume of 35.5 million flights. Moreover, the top 1000 airports (25%) cover 91% of the global air traffic volume. This is illustrated by the so-called Lorenz curve of Fig. 3.9, which notably deviates from the 45 degree line, thus indicating an uneven allocation of flights among the airports (Gini coefficient 5 0.80). Correspondingly, there are about 3000 airports with traffic volumes as low as to account for just 10% of the total volume. As a result, there are a relatively small number of airports with high traffic volumes, such as hub and other main airports and, on the other hand, a great number of airports with low traffic, forming thus  at least theoretically  an important capacity reservoir.

50

PART | I Basic concepts

Share of global flights (n = 35.5 million, 2016)

100% 90%

Top 1000 airports (24.7%) handle 32.2 million flights (90.7%)

80% 70% 60% 50%

Top 100 airports (2.5%) handle 15.9 million flights (44.8%)

40% 30% 20% 10% 0%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Share of airports (n = 4054, 2016)

FIGURE 3.9 Relative cumulative distribution of flights in the global airport network in 2016 [Official Airline Guide (OAG), 2016].

Basically, to measure the degree of capacity utilisation of a particular airport, the concepts of the 5% peak hour and the average hour traffic volume are applied. The average hour volume is defined by the annual aircraft movements divided by the annual operating hours of the airport. However, details of computing the average hour depend on the number of aircraft movements of an airport and will be described later in this book. Fig. 3.5 in Section 3.2 displayed a ranking curve for London Heathrow, which handled about 476,000 aircraft movements in 2016 and has been one of the most saturated airport in our data sample. The absolute value of the slope of the ranking curve serves as a measure for capacity utilisation of an airport. The CUI is therefore defined by the ratio of the average hour to the 5% PHTV: Capacity utilisation index ðCUIÞ 5

Average hour traffic volume 5% peak hour traffic volume

The selection of the average hour volume seems to be an intuitively sound choice; however, it is more or less arbitrary. Basically, any different hourly volume in the region of the average hour may be employed and thus it is only a matter of calibration, that is setting a critical value or a bandwidth of values, respectively, to separate constrained airports from those which are not constrained. The key point is to take two points on the curve to define a metric which approximates the slope of the curve in a fashion that should be both representative and comparable between airports of a

Capacity utilisation at airports worldwide Chapter | 3

51

minimum size (aircraft movements .70,000 per year). We have also tried a different ratio and divided the 5% peak hour by the 30th peak hour. The ranking of airports changed in a few cases, but, moreover, we felt that this ratio focused too much on traffic in annual peak hours and neglected operating hours with a lower degree of capacity utilisation. Therefore we have chosen to retain the original definition of the CUI to have a broader view on capacity utilisation of an airport with regard to the operating hours. In the case of London Heathrow, an airport with two runways, the CUI is 72/87 5 0.82 and it is among the highest ratios of airports that have been recorded for the year 2016. Here, the average hour traffic is computed only on the basis of such operating hours, which exceed five aircraft movements to reflect factors, such as night restrictions or low demand situations during night hours. Otherwise, these hours would keep the average value of aircraft movements per hour artificially low and thus suggest a lower degree of capacity utilisation, see for example hours 65008760 at London Heathrow in Fig. 3.5. In 2016 London Stansted, an airport with a single runway, handled almost 155,000 aircraft movements. Its CUI value was 24/40 5 0.59 in that year, as Fig. 3.10 displays. Again, the average hour is computed on the basis of operating hours, which exceed five aircraft movements. Likewise, London Stansted reaches its maximum runway capacity (almost) during peak hours (30th peak hour volume 44 aircraft movements, 5% peak hour volume 40 aircraft movements), but aircraft movements per hour decline thereafter rather rapidly. As a comparison, Fig. 3.11 shows Hanover airport, which has two principal runways and a low degree of capacity utilisation and thus ample capacity 30th peak hour volume: 44

50

366th hour or 5% peak hour volume: 40

Aircraft movements per hour

45 40

CUI: 0.59

35 30

Average hour volume: 24

25 20 15 10 5 0

1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570 600 630 660 690 720 750 780 810 840 870

Hours of the year

FIGURE 3.10 CUI and traffic ranking curve for London Stansted [Official Airline Guide (OAG), 2016]. CUI, capacity utilisation index.

52

PART | I Basic concepts 30th peak hour volume: 15

Aircraft movements per hour

20

406th or 5% peak hour volume: 12

18 16

CUI: 0.50

14 12 10 8

Average hour volume: 6

6 4 2 0

1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570 600 630 660 690 720 750 780 810 840 870

Hours of the year

FIGURE 3.11 CUI and traffic ranking curve for Hanover airport [Official Airline Guide (OAG), 2016]. CUI, capacity utilisation index.

reserves. Hanover handled about 49,500 aircraft movements in 2016. Therefore the ranking curve is more inclined compared to the London Heathrow case; however, it is similar to the ranking curve of London Stansted. The ratio of average hour to 5% peak hour volume is 6/12 5 0.50; the CUI is thus relatively high. However, Hanover airport does not reach its runway capacity limit in the highest peak hour (19 aircraft movements) and the ranking curve declines thereafter rather rapidly. For airports of the size of Hanover and smaller, the average hour is computed on the basis of all operating hours of the year, since hours with five aircraft movements and less form an important part in these cases. Compared to London Heathrow and London Stansted, Hanover’s CUI value of 0.50 has a different meaning and is of a more theoretical nature, because the 5% peak hour volume is well below actual runway capacity. This implies that the ratio of the average hour to the 5% peak hour volume does not really mirror a capacity utilisation rate. In fact, the true capacity utilisation of Hanover is lower. Airports, such as the Hanover case, are of less importance in CUI analyses, because the focus here is on, at least potentially, capacity constrained airports. The 5% peak hour volumes of these airports typically reach values near practical capacity. By means of ranking curves, capacity utilisation can be derived for each airport in the global network; CUI values are thus comparable between airports of different size and with different runway systems. Therefore we have basically defined three categories of airports with regard to the characteristics of the ranking curve and the CUI. First, there are airports that reach (almost) the capacity of their runway system during peak hours and aircraft movements per hour decline slowly thereafter (e.g. London Heathrow and Frankfurt airport). Second, there are airports that (almost) reach the capacity

Capacity utilisation at airports worldwide Chapter | 3

53

of their runway system during peak hours as well, but aircraft movements decline rather rapidly thereafter (e.g. London Stansted). Third, there are airports that do not reach the capacity of their runway system during peak hours (e.g. Hanover). In many cases, a boundary value of around 100,000 (70,000200,000, depending on the runway system) aircraft movements per year at an airport serves as a guideline to separate airports of the first two categories from category 3 airports; however, this is only a rough guideline that works in most cases and is refined for specific analyses (Gelhausen et al., 2013). Fig. 3.12 shows the distribution of CUI values for the top 1000 airports worldwide as well as the portion of the total global flights, which each airport handles. The majority of these airports ( . 900) range in the lower left section of the graph, that is they have a low CUI and serve only a small share of the total global flight volume. Another cluster of airports is located in the upper right section of Fig. 3.12. Member airports of this cluster have a high CUI value and they handle a notably larger share of the total global flight volume. Almost all of these airports are hubs, which are important for the global air traffic network. It is difficult to define a theoretically exact discrimination value for the CUI to separate congested airports from those which have ample capacity reserves. Moreover, congestion is a rather sneaky process, which gradually increases with traffic volumes. However, from empirical observations, we suggest that values in a range of about 0.650.70 serve as an indicator of significant congestion problems, if the corresponding

Airport share of total global flight volume

1.4% ATL

ORD

1.2%

1.0%

DFW LAX

DEN

0.8%

AMS

PVG FRA

PEK CLT

LHR

0.6%

0.4%

0.2%

0.0%

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Capacity utilisation index (Ratio of mean hourly flight volume to 5% peak hour flight volume)

FIGURE 3.12 Distribution of the CUI in relation to airports’ share of the total global flights in 2016 [Official Airline Guide (OAG), 2016]. CUI, capacity utilisation index.

54

PART | I Basic concepts

5% peak hour volumes reach near-capacity values. This is also supported by a comparison of Figs 3.8 and 3.12. Here, we find important airports, which have been operating at or near capacity, such as Frankfurt (FRA) and London Heathrow (LHR), and with high CUI values, such as Beijing (PEK) and Los Angeles (LAX). Airports with a CUI value of 0.65 or higher already cover about 35% of the total global flights, as global air traffic is concentrated on few airports with a high degree of capacity utilisation and correspondingly only small capacity reserves (if at all). Table 3.2 describes the development of the traffic situation and the CUI values over time at four example airports. Here, the years 2000, 2008 and 2016 have been chosen to give a brief overview. London Heathrow and Atlanta Hartsfield-Jackson airports are good examples of airports with consistently high traffic volumes and high CUI values. As annual and 5% PHTV increase, so does the CUI value. However, there are no major increases at

TABLE 3.2 Traffic development at selected airports between 2000 and 2016 [Official Airline Guide (OAG), 2016]. Annual aircraft movements

5% PHTV

CUI

2000 Atlanta Hartsfield-Jackson (ATL)

881,738

179

0.66

Du¨sseldorf (DUS)

168,414

36

0.75

London Heathrow (LHR)

464,342

85

0.83

Beijing (PEK)

169,700

41

0.72

200

0.67

2008 Atlanta Hartsfield-Jackson (ATL)

956,380

Du¨sseldorf (DUS)

208,721

47

0.69

London Heathrow (LHR)

479,294

86

0.85

Beijing (PEK)

430,154

81

0.67

2016 Atlanta Hartsfield-Jackson (ATL)

875,211

184

0.66

Du¨sseldorf (DUS)

206,006

46

0.70

London Heathrow (LHR)

476,226

87

0.82

Beijing (PEK)

606,105

88

0.79

Capacity utilisation at airports worldwide Chapter | 3

55

these airports as they are already near or at their capacity limits and have been for some time. In contrast, traffic volumes at Beijing airport have more than tripled from 2000 to 2016 and 5% peak hour volumes have gone up from 41 to 88 movements. However, CUI values have decreased between 2000 and 2008, because a third runway was opened late in 2007, which eased the capacity situation at Beijing airport somewhat. They then increased sharply between 2008 and 2016 because capacity stayed at the same level. London Heathrow and Beijing were, in fact, in 2016 the two major airports worldwide with the highest capacity utilisation, as expressed by the CUI value of over 0.79. As a further example, Du¨sseldorf is an airport with two dependent runways and a declared capacity well below its technical capacity due to a noise emission regulation, which limits the capacity to the level of one runway. In 2000 the coordinated capacity of Du¨sseldorf was in a range of 3438 aircraft movements per hour. By the end of 2005, capacity was increased to 45 movements per hour. Table 3.2 shows a 5% peak hour volume of 36 and a CUI value of 0.75 for 2000. In 2008 the 5% peak hour volume was 47 movements, but the CUI value had dropped from 0.75 to 0.69, although the traffic grew from more than 168,000 to almost 209,000 aircraft movements. Demand grew rapidly in those years and could be satisfied in due course by the capacity rise. After 2008, traffic remained stable; since the capacity did not grow further due to the existing regulation, the CUI went up only slightly to a level of 0.7. Thus the CUI enables us to give a correct description of the capacity situation development at Du¨sseldorf.

3.5 Case study: development of capacity utilisation at three example airports: San Diego (SAN), London Heathrow (LHR) and Beijing (PEK) The three airports have been selected for a comparison of the traffic and capacity utilisation development because they differ in their traffic volumes, structures and developments and their capacities, but not much in the degree of capacity utilisation as expressed by the CUI. London Heathrow is one of the most important hubs in the global network with primarily international traffic, of which the touristic route to Spain and the intercontinental route to the United States surpassed all other routes. The volume of aircraft movements at LHR has not grown for years (see Fig. 3.13) because traffic has reached its capacity. All more or less usable coordinated airport slots are used by incumbent airlines, which operate their slots with grandfather rights allocated to them years ago. Beijing airport is a major international hub as well, however, with rapidly growing traffic, the volume of which has surpassed the LHR volume since 2010 (see Fig. 3.13). The hub function is more pronounced due to the high portion of domestic OD and feeder traffic. In contrast to LHR and PEK, San Diego is a single runway airport, which,

56

PART | I Basic concepts 700,000

600,000

Aircraft movements

500,000

400,000

300,000

200,000

100,000

0

2000

2008 LHR

SAN

2016

PEK

FIGURE 3.13 Air traffic development at the three example airports San Diego (SAN), London Heathrow (LHR) and Beijing (PEK) [Official Airline Guide (OAG), 2016].

nevertheless, belongs to the top 30 airports in the United States with its traffic volume of about 173,000 aircraft movements. Most of these flights have domestic destinations, quite in contrast to LHR. As with LHR, traffic has not grown since the year 2000 in San Diego. However, this is also due to general air traffic saturation in the United States. While the US passenger demand has grown slightly in the decade up to 2016, by 16% (1.5% per year) on average, the flight volume has decreased in the same period by 213% (21.4% per year on average). Both the different traffic growth patterns and the different capacity utilisation at the three airports are also visualised by the development of the hourly traffic variation over the day. Figs 3.143.16 display the daily traffic patterns at SAN, LHR and PEK airports for the peak day of the years 2000, 2008 and 2016. As can be seen in Fig. 3.14, the capacity of San Diego airport is fairly well used over the daytime hours with volumes varying between 25 and 45 movements per hour. The hourly runway capacity has been estimated by the FAA to be 4857 movements, 48 in IFR and 57 in optimum conditions [Federal Aviation Administration (FAA), 2014]. Since the traffic volume has not grown from 2000 to 2016, the daily pattern did not change either. The hourly traffic distribution over the peak day at LHR shows hardly any difference between 2000, 2008 and 2016 (see Fig. 3.15). With the exception at noon time in 2000, the volumes vary between 80 and 90 movements during daytime hours and reach the declared capacity of up to 88 movements

Capacity utilisation at airports worldwide Chapter | 3

57

50

Aircraft movements per hour

45 40 35 30 25 20 15 10 5 0

0

1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 Hours of the day

SAN 2000

SAN 2008

SAN 2016

FIGURE 3.14 Development of hourly traffic pattern at the peak day of the years 2000, 2008 and 2016 at San Diego (SAN) airport [Official Airline Guide (OAG), 2016].

100

Aircraft movements per hour

90 80 70 60 50 40 30 20 10 0

0

1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 Hours of the day

LHR 2000

LHR 2008

LHR 2016

FIGURE 3.15 Development of hourly traffic pattern at the peak day of the years 2000, 2008 and 2016 at London Heathrow (LHR) airport [Official Airline Guide (OAG), 2016].

to a high degree. As previously stated, LHR is one of the airports with the highest capacity utilisation. In contrast to LHR, Beijing (PEK) airport shows a development of the hourly traffic distribution, which differs from year to year, due to the rapid traffic growth at this airport (see Fig. 3.16). Traffic has increased both during

58

PART | I Basic concepts

day and night hours and the variation of traffic during the day has decreased with growing traffic from 2000 to 2016, indicating a high capacity utilisation in 2016. Traffic ranking functions allow us to visualise the capacity situation of airports. In this case study, we were interested to see how the traffic ranking curves of the three selected airports have developed in relation to the capacity available and the traffic change over time. Figs 3.173.19 show the traffic 100

Aircraft movements per hour

90 80 70 60 50 40 30 20 10 0

0

1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 Hours of the day

PEK 2000

PEK 2008

PEK 2016

FIGURE 3.16 Development of hourly traffic pattern at the peak day of the years 2000, 2008 and 2016 at Beijing (PEK) airport [Official Airline Guide (OAG), 2016].

Aircraft movements per hour

60 50 40 30 20 10 0

1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570 600 630 660 690 720 750 780 810 840 870

Hours of the year SAN 2000

SAN 2008

SAN 2016

FIGURE 3.17 Traffic volume ranking curves of San Diego (SAN) airport in the years 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016].

Capacity utilisation at airports worldwide Chapter | 3

59

Aircraft movements per hour

100 90 80 70 60 50 40 30 20 10 0

1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570 600 630 660 690 720 750 780 810 840 870

Hours of the year LHR 2000

LHR 2008

LHR 2016

FIGURE 3.18 Traffic volume ranking curves of London Heathrow (LHR) airport in the years 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016].

Aircraft movements per hour

100 90 80 70 60 50 40 30 20 10 0

1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570 600 630 660 690 720 750 780 810 840 870

Hours of the year

PEK 2000

PEK 2008

PEK 2016

FIGURE 3.19 Traffic volume ranking curves of Beijing (PEK) airport in the years 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016].

ranking curves of San Diego (SAN), London Heathrow (LHR) and Beijing (PEK) airport for the years 2000, 2008 and 2016. The SAN traffic volume curves (see Fig. 3.17) differ only a little in their slope and position, indicating that both the traffic volume and the capacity situation in 2000, 2008 and 2016 did not vary a lot. Traffic grew slightly from around 190,000 aircraft movements in 2000 to almost 198,000 in 2008, and the volume ranking curve of 2000 moved up accordingly, in particular over many daytime hours. From

60

PART | I Basic concepts

2008 to 2016, traffic dropped to over 173,000 movements and, as can be seen, the ranking curve moved downwards below the 2000 curve. The top hourly traffic volume reached 45 aircraft movements and the 5% peak hour volume 40 movements in 2000. In 2008, these values went up to 50 and 42 and they decreased again to 47 and 37 movements in 2016. The slope of the curves is rather uniform over a range of about 6000 hours before hourly traffic of around 20 movements sharply drops to just a few movements during night hours. As already demonstrated by the hourly distribution of the peak day traffic at London Heathrow (LHR, see Fig. 3.15), the airport is operating at the capacity limit over almost all hours of the day. The high capacity utilisation is mirrored by the traffic ranking curves (see Fig. 3.18). There is hardly any difference in the shape and position of the three curves of the years 2000, 2008 and 2016 due to the invariance of capacity and traffic. As can be seen, the slope of the curves is rather flat in the high-volume range of 90 to about 70 aircraft movements, that is over many daytime hours until about 5000 hours, before volumes go down rapidly to zero movements. As in San Diego, LHR has no air traffic in about 1700 night hours. Except for Beijing airport in 2016, no other airport exhibits traffic ranking curves of such extreme ‘belly shape’, indicating the high capacity utilisation of the two parallel runways with a declared capacity of up to 88 movements in peak hours. Beijing (PEK) airport has experienced a totally different development of both traffic and capacity. Traffic has grown from around 170,000 aircraft movements to over 430,000 in 2008 and more than 606,000 movements in 2016, thus with an average growth rate of 8.3% per year. Until late in 2007, PEK handled the traffic on two parallel runways and then increased the capacity by an additional runway, which can be operated independently from the others. As can be seen in Fig. 3.19, the three traffic ranking curves differ greatly in their position and form. In 2000, hourly traffic did not exceed 50 movements and in most hours varied between 40 and 20 movements. In around 2000 hours, mainly at night, the airport did not have any aircraft movements. Traffic more than doubled in 2008 and the traffic volume function moved up accordingly, as top hourly volumes of 80 and more flights were handled on three independent runways. At the same time, the number of operating hours increased from about 6600 to around 8700 hours, thus encompassing all available hours of the year. The long middle part of the volume ranking curve is still inclined so that the capacity in these mainly daytime hours is not fully utilised. This changed in the following period to the year 2016. In that year the airport operated at or near capacity over many hours, as can be seen in the corresponding volume ranking curve. The slope is barely inclined over a time span of about 6000 hours at a mean level of around 85 movements, only then volumes drop from around 70 movements to rather small volumes, especially after around 6600 hours. The curve has thus a similar shape as the LHR curve with the exception that PEK operates more hours with high traffic volumes.

Capacity utilisation at airports worldwide Chapter | 3

61

0.9 0.8 0.7

CUI

0.6 0.5 0.4 0.3 0.2 0.1 0

2000

2008 LHR

SAN

2016 PEK

FIGURE 3.20 Development of CUI of the three selected airports San Diego (SAN), London Heathrow (LHR) and Beijing (PEK) [Official Airline Guide (OAG), 2016]. CUI, capacity utilisation index.

The three example airports differ in traffic volume, structure and development and in capacity; however, they share a high degree of capacity utilisation. For airports with traffic volumes at or near capacity during peak hours, the CUI is an indicator well suited to describe the degree of capacity utilisation. In Fig. 3.20 the CUI values of the three airports are displayed for the years 2000, 2008 and 2016. As can be seen, the airports have been highly utilised over the period from 2000 to 2016. The CUI was the highest in LHR with values of around 0.82, followed by SAN with values of around 0.71. PEK airport had a CUI of 0.72 in 2000, which went down slightly in 2008 to 0.67 in spite of a strong increase in traffic. The additional capacity through the third runway, however, caused the degree of utilisation to go down. In the following years, traffic continued to grow rapidly with no further addition of runway capacity so that the CUI went up to 0.79 in 2016. Since a CUI of around 0.85 can be regarded as a saturation value, a further increase of the capacity utilisation at PEK will not be feasible for a long time, given the strong demand growth. A new airport has been realised in late 2019 in the south of Beijing, which will take over the additional traffic.

3.6

Conclusion

In this chapter, we have mainly derived two methodological tools for estimating and evaluating airport capacity utilisation: traffic ranking curves and

62

PART | I Basic concepts

the CUI. Capacity utilisation of airports is regarded as a prime indicator of capacity reserves, or more importantly for high-volume airports, of a lack of capacity reserves. For assessing airport capacity reserves, the CUI has been defined as the ratio of the average hour traffic volume and the 5% PHTV. The hour has been chosen as a suitable time unit for measuring the capacity since the maximum throughput of an infrastructure like a runway can be achieved by the demand in real-life conditions only in short-time intervals. This is not the case for a whole year because of many low or zero demand periods at night or weekends. Traffic ranking functions have been developed to visualise the capacity situation at airports and to determine specific volumes at certain hours like the 5% peak hour volume. The functional form tells about the degree of capacity utilisation; the less inclined the slope of the curve over long periods of time, the closer is the airport reaching capacity in these periods, if and only if the top-ranking hourly volumes reach levels at or near capacity. If traffic ranking curves are belly shaped, as for instance those of London Heathrow or Frankfurt, then the 5% peak hour volumes are indicative of the practical capacity of the airport, thus allowing identification of airports with a high capacity utilisation. By means of the CUI, we are in a position to categorise airports according to their importance within the global air traffic system and to the level of capacity reserve or of capacity constraint, respectively. Thereby, we can identify those airports that are critical for the future growth of traffic, that is those airports which serve a large share of the global flight volume and, at the same time, suffer from capacity bottlenecks. The absolute slope of the traffic ranking function serves as a measure for capacity utilisation; the CUI as the ratio of the average hour traffic volume and 5% peak hour traffic volume is an indicator of that slope. For London Heathrow, for instance, the airport has been operating at capacity for many years, the CUI of the year 2016 has been calculated at 0.82. In contrast, London Stansted, an airport with capacity reserves, has a CUI of only 0.59. Traffic volumes reach capacity only in a few peak hours per year. Capacity utilisation indices have been developed on the basis of traffic ranking curves for each airport in the global network. The CUI values are thus comparable between airports of different size and with different runway systems. Based on traffic ranking curves, CUI values and annual traffic volumes  airports with small traffic volumes have been excluded since they will not face capacity problems in the near future  airports have been categorised according to their degree of capacity constraint. This will be the subject of Chapter 4, Constrained and under-utilised airports.

Capacity utilisation at airports worldwide Chapter | 3

63

References Ashford, N., Mumayiz, S., Wright, P., 2011. Airport Engineering, Planning, Design, and Development of 21st Century Airports. John Wiley & Sons, Hoboken. Brueckner, J.K., Zhang, Y., 2001. A model of scheduling in airline networks: How a hub-andspoke system affects flight frequency, fares and welfare. J. Transp. Econ. Policy 35 (2), 195222. Digital Aeronautical Flight Information Files (DAFIF), 2008. US Government, Washington D.C. Federal Aviation Administration (FAA), 1969. Airport Capacity Criteria used in Long-Range Planning, FAA Advisory Circular AC 150/5060-3A. Washington, DC. Federal Aviation Administration (FAA), 1983. Airport Capacity and Delay, FAA Advisory Circular AC 150/5060-5. Washington, DC. Federal Aviation Administration (FAA), 2014. Airport Capacity Profiles. Washington, DC. Gelhausen, M.C., Berster, P., Wilken, D., 2013. Do airport capacity constraints have a serious impact on the future development of air traffic? J. Air Transp. Manage. 28, 313. Givoni, M., Rietveld, P., 2009. Airline’s choice of aircraft size  explanations and implications. J. Transp. Res., A 43 (5), 500510. Horonjeff, R., McKelvey, F., Sproule, W., Young, S., 2010. Planning & Design of Airports. McGraw Hill, New York. Matthews, L., 1995. Forecasting peak passenger flows at airports. Transportation 22 (1), 5572. Official Airline Guide (OAG), 2008. Market Analysis. Reed Travel Group, Dunstable. Official Airline Guide (OAG), 2016. Market Analysis. Reed Travel Group, Dunstable. Official Airline Guide (OAG), 2017. Market Analysis. Reed Travel Group, Dunstable. Poole, M., 1972. Forecasting air traffic in peak periods. In: Paper Presented for Western European Airports Association, British Airports Authority, London. Pitfield, D.E., Caves, R.E., Quddus, M.A., 2010. Airline strategies for aircraft size and airline frequency with changing demand and competition: a simultaneous-equations approach for traffic on the north Atlantic. J. Air Transp. Manage. 16 (3), 151158. Urbatzka, E., 1991. A Method for Estimating Runway Capacity of German Airports for Single Runway Systems. German Aerospace Center (DLR), Cologne. Urbatzka, E., 1993. A Method for Estimating Runway Capacity  Analysis of Data of Hamburg Airport. German Aerospace Center (DLR), Cologne. Urbatzka, E., Wilken, D., 1997. Estimating runway capacities of German airports. J. Transp. Plann. Technol. 20 (2), 103129. Wang, P.T., Pitfield, D.E., 1999. The derivation and analysis of the passenger peak hour: an empirical application to Brazil. J. Air Transp. Manage. 5 (3), 135141. Weigelt, G., 1994. Ableitung eines funktionalen Zusammenhangs zwischen ja¨hrlichen und stu¨ndlichen Flugbewegungsaufkommen von Flugha¨fen. German Aerospace Center (DLR), Cologne. Wilken, D., Berster, P., Gelhausen, M.C., 2011. New empirical evidence on airport capacity utilisation: relationships between hourly and annual air traffic volumes. Res. Transp. Bus. Manage. 1, 118127.

Chapter 4

Constrained and under-utilised airports It is a well-known fact that traffic, regardless of the mode, is neither equally distributed over the network nor over the time of the day, week, or year. The spatial distribution of land uses and the temporal distribution of human activities and movements between different land uses cause concentrations of traffic flows on network sections and in time windows. Air traffic is no exception to this rule. In a former study on the global airport constraint situation (Gelhausen et al., 2013), it has been shown that in 2008 the cumulative distribution of air traffic in the global network of around 4000 airports was characterised by a high degree of concentration on a rather limited number of important airports, as was indicated by a Gini coefficient value of 0.8. While the mathematical details of how to compute airport capacities and forecast airport capacity enlargements is the subject of Chapter 7, Modelling future airport capacity and capacity utilisation, and Chapter 8, Modelling future airport capacity enlargements and limits, the research interest of this chapter is application-oriented. We describe the worldwide traffic distribution in detail, on the global scale as well as in world regions and selected countries. This includes statements on the great number of airports with rather small traffic volumes. Since we have to assume a correlation between traffic volumes and capacity utilisation, we analyse the degree of capacity utilisation by estimating airport capacities and volume-capacity ratios. As a result, we show that many of the high-volume airports are identical to those that operate at or near capacity. On the other hand, all airports with low traffic volumes have ample capacity reserves and would probably welcome additional traffic. In fact, the majority of airports worldwide are far from developing capacity problems in the near future. An interesting question is whether or not the degree of traffic concentration has changed in the past. With smaller and more cost-efficient commercial aircrafts in the market and more airports having been added to the network in some markets, one would assume that traffic has deconcentrated. As will be shown, this is the case on the global scale, however, only to a rather small degree. Air traffic remains very concentrated. Similar developments have occurred in world regions and countries, although with some variation. Airport Capacity Constraints and Strategies for Mitigation. DOI: https://doi.org/10.1016/B978-0-12-812657-8.00004-X © 2020 Elsevier Inc. All rights reserved.

65

66

PART | I Basic concepts

Air traffic, as described in the following chapter, is given by the number of flights or aircraft movements (take-offs and landings) at airports. All airports listed in the OAG data file which provides scheduled air services to other airports have been included in the analysis of traffic distribution and capacity utilisation. While it is obvious that the biggest airports at the top end, such as Chicago O’Hare and Atlanta Hartsfield-Jackson, are enlisted in the list of all airports, the network defined by OAG includes airports at the low end, which may only have a few regular services within a week or a year. Since there are only relatively few airports with high traffic volumes of say over 100,000 aircraft movements per year, there are many airports with rather low traffic volumes of some hundred or thousand movements, the average annual traffic volume in 2016 was only about 17,500 aircraft movements per airport. The total number of airports varies from year to year depending on the existence of scheduled services. In some countries, airport networks increase in size with growing demand, whereas in other countries, the number of airports served by regular flights decreases. In total, the number of airports with scheduled services slightly increased from 4035 in 2000 to 4054 in 2016, as can be seen in Table 4.1. In contrast, the traffic volume has grown from 27.9 million flights (corresponding to 55.8 million aircraft movements) in 2000 to 35.5 million flights in 2016. Since the number of airports stayed almost stable, the average traffic volume per airport has grown from about 6900 in 2000 to around 8700 flights in 2016. The corresponding average daily traffic volume has thus increased from just 19 to 24 flights. As has been mentioned, the analysis of traffic distribution and concentration has been carried out for major world regions as well as for some countries. The world regions are (definitions can be found in the appendix) G G G G G G G

Africa, Asia, Europe, Middle East, North America, South America and Southwest Pacific.

The selected countries are two big states: China, which represents a growing air transport market, and the United States, which is a huge market with more or less saturated demand, as well as two European states with a centralised and a less centralised network, namely France and Germany. The number of airports in each region with the corresponding traffic volume in 2016 is shown in Tables 4.2 and 4.3. The North American market, dominated by the US market, is still the most important air traffic region among the listed world regions and counts

Constrained and under-utilised airports Chapter | 4

67

TABLE 4.1 Development of the number of airports and take-offs [Official Airline Guide (OAG), 2016]. Year

Number of airports

Number of take-offs

2000

4035

27,938,423

2001

3962

27,662,485

2002

3911

26,514,658

2003

3826

26,278,155

2004

3826

27,589,145

2005

3816

28,581,862

2006

3786

28,899,965

2007

3782

30,145,403

2008

3790

30,274,483

2009

3834

29,189,147

2010

3849

30,577,632

2011

3882

31,596,311

2012

3810

31,847,041

2013

3936

32,220,615

2014

3944

33,002,972

2015

3962

34,027,337

2016

4054

35,454,054

more than 1000 airports, which handled in 2016 over 10 million flights, equal to 20 million aircraft movements. The Asian market follows with 921 airports with regular services and almost as many flights as in North America. Europe is in third place with 685 airports, which handled 8.2 million flights. Of the total air traffic of 35.5 million flights, 80% is concentrated in these three regions, which have, however, just 65% (2624) of all airports worldwide. While the total number of airports with regular services has changed only slightly from 2000 to 2016, the airport networks in world regions have developed in different ways. In North America, the number of airports has decreased from 1069 in 2000 to 1018 in 2016, whereas in Asia, networks have grown strongly from 650 to 921 airports. In the Middle East, too, airport density has grown from 99 to 111 airports. In Europe, the number of airports has practically not changed in that time span of 16 years. In

68

PART | I Basic concepts

TABLE 4.2 Number of airports and flights by world region [Official Airline Guide (OAG), 2016]. World region

Number of airports

Number of take-offs

Africa

388

1,206,038

Asia

921

10,149,228

Europe

685

8,234,691

Middle East

111

1,239,500

North America

1018

10,181,932

South America

530

3,312,908

Southwest Pacific

401

1,129,757

World

4054

35,454,054

TABLE 4.3 Number of airports and flights for selected countries [Official Airline Guide (OAG), 2016]. Country

Number of airports

Number of take-offs

China

213

3,975,428

France

60

726,677

Germany

34

937,459

United States

713

9,042,414

the other regions of Africa, South America and the Southwest Pacific network density has gone down.

4.1

Overview: point-to-point versus hub networks

Pure hub-and-spoke (HS) and point-to-point (P2P) systems are two contrary systems of flight network organisation; however, in many real cases, there are hybrid systems that are a mixture of both the systems. A true HS system needs a smaller number of flights to connect each node compared with a pure P2P system, thus leading to a more efficient system in terms of resources employed (economics of density). Thus an HS system can generate more routes and frequencies compared with a pure P2P system because of the higher traffic volumes and the high share of transfer passengers (Redondi et al., 2011, 2012). This is particularly true for intercontinental destinations,

Constrained and under-utilised airports Chapter | 4

69

where many origin destination (OD) markets are less dense and cannot be viably served at a reasonable frequency; see, for example, Wilken et al. (2016). Therefore local passengers benefit from a hub system with high transfer passenger volumes as well, even if they take a nonstop flight. But how did these two systems and their hybrids evolve over time? A main driving force developing these systems has been airline deregulation, which fostered both airline alliances and low-cost carriers (LCCs). Airline deregulation started more than 40 years ago when the US Airline Deregulation Act was approved by the US Congress (e.g. Goetz and Vowles, 2009), and more countries have followed in subsequent years. For example, deregulation began in Australia in the early 1980s (e.g. Hooper, 2005) and in Europe in the late 1980s (e.g. Button et al., 1998; Ehmer et al., 2000). Deregulation pushed HS systems, which were more spatially, but in particular more temporally, concentrated on coordinated incoming and outgoing flights (wave system). Before deregulation, airline networks were already spatially concentrated because of the system of bilateral agreements and national carriers. However, these star-shaped networks were not temporally coordinated and, in particular, international transfer opportunities existed rather by accident (Burghouwt and de Wit, 2005). Nevertheless, deregulation led to intense competition between the main hubs, as typically a few hubs compete for the same OD markets. In this regard, European hubs have a geographical advantage compared with American and Asian hubs relative to the main markets, but this might change in the future with the emerging markets in East Asia and the rise of the Gulf carriers (Redondi et al., 2011). Deregulation led a number of legacy carriers in the United States and Europe to switch from P2P systems to temporally coordinated HS systems (Burghouwt and de Wit, 2005). However, as already mentioned, European networks were already spatially concentrated, caused by bilateral agreements and national flag carriers being tied to their home base. Temporally coordinated HS systems have a number of cost- and demand-side advantages (e.g. Button, 2002) that airlines tried to realise by reorganising their network more towards a HS model. However, deregulation also gave rise to LCCs entering and developing new markets, mainly operating in P2P networks with sufficient OD demand volume to make nonstop flights viable (e.g. Berster et al., 2012; ReynoldsFeighan, 2001; Williams, 2001). In fact, the strategy of hub concentration of legacy carriers opened the door for LCCs (Dobruszkes, 2006). These markets are mainly in short- and medium-haul travel, for example, within Europe or between Europe and North Africa, which are rather dense or where sufficient demand can be generated by direct services and low fares. While there is still an ongoing discussion as to whether the LCC model is transferable to longhaul routes (see, e.g. Daft and Albers, 2012), hubbing makes many markets viable that otherwise would have been less developed. Demand for longdistance air travel is less price elastic than demand for short-haul and

70

PART | I Basic concepts

medium-haul air travel, so that the potential for demand generation is limited and dense markets or feeder traffic is needed (Morrell, 2008). Pels (2008) concluded that a sufficient level of demand is mandatory for long-haul lowcost services. Yet in many cases, LCCs would compete on such routes with well-established hub carriers, and they cannot play fully to their strengths as operational efficiency, as a major competitive advantage, is less important in long-haul travel than in short- or medium-haul markets, due to longer flights and therefore fewer turnarounds (Wensveen and Leick, 2009). Estimates of the cost advantage of LCC in long-haul travel ranges from up to 10% (Moreira et al., 2011) to 60% at most (Morrell, 2008). A random connecting system or ‘self-help hubbing’ (Malighetti et al., 2008) may generate some transfer passengers, although far fewer than at a fully-fledged hub. For example, the share of transfer passengers of Ryanair at London Stansted was 17.2% (O’Connell and Williams, 2005). However, compared with services of legacy carriers, self-help hubbing is typically less convenient for passengers as the costs of the search for connecting flights, baggage handling during stopover, and the risk of missed connections are entirely borne by the passenger (Grimme, 2011). As we can see, hubbing is necessary for airlines to serve thin markets, in particular in long-haul travel, at a relatively high frequency. Furthermore, larger aircrafts are typically employed in intercontinental long-haul markets for technical and economic reasons, which intensifies the problem of maintaining attractive flight schedules on thin routes and makes hubbing necessary (Wilken et al., 2016). One of the negative effects of increased hubbing is airport capacity constraints, particularly during peak hours. While only a relatively small number of about ten airports were heavily capacity constrained, that is during most daytime hours, in 2008 (the year before the worldwide economic downturn), there was a fairly high degree of average capacity utilisation among the largest 100 airports, which accounts for about 45% of flights worldwide (Gelhausen et al., 2013). Key measures to mitigate capacity constraints at airports include increasing capacity by adding new runways, reorganisation of traffic either to off-peak times or less congested airports, and using aircrafts with higher seat capacity (see Chapter 3: Capacity utilisation at airports worldwide).

4.2

Development of global air traffic distribution over time

By ranking airports by traffic volume, we can show the traffic distribution over all airports, both in a direct and in a cumulative way. In Fig. 4.1, the distribution of the global air traffic is shown for the year 2016. For reasons of comparability between charts the number of flights (vertical axis) and of airports (horizontal axis) are shown as shares of the total number. Each airport in the global network accounts therefore for 0.0257% of all airports. The airport Dallas/Fort Worth (DFW) had, for example, a traffic volume of

Constrained and under-utilised airports Chapter | 4

71

Share of global flights (n=35.5 million, 2016)

1.4%

1.2%

ATL ORD

1.0%

0.8%

0.6%

DFW LAX PEK DEN CLT LHR PVG AMS

0.4%

0.2%

0.0% 0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Share of ariports in the world (n=4054)

FIGURE 4.1 Air traffic distribution in the global airport network 2016 [Official Airline Guide (OAG), 2016].

over 323,000 flights in 2016, which corresponds to roughly 0.9% of the total flight volume of 35.5 million flights worldwide. The average traffic volume per airport in 2016 is about 8700 flights, which corresponds to 0.0245% of the total flight volume. The biggest airport in terms of aircraft movements is Atlanta Hartsfield-Jackson (ATL) with almost 438,000 flights, which is 50 times the average airport traffic. The distribution, as shown in Fig. 4.1, shows that only about 15% of all airports have higher than average volumes, which also means that 85% of all airports worldwide have traffic volumes below 8700 flights per year. Traffic is thus not equally distributed over all airports but rather concentrated on a relatively small number of airports. The concentration of traffic becomes even more evident if one refers to the cumulative distribution, which is shown in Fig. 4.2 (which is similar to Fig. 3.9). The two axes have the same dimensions as in Fig. 4.1. As can be seen, traffic share increases sharply over just a small share of all airports; 50% of total traffic is handled by only 3% and 90% is handled by 23% of all airports. In other words, the biggest 120 airports (corresponding to about 3%) handle half of the total traffic, that is 17.75 million flights, while the other 3935 airports handle the same volume of traffic, each one, on average 4500 flights per year. Furthermore, the biggest 930 airports (23%) handle almost 32 million flights, while the other 3125 airports handle just 3.4 million flights, on average thus 1100 flights per year. Converted to a daily volume, this would mean not more than three flights per day. The largest four airports in 2016 were located in the United

72

PART | I Basic concepts

Share of global flights (n=35.5 million, 2016)

100% 90% 80%

70% 60% 50% 40% 30% 20% 10% 0%

0%

10%

20%

30%

40% 50% 60% Share of airports (n=4054)

70%

80%

90%

100%

FIGURE 4.2 Cumulative distribution of air traffic in the global airport network in 2016 [Official Airline Guide (OAG), 2016].

States [Chicago O’Hare (ORD), Atlanta Hartsfield-Jackson (ATL), Dallas/ Fort Worth (DFW) and Los Angeles (LAX)], followed by Beijing (PEK) airport. The largest European airport is London Heathrow (LHR) in the eighth place. The global airport network consists thus of a small number of airports with high traffic volumes, while the great majority of airports handle traffic volumes in the order of just a few flights a day. This latter group of airports does not have the same capacity problems as airports in the first group may have to struggle with; their main concern is often to attract more traffic to the airport in order to cover the cost of operations. For the years 2000, 2008 and 2016, we have analysed the traffic distribution globally and in networks of world regions and some selected countries and have confirmed the result of 2016, a high degree of traffic concentration on a relatively small number of important airports, often hub airports, and have identified a great number of airports with rather small traffic volumes well below the capacity limit. For the three years, we have carried out distribution analysis and measured concentration indicators, such as the Gini coefficient and the share of airports, which handle 50% and 90%, respectively, of the total traffic volume. Fig. 4.3 shows the temporal development of the Lorenz curve of global take-offs for the years 2000, 2008 and 2016. On a global scale, traffic concentration remained more or less the same. If we limit the sample to airports that have at least 1000 take-offs per year, concentration of traffic at a fairly small number of airports is still very high, as illustrated in Fig. 4.4. The number of airports that have at least 1000

Constrained and under-utilised airports Chapter | 4

73

100%

Share of aircraft movements

90% 80% 70%

60% 50% 40%

30% 20% 10%

0%

0%

10%

20%

30%

40% 50% 60% Share of airports 2000

2008

70%

80%

90%

100%

2016

FIGURE 4.3 Lorenz curve for aircraft movements at airports worldwide in 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016]. 100% 90%

Share of aircraft movements

80% 70% 60%

50% 40% 30%

20% 10% 0%

0%

10%

20%

30%

40% 50% 60% Share of airports 2000

2008

70%

80%

90%

100%

2016

FIGURE 4.4 Lorenz curve for aircraft movements at airports worldwide with at least 1000 take-offs per year in 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016].

flights per year is 2087, 2050 and 2222 in 2000, 2008 and 2016. While traffic concentration increased only very little between 2000 and 2016, the total number of airports with at least 1000 take-offs per year increased by about 6%. However, because both Figs 4.2 and 4.3 basically produce qualitatively similar results, we retained the full airport sample for further analyses.

74

PART | I Basic concepts

TABLE 4.4 Development of the global Gini coefficient from 2000 to 2016 [Official Airline Guide (OAG), 2016]. Year

Share of airport with 50% of traffic (%)

Share of airport with 90% of traffic (%)

Gini coefficient

2000

2.6

25.2

0.8320

2008

3.0

24.2

0.8306

2016

3.1

23.3

0.8354

As can be seen, the distribution pattern hardly changed over time. The global air traffic was concentrated in 2000 and has been concentrated since then. In 2000, the global airport network with scheduled services consisted of 4035 airports, and in 2016, the network size was nearly the same with 4054 airports; however, traffic grew from 27.9 to 35.5 million flights. Nevertheless, the share of airports with a 50% and 90% traffic share as indicators of traffic concentration changed only slightly, as can be seen in Table 4.4. The Gini coefficient had values of around 0.83 in all years and thus stands for a highly concentrated distribution of traffic. If we look at the share of airports which handle 50% of the total traffic we see a slight decrease of concentration; in 2000 only 2.6% of all airports handled half of the total traffic, 16 years later this share increased to 3.1%. On the other hand, the share of airports which handled 90% of total traffic decreased slightly from 25.2% to 23.3%, thus indicating a marginal rise of concentration. But before we turn to regional air traffic distributions in the next section, we take a look at global air traffic volume growth between 2000 and 2016, as illustrated by Table 4.1. Fig. 4.5 shows the traffic growth from the year 2000 to 2008 and 2016. The number of global take-offs at airports increased from nearly 28 million in 2000 to 30 million in 2008 and to more than 35 million take-offs in 2016. In 2009, the year of the global financial crisis, the number of global take-offs dropped to 29 million but continued to grow thereafter. The compound annual growth rate (CAGR) of flights worldwide for the period from 2000 to 2016 is 1.5%, thus flights grew on average by 1.5% per year. However, if we split the period into two sub-periods, 2000 08 and 2008 16, we detect a rapid increase of the CAGR. While the CAGR was only 1.0% for the period from 2000 to 2008, it doubled to 2.0% in the period from 2008 to 2016.

4.3

Regional characteristics of air traffic distribution

If, on the global scale, traffic is spatially distributed in such a way that a relatively small number of airports with big flight volumes, in many instances

Constrained and under-utilised airports Chapter | 4

75

40,000,000

35,000,000

Number of global take-offs

30,000,000

25,000,000

20,000,000

15,000,000

10,000,000

5,000,000

0

2000

2008

2016

FIGURE 4.5 Growth of take-offs at airports worldwide from 2000 to 2016 [Official Airline Guide (OAG), 2016].

identical to hub airports, handle a great portion of the total traffic, can we assume similar concentration patterns in world regions? To examine this, the traffic distribution and cumulative distribution have been analysed region by region for the years 2000, 2008 and 2016. Fig. 4.6 illustrates how the traffic distribution between world regions in terms of take-offs evolved from 2000 to 2008 and 2016. North America was still one of the largest market in 2016, but its share of global take-offs declined throughout the period from 48% to 29%. This was mainly due to a stagnating US market. On the other hand, the share of the Asian market more than doubled. In 2000, 11% of the traffic was handled at Asian airports, but their share increased to 19% in 2008 and 29% in 2016. The Asian market grew thus to about the same size as the North American market. The third largest market in 2016 was Europe, and its share fluctuated between values of 23% (2000) and 26% (2008). The remaining four regions had a fairly small share of global flights; however, the airports in Africa and, in particular, the Middle East showed a continued positive traffic growth. Table 4.5 illustrates the CAGR of take-offs by world region for the period from 2000 to 2016, as well as the sub-periods 2000 to 2008 and 2008 to 2016. While on the global scale we find accelerating growth with an average growth rate of 1% in the first period of 2000 08 and 2% in the second period to 2016, the growth pattern varies considerably in world regions. While growth rates are positive in four out of seven regions in the first period and remain positive in six regions in the second period, we find a

76

PART | I Basic concepts

Share of aircraft movements by world region

50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0%

Southwest Pacific

Africa

Europe 2000

Middle East 2008

Asia

North America South America

2016

FIGURE 4.6 Traffic distribution between world regions from 2000 to 2016 [Official Airline Guide (OAG), 2016].

TABLE 4.5 Air traffic growth (compound annual growth rate of take-offs) by world region for the period from 2000 to 2016 [Official Airline Guide (OAG), 2016]. World region

2000 08 (%)

2008 16 (%)

2000 16 (%)

Africa

4.25

3.65

3.95

Asia

7.21

7.53

7.37

Europe

2.36

0.73

1.54

Middle East

5.68

7.52

6.59

North America

21.68

21.49

21.59

South America

20.64

2.76

1.04

Southwest Pacific

21.17

0.55

20.31

World

1.01

1.99

1.50

negative trend in Europe and, in particular, in North America. Correspondingly, we find weak economic growth in North America and the European Union. While gross domestic product (GDP) grew by 2.40% and 2.35% per year on average during the period from 2000 to 2008, respectively, these rates fell to 1.72% and 0.73% per year for the period from 2008 to 2016 (The World Bank, 2016).

Constrained and under-utilised airports Chapter | 4

77

In summary, air traffic growth was strong in Africa, Asia and the Middle East from the year 2000 to 2008; however, growth rates have declined since 2008 in some regions, particularly in Europe. From our viewpoint, this is in part due to rather weak economic growth since the global financial crisis in 2008/09, as GDP is typically the main driver of air traffic development [e.g. International Air Transport Association (IATA), 2008]. In particular, welldeveloped markets such as North America and Europe suffer from restricted growth, while emerging markets such as Asia and the Middle East are still growing strongly, which again coincides with economic development. However, in the case of the US market a number of airline mergers due to economic problems also play an important role in explaining the traffic development there. In Table 4.6 the main indicators of traffic concentration in the world regions are summarised for the year 2016. These are the share of airports in the region which handle 50% and 90%, respectively, of the total traffic of that region. North America is the region with the highest degree of traffic concentration as measured by the Gini coefficient of 0.8668 (global value: 0.8354). A further indicator of concentration is the share of airports which handles a certain part of the total traffic, for instance 90%. With this measure, as well, North America has the highest concentration of traffic in airports. Only 17.8% of all airports, equal to 181 airports, handle 90% of the North

TABLE 4.6 Traffic concentration by world region [Official Airline Guide (OAG), 2016]. World region

Share of airports with 50% of region traffic (%)

Share of airports with 90% of region traffic (%)

Number of airports

Gini coefficient

Africa

4.6

34.0

388

0.7676

Asia

3.9

25.4

921

0.8168

Europe

4.7

26.9

685

0.7986

Middle East

5.4

22.5

111

0.8193

North America

2.2

17.8

1018

0.8668

South America

5.1

34.3

530

0.7630

Southwest Pacific

1.7

22.4

401

0.8584

World

3.1

23.3

4054

0.8354

Share of aircraft movements at North American airports

78

PART | I Basic concepts 100% 90% 80% 70% 60% 50% 40% 30% 20%

10% 0%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Share of North American airports 2000

2008

2016

FIGURE 4.7 Lorenz curve for aircraft movements at North American airports [Official Airline Guide (OAG), 2016].

American air traffic. Traffic in the other 837 airports accounts for just 10% of the total. In contrast, in South America, more than one-third of all airports (34.3%) have a 90% traffic share. In the following paragraphs, we will describe the seven world regions in more detail. Fig. 4.7 displays the Lorenz curve for North America. This is the region with the greatest number of airports; 1018 airports have been identified as airports with regularly scheduled services in 2016. Furthermore, the airports with the highest traffic volume in terms of flights are to be found in North America. The biggest airports worldwide are Atlanta Hartsfield-Jackson (ATL) with nearly 438,000 flights, followed by Chicago O’Hare (ORD) with almost 428,000 and Dallas/Fort Worth (DFW) with more than 323,000 flights. There are 27 airports with more than 100,000 flights, and 50% of the total traffic of North America (10.2 million flights) is handled by just 22 airports (2.2%). The Gini coefficient of the cumulative distribution has a value of 0.8668, which is the highest value of all regions. As a result, the North American region has both the highest number of airports and highest degree of concentration of any world region. Nevertheless, concentration still increased a little between the years 2000 and 2016 despite the fact that takeoffs decreased by 1.59% on average per year, as shown in Table 4.5. Fig. 4.8 shows the Lorenz curve for South America. This region has a network of 530 airports with scheduled services in 2016, which handle a traffic volume of 3.3 million flights. The three biggest airports are Mexico (MEX), Sao Paulo (GRU) and Bogota (BOG), which are the only airports in South America with volumes exceeding 100,000 flights in 2016. All other

Share of aircraft movements at South American airports

Constrained and under-utilised airports Chapter | 4

79

100% 90% 80% 70% 60% 50% 40% 30% 20%

10% 0%

0%

10%

20%

30% 40% 50% 60% 70% Share of South American airports 2000 2008 2016

80%

90%

100%

FIGURE 4.8 Lorenz curve for aircraft movements at South American airports [Official Airline Guide (OAG), 2016].

airports handle much smaller traffic volumes. These three airports handle about 463,000 flights, which is just 14% of the total traffic of South America. This is one of the smaller world regions in terms of air traffic volume and the number of airports considered and has the lowest degree of concentration of any of the seven world regions (Gini coefficient of 0.7630 in 2016). Concentration increased between the years 2000 and 2016, especially in the first half, that is between the years 2000 and 2008. Furthermore, takeoffs increased by 1.04% per year between the years 2000 and 2016, and even by 2.76% per year during the period 2008 16. However, concentration increased less during the 2008 16 period. Fig. 4.9 illustrates the Lorenz curve for Europe, which comprises 685 airports for 2016 and thus is the third-largest world region in terms of airports considered. The biggest airports in Europe are London Heathrow (LHR), Amsterdam (AMS), and Istanbul (IST), followed by Frankfurt (FRA) and Paris Charles de Gaulle (CDG) with traffic volumes only marginally smaller than those of the top three airports. Out of the total of 685 airports in Europe, 22 have traffic volumes of over 100,000 flights, and 50% of the total traffic of Europe (8.2 million flights) is handled by just 32 airports. The other 653 airports handle altogether 4.1 million flights, each airport on average about 6300 flights per year. The Gini coefficient of 0.7986 is in halfway position, which is a result of both a high degree of airline alliances and LCC traffic. While airline alliances prefer hub airports for their operations, LCCs in Europe tended to choose secondary and regional airports. However, LCCs are increasingly offering flights from larger or even some hub airports. As a

80

PART | I Basic concepts

Share of aircraft movements at European airports

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Share of European airports 2000

2008

2016

FIGURE 4.9 Lorenz curve for aircraft movements at European airports [Official Airline Guide (OAG), 2016].

result, concentration decreased slightly between 2000 and 2008 and remained more or less constant between 2008 and 2016. There was an increase of take-offs during the period 2000 08 of 2.4% per year; however, traffic grew by only 0.7% per year during the 2008 16 period. Fig. 4.10 displays the Lorenz curve for Asia, which comprises 921 airports in 2016. As can be seen, the distribution of traffic is similar to the global distribution; traffic is concentrated in a relatively small number of big airports, such as Beijing (PEK), Shanghai Pudong (PVG), Tokyo Haneda (HND) and a few others. Only 24 out of 921 airports in Asia have traffic volumes of over 100,000 flights, and 53 airports have traffic volumes of more than 50,000 flights per year. On the other hand, nearly 870 airports in Asia have traffic volumes of less than 50,000 flights, or on average less than 140 flights a day. Looking at the traffic distribution in Asia, the concentration of traffic becomes evident through the skewedness of the distribution function. Asia is thus the second-largest world region in terms of airports considered and has a degree of concentration similar to Europe (Gini coefficient of 0.8168 in 2016). Concentration especially increased between the years 2000 and 2008, which was accompanied by large traffic growth of 7.2% per year (see Table 4.5). Between 2008 and 2016 the number of takeoffs grew even stronger at a rate of 7.5% per year, but concentration increased only slightly. Fig. 4.11 shows the Lorenz curve for the Middle East, which comprises only 111 airports in 2016 and is thus by far the smallest world region in terms of airports considered. However, the Middle East region has about the

Constrained and under-utilised airports Chapter | 4

81

Share of aircraft movements at Asian airports

100%

90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

0%

10%

20%

30%

40% 50% 60% Share of Asian airports 2000

2008

70%

80%

90%

100%

2016

FIGURE 4.10 Lorenz curve for aircraft movements at Asian airports [Official Airline Guide (OAG), 2016].

same traffic as Africa and the Southwest Pacific, namely about 1.2 million flights. The number of airports, however, at 111 is significantly smaller than in the other regions. By far the biggest airport in the Middle East is Dubai (DXB), followed by Doha (DOH) and Jeddah (JED), and these three airports have traffic volumes exceeding 100,000 flights. These three airports and Riyadh (RUH), Abu Dhabi (AUH) and Tel Aviv (TLV) handle over 50% of the total air traffic of the Middle East. The level of concentration is average compared to the other world regions (Gini coefficient of 0.8193) and increased especially between the years 2000 and 2008, which was accompanied by large traffic growth of 5.7% per year (see Table 4.5). Between 2008 and 2016, concentration remained constant, although traffic continued to grow by 7.5% per year. Fig. 4.12 displays the Lorenz curve for the Southwest Pacific, which comprises 401 airports in 2016 and is thus the third-smallest world region in terms of airports considered. The main contributor to the air traffic of this region is Australia, and the three biggest airports of this region are the Australian airports of Sydney (SYD), Melbourne (MEL) and Brisbane (BNE). Only SYD and MEL airports have traffic volumes exceeding 100,000 flights, the other 399 airports have smaller volumes. Almost 50% of the total traffic of 1.1 million flights is handled by just six airports, which means that the other 395 airports have, on average, traffic volumes of 1485 flights per year. However, despite the decrease of take-offs between 2000 and 2016 (20.31% per year), concentration increased consistently, and for 2016, the Southwest Pacific region exhibits the second-largest Gini coefficient of 0.8584 among the seven world regions.

Share of aircraft movements at airports in the Middle East

82

PART | I Basic concepts 100% 90%

80% 70% 60%

50% 40% 30%

20% 10% 0%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Share of airports in the Middle East 2000

2008

2016

Share of aircraft movements at airports in the Southwest Pacific

FIGURE 4.11 Lorenz curve for aircraft movements at airports in the Middle East [Official Airline Guide (OAG), 2016].

100% 90% 80% 70% 60% 50% 40% 30% 20%

10% 0%

0%

10%

20%

30% 40% 50% 60% 70% Share of airports in the Southwest Pacific 2000 2008 2016

80%

90%

100%

FIGURE 4.12 Lorenz curve for aircraft movements at airports in the Southwest Pacific [Official Airline Guide (OAG), 2016].

Fig. 4.13 illustrates the Lorenz curve for Africa, which comprises 388 airports in 2016. Africa is, like the Southwest Pacific and the Middle East, a region with still rather moderate traffic volumes, but growing traffic demand. Johannesburg (JNB), Cairo (CAI) and Nairobi (NBO) are the biggest airports

Constrained and under-utilised airports Chapter | 4

83

Share of aircraft movements at African airports

100% 90% 80%

70% 60% 50%

40% 30% 20% 10% 0%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Share of African airports 2000

2008

2016

FIGURE 4.13 Lorenz curve for aircraft movements at African airports [Official Airline Guide (OAG), 2016].

in Africa, but only Johannesburg reaches a traffic volume of more than 100,000 flights. Eighteen of the 388 airports handle 50% of the total traffic of Africa, which counted in 2016 about 1.2 million flights. While take-offs increased between 2000 and 2016 by 4% per year, traffic concentration has remained almost constant. The Gini coefficient has a value of 0.7676 and is thus the second-lowest among the seven world regions. The rather low degree of concentration seems to be a result of less developed markets and airlines (Gelhausen and Berster, 2017). Not surprisingly, we find similar distribution functions in all world regions. Air traffic is highly concentrated in few airports, often identical to hub airports, whereas the majority of airports have rather small flight volumes, absolutely and compared with airport capacity. However, there is some variation in traffic concentration between different world regions which depends on factors often related to regional circumstances. The share of airline alliances, LCCs and of secondary and regional airline traffic seems to play an important role. Airline alliances depend more on hub traffic than LCCs and regional airlines, which are more focused on smaller and regional airports (Gelhausen and Berster, 2017). However, this could change in future if LCCs increasingly offer flights from hub airports. On the other hand, traffic growth seems to be less suitable for explaining traffic concentration in various cases, especially in more mature markets. In theory, one could extend the concentration analysis to each country of the world in order to study the variation of concentration. Although such an analysis could be carried out, the results would probably not justify the

84

PART | I Basic concepts

effort; instead, four countries have been selected: France and Germany in Europe, the United States in North America and China in Asia. France has a network which is concentrated in Paris; most other regional airports have strong links with the capital’s airports but handle much smaller volumes than the Paris airports Orly and Charles de Gaulle. Germany has a less centralised network, however, with Frankfurt and Munich as the two principal hub airports. The United States and China are, of course, much bigger countries than those in Europe. While the US air transport market is rather saturated, and the European markets will be more or less approaching saturation in the coming years, the Chinese market is still relatively young and growing dynamically. In Fig. 4.14 the cumulative traffic distribution in the airport networks of these countries is shown for the year 2016. While in Germany there are only three airports and in France two airports with more than 100,000 flights a year, there are many more airports in China and, especially, in the United States with traffic volumes exceeding 100,000 take-offs. It can be seen that the biggest airports differ in traffic volume, and that the decline in traffic is particularly pronounced in the United States, but it similarly exists among the big airports in the other countries. If we concentrate on the biggest ten airports in each country, we see that the volume span between the biggest and the tenth biggest airport in Germany reaches from about 19,000 flights to over 227,000 at Frankfurt airport (FRA), in France, from around 9000 to more than 224,000 at Paris Charles de Gaulle airport (CDG), in China, from about 129,000 to over 100%

Share of aircraft movements at German/French/Chinese/US airports

90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

0%

10%

20%

30% 40% 50% 60% 70% 80% Share of German/French/Chinese/US airports France

Germany

China

90%

100%

US

FIGURE 4.14 Cumulative air traffic distribution in the airport networks of Germany, France, China and the United States in 2016 [Official Airline Guide (OAG), 2016].

Constrained and under-utilised airports Chapter | 4

85

303,000 take-offs at the biggest airport in Beijing (PEK), and in the United States from around 200,000 to about almost 438,000 flights in Atlanta (ATL). In Germany and France, the tenth biggest airports are already airports with small traffic volumes, whereas in China and the United States, the tenth biggest airports are big airports with volumes exceeding 100,000 aircraft movements. The cumulative distribution of air traffic reveals more of the concentration of traffic than just the direct distribution. As can be seen, the country-specific traffic distribution functions are similar to those of the corresponding world regions. Their skewedness signals the high degree of traffic concentration in the biggest airports in the country, which form a relatively small part of the overall network, and the great number and share of airports with small traffic volumes, on the other hand. Traffic concentration is already rather high in Germany 90% of the total traffic is handled by 26% of all airports. Nevertheless, the concentration is still higher in the other countries. In the United States, 90% of total traffic is concentrated in just 18% of the 713 airports, while the other 82% of airports handle just 10% of total traffic. The biggest 19 airports of the US are responsible for 50% of total flight volume of 9 million take-offs. Traffic in France is concentrated in Paris. The two Parisian airports Charles de Gaulle and Orly handle 47% of the total traffic of France (about 727,000 flights in 2016), while 90% of the total traffic is concentrated in 22% of all airports (60). In China, traffic is concentrated in 12 airports (of 213 in total), which handle about 50%, and 51 airports which handle 90% of total traffic of almost 4 million flights. This means that 162 airports had an overall traffic of 400,000 flights in 2016, or on average a traffic volume of just 2500 flights per airport per year. Air traffic is thus rather concentrated in the selected countries and in many other countries, as the distribution functions of the world regions have shown, on a relatively small number of airports, and in most cases, only these airports handle high flight volumes of say more than 100,000 flights, while the great majority of airports deal with only small traffic volumes of some 1000 flights a year. These volumes are well below the capacity of airports so that these airports are at least theoretically in a position to take over a much greater share of the total traffic. Due to the HS concept followed by many scheduled carriers, traffic tends to concentrate in hub airports, thus creating high traffic volumes at peak times and often high degrees of capacity utilisation as well. Other airports depend very much on their catchment in the surrounding area whether or not high traffic volumes are achieved. Many regional and peripheral airports do not have a strong catchment area, so that they are and will be lacking traffic. In recent years, these airports have partly lost traffic because of low growth rates of demand and the change in aircraft size (passengers and seats per aircraft, respectively). Airlines have, in general, increased seat capacity per aircraft to be more competitive. At regional airports this has often led to suspending routes with insufficient demand. As

86

PART | I Basic concepts

Share of aircraft movements at German/French/Chinese/US airports

a consequence, these airports suffer from low traffic volumes, since airports have problems surviving economically because often the number of flights is not high enough to cover operating costs. Many airports are thus underutilised. Examination of the development of traffic distribution in the four selected countries shows a similar picture as on the global scale, as Fig. 4.15 illustrates. Traffic has been concentrated in important airports, and the level of concentration did not change significantly. In both Germany and France, the degree of concentration has increased over time, if we look at the traffic share of the most important airports. In Germany, the top four out of 33 airports handled a traffic share of 59% in 2000 and 64% in 2016. In France, the two Parisian airports, as the most important ones, had a traffic share of 44% in 2000. This share increased to 47% in 2016. If we look at the share of airports which handled 90% of total traffic, in Germany we see a decrease of airport share, corresponding to a decrease of the number of airports, from 28% to 26%, thus a slight increase of traffic concentration. In France, the share of airports with 90% traffic share stayed constant over time at 22%. The development of traffic concentration in China and the United States has been similar to the development in the two European countries, however, against a fundamentally diverging background of network and traffic development in these regions. While in China both the number of airports and flights has grown substantially, the contrary occurred in the United States where both the traffic and network

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Share of German/French/Chinese/US airports France 2000

France 2016

Germany 2000

Germany 2016

China 2000

China 2016

US 2000

US 2016

FIGURE 4.15 Cumulative air traffic distribution in airport networks in Germany, France, China and the United States from 2000 to 2016 [Official Airline Guide (OAG), 2016].

Constrained and under-utilised airports Chapter | 4

87

size decreased. In China, the share of airports that handled 50% of total traffic decreased from 8% in 2000 to 6% in 2016, equivalent to an increase in traffic concentration. The absolute number of corresponding airports increased from eight to 13 airports; however, the total number of airports with regular services increased significantly from 113 to 213 airports. In the United States, the share of airports that handled half of total traffic decreased from just 3.2% in 2000 to 2.6% in 2016, indicating not only a high degree of concentration, but an even further intensifying of concentration. If we look at the share of airports which handled 90% of total traffic, this share decreased in China from 31% in 2000 to 24% in 2016, and in the United States from 24% to just 18% in 2016. By both indicators, the 50% and 90% traffic share, the already high level of traffic concentration moved upwards even further in both countries.

4.4

How to describe levels of airport capacity constraint

As has been demonstrated in Chapter 3, Capacity utilisation at airports worldwide, a series of important airports suffer from capacity bottlenecks now or are likely to in the future. However, we have to differentiate between occasional and peak hour and severe bottlenecks. In this chapter, we propose and describe quantitatively different levels of capacity constraint of airports in the global network. So far, a number of studies have shown the airport capacity problem by identifying constrained airports based on their capacity, forecasts of demand and thus the unaccommodated demand (Eurocontrol, 2018). Capacity constraint is not a black-and-white problem, but one with a great range of severity. Airport congestion encompasses the whole transition area between constrained flow conditions at some peak hours and dense traffic conditions, with high delays for each aircraft over longer periods of time. With airport capacity problems, we mean not only the level of capacity utilisation described by the ratio of demand (in this case the number of aircraft movements) and capacity but also the negative effects of a near-capacity utilisation on the performance of the airport. The main effects are certainly the growing complexity of operations and increasing delays of flights when traffic approaches the capacity. While we have found an answer to the problem of how to measure capacity utilisation, we have a serious data problem of describing aircraft delays at airports in relation to the level of capacity utilisation on a global scale. Since delay data that is comparable among airports is not available for the time being, we pursue an approach which describes the intensity of capacity constraint in relation to the relative amount of time in which aircraft movements at an airport are handled under near-capacity conditions. We think that besides the delay statistic, the time statistic is a well-suited indicator of the constraint problem. The task is then to determine for airports, which have been identified as airports with capacity problems, the practical capacity and

88

PART | I Basic concepts

to estimate the number of hours in which they operate at or near capacity. The methodological tool to quantify these indicators consists of traffic ranking functions, which have been developed for global airport capacity studies and are described in Section 3.2. Before applying volume ranking curves for estimating the level of airport constraint, we have to identify those airports which may be or are already constrained. For that, we have selected those airports with a threshold traffic volume of 70,000 aircraft movements in 2016. We could have selected other threshold volumes as well. The decision to choose the volume of 70,000 aircraft movements was influenced by the fact that single runway airports, with a volume in that order, have a 5% peak hour volume of around 20 movements, which corresponds to about 50% of the hourly capacity of a runway under IFR conditions. Airports with smaller volumes can be seen as airports of rather regional importance without any capacity problems in the near future. In 2016, there were 229 airports worldwide with traffic volumes exceeding 70,000 aircraft movements, handling about two-thirds of the total global flight volume. In order to identify constrained airports, in the sample of 229 airports, we have applied an approach which follows a step-wise procedure of eliminating airports with low annual traffic volumes lower than defined threshold volumes by capacity class of the airport and then with lower than defined 5% peak hour volumes. If the values of these indicators exceed certain threshold values in the case of the 5% peak hour volume depending on the capacity class of the airport (single runway, two parallel runways, etc.) then the airport can be regarded as an airport with congestion problems over more or less operating hours of the year, the duration depending on the value of the thresholds. The relative time span during a year in which airports operate at or near capacity can be visualised in traffic volume ranking curves, provided the ranking curves are comparable between airports and the time dimension is expressed as the relative time of the year in which the airport operates over or at a certain volume. Given the lower threshold volume of 70,000 movements for a single runway airport, we have estimated corresponding values for the higher capacity classes analogous to the single runway class. Thereby, we have relied on estimates of annual service volumes which have been calculated on the basis of a data envelopment analysis (DEA) and a follow-up regression analysis, as described in Chapter 7, Modelling future airport capacity and capacity utilisation. Table 7.4 displays model estimates of annual service volumes of generic airports with different numbers of runways and operating hours. As can be seen, the annual service volume as a proxy of the annual capacity of a single runway airport has been calculated to about 238,000 aircraft movements per year. The threshold volume of 70,000 movements corresponds to about 30% of that capacity. In a similar analogy, the threshold volumes of

Constrained and under-utilised airports Chapter | 4

89

the higher capacity classes have been derived; the corresponding values for airports with up to five runways are as follows: G

G

G G G G

Airports with two independent parallel runways: 140,000 aircraft movements Airports with two dependent parallel runways: 125,000 aircraft movements Airports with two crossing runways: 125,000 aircraft movements Airports with three runways: 165,000 aircraft movements Airports with four runways: 195,000 aircraft movements Airports with five runways: 220,000 aircraft movements

The threshold values for the higher capacity class airports, of which there are only very few, have been derived accordingly. The second and more stringent threshold is the 5% peak hour volume. Here, we have not taken the actual value of 5% peak hour volume of each airport, but a relative value, that is the ratio of the actual 5% peak hour volume and the maximum 5% peak hour volume of the corresponding airport capacity class as derived from the DEA and regression analysis described in Chapter 7, Modelling future airport capacity and capacity utilisation. By retaining the relative 5% peak hour volume, we achieve a better comparability between constrained airports regarding the time duration of constrained traffic conditions. The threshold value for including airports into further constraint analysis has been set to 70% of the maximum 5% peak hour volume. The latter represents a practical hourly capacity and is calculated by dividing the so-called maximum average hourly volume of an airport by the highest attainable CUI of the corresponding airport capacity class (for a more detailed description see Section 7.2). The modelled maximum 5% peak hour volumes by airport capacity class are given in Table 7.6 of Chapter 7, Modelling future airport capacity and capacity utilisation. It should be noted that these capacity values are the result of an analysis which has no relationship with capacity estimates leading to the ‘declared capacity’ in the process of the IATA-type slot coordination at coordinated airports. The corresponding threshold values of the 5% peak hour volume for generic airports with up to five runways are as follows: G G G G G G G

Single runway airports: 30 aircraft movements Airports with two independent parallel runways: 60 aircraft movements Airports with two dependent parallel runways: 55 aircraft movements Airports with two intersecting runways: 55 aircraft movements Airports with three runways: 75 aircraft movements Airports with four runways: 95 aircraft movements Airports with five runways: 110 aircraft movements

The threshold values for the higher capacity class airports, of which there are only very few, have been derived accordingly. In the constraint analysis,

90

PART | I Basic concepts

the maximum 5% peak hour volumes have been calculated for each airport separately, the generic airport values are given here as proxy values of the actual ones. Taking Geneva airport as an example, we see that the single runway airport had an annual traffic volume of about 148,000 aircraft movements (according to OAG) in 2016 with a 5% peak hour volume of 35 movements, thus seven movements more than the Geneva specific threshold volume of 28 movements. Geneva airport is, therefore, further examined with respect to the duration in which the airport is operated under constrained conditions. As has been mentioned, to reveal the degree of constraints, we use again the instrument of traffic ranking curves. Traffic ranking functions, as discussed in Chapter 3, Capacity utilisation at airports worldwide, show the distribution of capacity utilisation within a given year. The ranking function allows determination of the number of hours when traffic exceeds specific values. These values have been taken as actual volumes. They may, however, also be taken as relative values, in particular as a percentile of the socalled 5% peak hour capacity (or maximum 5% peak hour volume). Equally, we can convert the absolute number of hours of the year into the share of the total number of operating hours. The resulting ‘relative’ traffic ranking function has the same shape as the absolute counterpart used so far; however, the hourly volumes are related to the 5% peak hour capacity, and the hours are expressed as shares of the total number of operating hours. We are now in a position to read directly the relative time span of a year in which the airport handles traffic volumes exceeding certain percentages of the practical hourly capacity, for instance more than 70%. In searching for a measure of the constraint intensity, we have to ask, at what traffic volume does the airport begin to experience flight delays that grow in relative terms faster than the number of additional flights? Lacking delay data for today’s traffic at high-volume airports, we have to assume a certain level of traffic in relation to the capacity. We have decided to draw the dividing line between free traffic flow conditions and those when first constraints and delays occur at the airborne or groundside of the runway, at 70% of the practical capacity as expressed by the maximum 5% peak hour volume. For a single runway airport, the threshold volume would thus be around 30 aircraft movements per hour. Furthermore, we were interested in defining capacity utilisation classes which serve as indicators of the time duration in which the airport operates at or near capacity. After examining the traffic ranking functions of the selected high-volume airports, we have grouped the airports into four classes, whereby all classes have as a lower 5% peak hour volume limit 70% of the 5% peak hour capacity and otherwise differ in the relative amount of time when the lower limiting hourly volume is exceeded. The time class limit of groups is set to 5%, 35% and 65%. The four capacity utilisation classes are thus defined as follows:

Constrained and under-utilised airports Chapter | 4 G

G

G

G

91

Class A: Airports with 5% peak hour traffic volumes exceeding 70% of the 5% peak hour capacity in more than one hour and in less than 5% of all operating hours of the year. Class B: Airports with 5% peak hour traffic volumes exceeding 70% of the 5% peak hour capacity in more than 5% and in less than 35% of all operating hours of the year. Class C: Airports with 5% peak hour traffic volumes exceeding 70% of the 5% peak hour capacity in more than 35% and in less than 65% of all operating hours of the year. Class D: Airports with 5% peak hour traffic volumes exceeding 70% of the 5% peak hour class in more than 65% of all operating hours of the year.

It should be noted that the capacity utilisation as expressed by these classes may differ greatly from capacity utilisation which is related to the declared capacity at slot coordinated airports. Typically, but not always, declared capacities are lower than the maximum 5% peak hour volumes derived here as estimates of practical capacity. As an example, the relative traffic ranking curve for Geneva airport is shown in Fig. 4.16. Dividing lines are drawn, on the one hand, at the 100% and 70% level of the volume capacity ratio and, on the other hand, at the 5%, 35% and 65% time share of the total number of annual operating hours. As can be seen, the ranking curve intersects the 70% level of capacity share between the 5% and 35% lines of the annual time share (at about 10%). The airport therefore belongs to the capacity utilisation class B. In other words, by handling traffic volumes 140%

A

C

B

D

Share of aircraft movements

120% Estimated 5% peak hour volume=100% 100%

80%

60%

40%

20%

0%

0%

10%

20%

30%

40% 50% 60% Share of hours of the year

70%

80%

90%

100%

FIGURE 4.16 Relative traffic ranking curve of Geneva airport (GVA, capacity utilisation class B) in 2016 [Official Airline Guide (OAG), 2016].

92

PART | I Basic concepts

exceeding 70% of the practical hourly capacity in 10% of all operating hours in 2016, Geneva airport is regarded as a constrained airport with peak hour problems. However, the airport has not yet been confronted with severe capacity problems over longer time periods.

4.5

Capacity constrained airports and under-utilised airports

For classifying airports according to the level of constraint, airports with traffic volumes exceeding 70,000 aircraft movements in 2016 have been identified in the first step. As has been mentioned, the total number of those airports was 229 in 2016. They have been categorised by runway class since the lower threshold volume grows with the capacity class, as described in the former chapter. Therefore 57 airports have been identified which are not yet constrained, since their hourly traffic volumes in all capacity classes or their annual traffic volumes in capacity classes with more than one runway do not reach the level of the lower threshold volume of each capacity class. This means that 172 airports remain with 5% peak hour volumes exceeding 70% of the practical hourly capacity in at least one hour of the year. For these airports, relative traffic ranking curves have been derived in order to determine the duration of operations at or near capacity as part of the annual operating time. Figs 4.17 4.19 display the relative traffic ranking curves of 35 airports belonging to capacity utilisation classes B, C or D. In total, 137 airports have been identified as airports with 5% peak hour traffic volumes exceeding 70% of the practical hourly capacity in at least one hour and in less than 5% of all operating hours, thus belonging to the capacity utilisation class A. The question is whether or not these airports may be classified as congested airports. If an airport experiences just a few hours within a year where traffic levels approach capacity, one would probably not yet speak of a constrained airport. If, on the other hand, hourly traffic approaches capacity in 5% of all operating hours, meaning in around 300 400 hours per year, then some airport operators would most likely regard such an airport as one having a capacity problem. We would not contradict airport operators in such a situation. However, we think that in this global analysis, an airport should be regarded as a constrained airport if traffic reaches levels at or near capacity in more than 5% of all operating hours of a year. We have therefore classified airports in capacity utilisation classes B, C and D as constrained airports. We have to realise here that ultimately a statement as to whether or not an airport has a capacity problem depends certainly on objective criteria, the evaluation of which, however, is also a normative one and often leads to political considerations and judgement. As can be seen in Fig. 4.17, as demonstrated by the traffic ranking over all operating hours of the year 2016, 20 airports in the global network have hourly traffic distributions with a peak hour capacity utilisation of more than 70% of the practical capacity in more than 5% and less than 35% of all

Constrained and under-utilised airports Chapter | 4 B

A

140%

C

D

DEN SEA SAN LIM STN GVA DFW CLT IST EWR DEL PHX SIN BKK BOM MEL BLR SUB AKL XMN

120% Share of aircraft movements

93

Estimated 5% peak hour volume=100% 100%

80%

60%

40%

20%

0%

0%

10%

20%

30%

40% 50% 60% 70% Share of hours of the year

80%

90%

100%

FIGURE 4.17 Relative traffic ranking curves of airports in capacity utilisation class B in 2016 [Official Airline Guide (OAG), 2016].

140%

B

A

C

D CGK DXB HKG

Share of aircraft movements

120%

LAX SAW

Estimated 5% peak hour volume=100%

ATL

100%

MEX LGW ORD

80%

CAN MUC DUB

60%

40%

20%

0%

0%

10%

20%

30%

40% 50% 60% 70% Share of hours of the year

80%

90%

100%

FIGURE 4.18 Relative traffic ranking curves of airports in capacity utilisation class C in 2016 [Official Airline Guide (OAG), 2016].

94

PART | I Basic concepts

140%

B

A

C

D PEK LHR LGA

Share of aircraft movements

120% Estimated 5% peak hour volume=100% 100%

80%

60%

40%

20%

0% 0%

10%

20%

30%

40% 50% 60% 70% Share of hours of the year

80%

90%

100%

FIGURE 4.19 Relative traffic ranking curves of airports in capacity utilisation class D in 2016 [Official Airline Guide (OAG), 2016].

operating hours. These airports are therefore classified as airports of capacity utilisation class B. They are constrained on average in around 20% of all operating hours, corresponding to about three to four hours per day. Among them are important traffic nodes such as Denver (DEN), Dallas/Fort Worth (DFW) and Newark (EWR) in North America, Singapore (SIN), Bangkok (BKK) and Delhi (DEL) in Asia and London Stansted (STN), Geneva (GVA) and Istanbul (IST) in Europe. In 2016, the airports in this group handled a traffic volume of 6.28 million aircraft movements, corresponding to 9% of the global traffic of 70.9 million movements. Twelve airports are in the capacity utilisation class C (see Fig. 4.18); they handle peak traffic volumes approaching capacity in 35% 65% of all operating hours of a year, or in about six to 11 hours per day. On average, these airports are constrained in about half of their operating time with severe impacts on the complexity of operations and traffic delay. The 12 airports are Atlanta Hartsfield-Jackson (ATL), Chicago O’ Hare (ORD) and Los Angeles (LAX) in North America, Guangzhou (CAN), Jakarta (CGK) and Hong Kong (HGK) in Asia, Mexico City (MEX) in South America, Dubai (DXB) in the Middle East and Munich (MUC), London Gatwick (LGW), Istanbul Sabiha Go¨kcen (SAW) and Dublin (DUB) in Europe. These airports have a traffic share of 8% of the global traffic in 2016 with 5.5 million aircraft movements. Beijing (PEK), London Heathrow (LHR) and New York LaGuardia (LGA) are the airports with the highest degree of congestion. In more than

Constrained and under-utilised airports Chapter | 4

95

65% of all operating hours, their peak hour volumes exceed levels of over 70% of the practical capacity (see Fig. 4.19). These three airports belong thus to the capacity utilisation class D. While traffic in Beijing and New York LaGuardia reaches volumes at or near capacity in about 70% of all operating hours, London Heathrow is congested in about 80% of all hours of operation. No other airport reaches such a high capacity utilisation. The high degree of congestion has already prevailed for several years with the consequence that London Heathrow has not participated in the overall traffic growth. The airport is fully slot coordinated, and all slots available have been used by carriers for many years. Since the capacity of the runway system has remained constant, flight volume has not risen either. The number of passengers has grown due to the fact that the seat capacity per flight has been increased by some airlines. The three airports of class D handled a traffic volume of 1.45 million aircraft movements in 2016, corresponding to 2% of the global traffic volume. The spatial distribution of constrained and non-constrained airports reveals a concentration of constrained airports in Asia, North America and Europe, as can be seen in Table 4.7. The global network consists of more than 4000 airports, of which more than 2600 (64%) are located in the three major air traffic regions. Thirty-five airports have been identified as constrained airports (in capacity utilisation classes B, C and D) worldwide, and only five of them are located in the Middle East, South America and the Southwest Pacific. None of the airports in Africa are yet constrained. The other 30 (86%) constrained airports are Asian, European and North American airports. All airports in these regions handle a traffic volume of 57 million aircraft movements, corresponding to about 80% of the global traffic volume of 70.9 million movements in 2016. The 30 constrained airports alone handle almost 12 million aircraft movements, corresponding to 17% of the global traffic and 21% of the traffic in these regions. While 83% of the global air traffic is operated at over 4000 airports in non-constrained conditions, 17% is concentrated in just 30 major airports, which suffer from traffic congestion occurring daily for several hours. As can be seen in Table 4.7, 57 airports are classified as class 0 airports, which are not constrained, since their hourly or annual traffic volumes are lower than the corresponding threshold volumes defined for each airport capacity class. These airports have, however, annual volumes exceeding 70,000 aircraft movements in 2016. For example, the single runway airport of Glasgow (GLA) belongs to this group. Glasgow exceeds the 70,000 movement threshold volume with 85,000 aircraft movements in 2016. However, with a 5% peak hour volume of 22 movements, it does not reach the lower hourly threshold volume of around 30 movements. Other examples of airports in this group are Copenhagen (CPH) and Milan Linate (LIN) in Europe; Seoul (ICN), Shanghai (SHA) and Hyderabad (HYD) in Asia; Boston (BOS), Las Vegas (LAS), Vancouver (YVR) and Honolulu (HNL) in North America; Cairo (CAI), Nairobi (NBO) and Addis Ababa (ADD) in

TABLE 4.7 Airports by capacity utilisation class and world region [Official Airline Guide (OAG), 2016]. World region

Number of airports in capacity utilisation class 0 ,70,000 annual aircraft movements

Total number of airports A

B

C

D

. 70,000 annual aircraft movements Without constraints

Constrained

Africa

381

4

3

0

0

0

388

Asia

853

7

50

7

3

1

921

Europe

627

11

40

3

4

1

685

Middle East

100

5

5

0

1

0

111

North America

960

23

24

7

3

1

1018

South America

511

5

11

1

1

0

530

Southwest Pacific

393

2

4

2

0

0

401

World

3825

57

137

20

12

3

4054

Σ

229

Constrained and under-utilised airports Chapter | 4

97

Africa; Riyadh (RUH), Abu Dhabi (AUH) and Tel Aviv (TLV) in the Middle East; Buenos Aires (AEP) and Cancun (CUN) in South America and Adelaide (ADL) in the Southwest Pacific. Another 137 airports are in the capacity utilisation class A, which experience some congestion in peak hours, although only for short periods in about 300 400 hours per year. Although these airports have not been classified as constrained airports, they are on the brink of congestion, and with further traffic growth, it is a matter of some years before they will be constrained as well, if no capacity enhancing measures are realised. This group is by far the biggest among the four capacity utilisation classes. About 50 of them can be found in the group of 200 top-ranking airports with the highest traffic volumes. While the share of class A airports (137) in the global network is not yet even 4%, almost one-quarter of the most important airports are class A airports, meaning that a greater number of the highest volume airports are approaching capacity limit in the near future than in the global network of over 4000 airports. Examples of class A airports in Europe are Frankfurt (FRA), Rome (FCO), Barcelona (BCN), Moscow (SVO), Vienna (VIE), Manchester (MAN), Palma de Mallorca (PMI) and Nice (NCE). Some of these airports may reach capacity limits at present more often than shown in this analysis, as is the case in Frankfurt or Du¨sseldorf (DUS). The reason is that these airports reach traffic volumes which approach their declared capacity, as they are fully coordinated. Typically, declared capacities are smaller than the practical capacity we have estimated, which is solely based on runway capacity, whereas the declared capacity is a compound measure of all capacity constraining elements of an airport. If we mirrored the relative volume ranking against declared capacity, these airports would most likely be in the higher capacity utilisation classes. Europe and Asia have a relatively high share of class A airports in the network of high-volume airports (with volumes exceeding 70,000 movements). Around 40 of the 50 class A airports are located in these two world regions. Examples in Asia are Chengdu (CTU), Kuala Lumpur (KUL), Manila (MNL), ¨ ru¨mqi (URC) and Bangkok (DMK), Ho Chi Minh (SGN), Fukuoka (FUK), U Denpasar (DPS). Class A airports in North America are New York (JFK), San Francisco (SFO), Minneapolis (MSP) and Miami (MIA). In Africa, Cape Town (CPT) is an example and in the Middle East, Doha (DOH) belongs to the class A airports group; in South America, Bogota (BOG) and Sao Paulo (GRU); and in the Southwest Pacific region, Sydney (SYD).

4.6 Case study: development of capacity utilisation by capacity utilisation class at three example airports: San Diego (SAN), London Heathrow (LHR) and Beijing (PEK) As described in Section 3.5, the airports of San Diego (SAN), Beijing (PEK) and London Heathrow (LHR) have been selected as case study airports,

98

PART | I Basic concepts

because they represent important nodes in the global airport network and are examples of constrained airports. They differ, however, in traffic volume, structure and capacity development. While traffic volumes at LHR and SAN have not grown since the year 2000 in LHR because of capacity limitations and in SAN primarily because of demand saturation traffic in PEK has developed dynamically in that period (see Fig. 3.13). In Section 3.5, we have shown the development of capacity utilisation at these airports by means of traffic ranking curves. They have revealed that the airports handle traffic volumes at or near capacity over many hours of the year and that peak hour and average hour traffic volumes have increased over time in SAN and PEK, but not in LHR due to capacity limitations. Ranking functions with absolute values of traffic volume and hours of the year have not yet allowed reading of the capacity utilisation in relation to capacity and share of total operating time of a year. We have therefore developed relative traffic ranking curves with hourly volumes as shares of capacity and time duration as share of annual operating time, as described in this chapter. By means of relative traffic ranking curves of the years 2000, 2008 and 2016, we are able to show the severity and development of airport capacity constraint at the selected airports (see Figs 4.20 4.22). While the capacity of SAN has barely changed over time, traffic increased slightly from almost 190,000 aircraft movements in 2000 to nearly 198,000 in 2008 and decreased again until 2016 to about 173,000 movements. Similar to the absolute traffic ranking curves (see Fig. 3.17), the relative curves vary only slightly (see Fig. 4.20). The 2000 curve moves upwards in 2008 and moves down again 140%

B

A

C

D SAN 2000

Share of aircraft movements

120%

SAN 2008

SAN 2016

Estimated 5% peak hour volume=100% 100%

80%

60%

40%

20%

0%

0%

10%

20%

30%

40% 50% 60% 70% Share of hours of the year

80%

90%

100%

FIGURE 4.20 Relative traffic ranking curves of San Diego (SAN) airport in the years 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016].

Constrained and under-utilised airports Chapter | 4 C

B

A

140%

99

D LHR 2000

Share of aircraft movements

120%

LHR 2008 LHR 2016

Estimated 5% peak hour volume=100% 100%

80%

60%

40%

20%

0%

0%

10%

20%

30%

40% 50% 60% 70% Share of hours of the year

80%

90%

100%

FIGURE 4.21 Relative traffic ranking curves of London Heathrow (LHR) airport in the years 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016].

140%

A

C

B

D PEK 2000 PEK 2008

Share of aircraft movements

120%

PEK 2016

Estimated 5% peak hour volume=100%

100%

80%

60%

40%

20%

0%

0%

20%

40% 60% Share of hours of the year

80%

100%

FIGURE 4.22 Relativetraffic ranking curves of Beijing (PEK) airport in the years 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016].

in 2016, with different points intersecting the capacity share line at 70%. In 2000, the airport handled traffic volumes exceeding 70% of the practical capacity in about 28% of all operating hours, thus belonging to capacity utilisation class B. Due to the traffic increase, the airport became more

100

PART | I Basic concepts

constrained and handled volumes at or near capacity in about 40% of all operating hours of the year, thus changing to capacity utilisation class C. With lower traffic in 2016, the degree of constraint at San Diego airport diminished, in only around 12% of all hours of the year were traffic volumes above the 70% capacity line. The airport moved back into capacity utilisation class B. At first sight, the situation at London Heathrow (LHR) airport looks similar to San Diego, because traffic and capacity at both airports did not change much over time. The degree of capacity utilisation, however, was significantly higher at LHR, as can be seen in Fig. 4.21. The three relative traffic ranking curves of the years 2000, 2008 and 2016 are almost identical and intersect the 70% capacity share line at about 80%, meaning that in about 80% (equivalent to about 5400 hours) of all operating hours of the year, the airport handled hourly traffic volumes approaching capacity. LHR belongs to capacity utilisation class C and was thus highly constrained over the whole period from 2000 to 2016. The capacity estimated for the capacity utilisation analysis was not reached even in absolute peak hours, since the planned traffic at LHR was always below the declared capacity of around 88 aircraft movements, which was slightly lower than the capacity estimated by means of the DEA applied here. The traffic and capacity of Beijing (PEK) airport developed in a completely different way to that of San Diego and London Heathrow. Traffic grew from about 170,000 aircraft movements in 2000 the same volume as San Diego handled in 2016 to over 600,000 movements in 2016 and the capacity went up late in 2007 from two to three independent runways. As already seen in the discussion of the absolute traffic ranking curves of PEK (see Fig. 3.19), the airport handled the traffic volume in 2000 in ‘free-flow’ conditions without any capacity problems. This is accentuated by the relative ranking curve, as shown in Fig. 4.22. In most daytime hours the airport operated with volumes well below the practical capacity. The no-constraint condition of 2000 changed in 2008 when the airport handled an annual volume of around 430,000 movements, the and hourly peak volumes exceeded 70% of the capacity in more than 10% of all operating hours of the year. The airport has been categorised thus as a capacity utilisation class B airport. Further traffic growth of more than 170,000 aircraft movements until 2016 caused a prolongation of the time duration of constrained traffic conditions. The relative ranking function reveals that the airport was charged with traffic volumes at or near capacity in nearly 70% of all operating hours of the year, just as LHR, which belongs to capacity utilisation class D. It seems that London Heathrow has reached a maximum utilisation of capacity since the airport was not in a position to increase the degree of utilisation although the air transport demand in the greater London area continued to grow. LHR has been a class D airport with traffic volumes at or near capacity for about 80% of the time for more than 16 years. We can therefore expect that Beijing

Constrained and under-utilised airports Chapter | 4

101

airport will be able to grow a little further until an utilisation rate of around 80% is reached. The additional demand for flights in the Beijing region will then have to be handled by the new airport south of the capital city.

4.7

Conclusion

This chapter is subdivided into two parts, a descriptive part dealing with the air traffic distribution in the global airport network and in regional networks, and an analytical part describing the method of identifying constrained airports and categorising them according to the degree of capacity constraint. Identifying constrained airports also revealed the great number of airports with rather small traffic volumes. These airports have ample capacity reserves and would welcome additional traffic. The total number of airports with regular scheduled services, which has been the subject of the concentration and constraint analysis, has slightly increased from 4035 in 2000 to 4054 in 2016. In contrast, the traffic volume has become much stronger, from 27.9 million flights (corresponding to 55.8 million aircraft movements) in 2000 to 35.5 million flights. Since the number of airports remained almost stable, the average traffic volume per airport has grown from about 6900 in 2000 to around 8700 flights in 2016. The cumulative global traffic distribution has shown that only about 15% of all airports (equal to about 600 airports) handle above-average traffic volumes. In other words, 85% handle traffic volumes smaller than 8700 flights per year. It is obvious that roughly 3400 airports are far from having serious capacity bottlenecks and, at least in theory, they form a great capacity reservoir for future traffic growth. In reality, traffic has been concentrated in a small number of important airports, for example, hub airports. The high degree of concentration has been confirmed in the 2016 analysis. Half of the global traffic is handled by just 120 airports (3% of all airports), that is almost 300,000 aircraft movements per year per airport, and the other half by the remaining 3934 airports, corresponding to just 9000 annual aircraft movements per airport. The spatial concentration of the population and the need of airlines to bundle air traffic in HS networks are prime factors of traffic concentration in regional areas as well as the global network. The analysis of concentration in 2000, 2008 and 2016 has shown that the traffic distribution has hardly changed over time; the global air traffic was concentrated in 2000 and has been concentrated since then. Air traffic in world regions has been similarly as concentrated as the global traffic, with high Gini coefficients varying only slightly around the global value of 0.8354. North America is the region with the highest degree of traffic concentration as measured by the Gini coefficient of 0.8668. The country-specific distribution functions of the selected countries of China, United States, France and Germany are similar to those of the corresponding world regions. Their skewness signals the high degree of traffic

102

PART | I Basic concepts

concentration on the biggest airports in each country, which form a relatively small part of the overall network, and the great number and share of airports with small traffic volumes. To identify constrained airports and categorise them according to the level of constraint, we have to give an answer to the problem of how to define and measure capacity constraint. It is certainly not a black-andwhite problem, but one with a great range of severity. Airport congestion encompasses the whole transition area between constrained flow conditions in some peak hours and dense traffic conditions with high delays for each aircraft over longer periods of time. Due to the lack of delay data comparable between airports, we have pursued an approach which describes the intensity of capacity constraint in relation to the relative amount of time in which aircraft movements at an airport are handled at or near capacity. For those airports, which have been identified as airports with capacity problems, we had to determine the practical hourly capacity and estimate the number of hours in which they operate at or near capacity. The methodological tool to quantify these indicators consists of traffic ranking functions (see Section 3.2). In order to identify constrained airports, we have in the first step selected those airports with a threshold traffic volume of 70,000 aircraft movements in 2016. These airports have peak hour traffic volumes typically not exceeding 50% of the practical capacity of around 40 movements in the case of single runway airports and hence no capacity problems in the near future. Thus 229 airports were identified as airports with more than 70,000 movements; another 57 airports have traffic volumes smaller than the threshold values of the higher capacity class airports with two and more runways. The remaining 172 airports are likely to operate at or near capacity. In order to identify constrained airports among the sample of 172 airports, in the second step, we have eliminated airports with 5% peak hour volumes which are smaller than predefined threshold volumes by capacity class of the airport. For airports exceeding these threshold volumes, we have determined in a further step the relative time span during a year in which they operate at or near capacity by means of traffic ranking functions with traffic volumes and hours operating at each traffic volume in relative terms. Traffic volumes are expressed as shares of practical capacity and the number of hours as shares of the total operating time per year. Four capacity utilisation classes have thus been derived. Airports are in the lowest class A if their 5% peak hour volumes exceed 70% of the 5% peak hour capacity in more than one hour and in less than 5% of all operating hours of the year. Airports belonging to the highest capacity utilisation class D are those with 5% peak hour volumes exceeding 70% of the 5% capacity in more than 65% of all operating hours. The class A, with 137 airports, is the biggest among the four capacity utilisation classes. Class A airports experience some congestion in peak hours, however, only for short

Constrained and under-utilised airports Chapter | 4

103

periods in up to about 300 400 hours per year. As such, we have not classified them as congested airports, although we would not contradict operators of airports with traffic volumes approaching capacity in 5% of operating hours, if they think that their airport has a capacity problem. In our global analysis, we have classified the class B, C and D airports as constrained. Thus 35 airports have been more or less constrained in 2016, with London Heathrow, New York LaGuardia and Beijing being the most constrained airports of class D. Other predominant examples of constrained airports are Atlanta Hartsfield-Jackson, Chicago O’ Hare, Los Angeles and Newark in North America; Mexico City in South America; Singapore, Bangkok, Delhi, Jakarta and Hong Kong in Asia; Munich, London Gatwick and Istanbul (IST) in Europe and Dubai in the Middle East. The 35 congested airports are the main airports of the global network; they handled a traffic volume of 13.3 million aircraft movements, corresponding to a share of nearly 19% of the global traffic in 2016. The global constraint analysis has clearly confirmed that traffic is concentrated on a relatively small number of airports. Furthermore, these are the most important nodes of the network, and they are to a great extent confronted with severe capacity problems. On the other hand, the great majority of over 4000 airports with scheduled traffic are not utilised to a high degree. Most of them have enough capacity for handling the traffic to be expected in the next decades. However, those which are already constrained today face growing problems in keeping up with the traffic forecast by industry and intergovernmental institutions like ICAO. In Part 3 of the book, we will look at how the airports expected to be constrained in the future will be able to manage the traffic growth.

References Berster, P., Gelhausen, M.C., Wilken, D., 2012. Demand and supply development patterns of low-cost carriers in Africa, America, Europe, Australia and Asia. In: 16th Annual World Conference of the Air Transport Research Society, Tainan, 27 30 June 2010. Burghouwt, G., de Wit, J., 2005. Temporal configurations of European airline networks. J. Air Transp. Manage. 11, 185 198. Button, K., 2002. Debunking some common myths about airport hubs. J. Air Transp. Manage. 8, 177 188. Button, K., Haynes, K., Stough, R., 1998. Flying Into the Future. Air Transport Policy in the European Union. Edward Elgar, Cheltenham. Daft, J., Albers, S., 2012. A profitability analysis of low-cost long-haul flight operations. J. Air Transp. Manage. 19, 49 54. Dobruszkes, F., 2006. An analysis of European low-cost airlines and their networks. J. Transp. Geogr. 14, 249 264. Ehmer, H., Berster, P., Basedow, J., Jung, C., 2000. Liberalisierung im Luftverkehr Deutschlands Analyse und wettbewerbspolitische Empfehlungen, German Aerospace Center (DLR), Cologne.

104

PART | I Basic concepts

Eurocontrol, 2018. European Aviation in 2040 Challenges of Growth. Brussels, Belgium. Gelhausen, M.C., Berster, P., 2017. Domination of hub-and-spoke systems. In: Finger, M., Button, K. (Eds.), Air Transport Liberalization: A Critical Assessment. Edward Elgar Publishing, Cheltenham. Gelhausen, M.C., Berster, P., Wilken, D., 2013. Do airport capacity constraints have a serious impact on the future development of air traffic? J. Air Transp. Manage. 28, 3 13. Goetz, A.R., Vowles, T.M., 2009. The good, the bad, and the ugly: 30 years of US airline regulation. J. Transp. Geogr. 17 (4), 251 263. Grimme, W., 2011. The evolution of the low-cost carrier business model connections, hubbing and interlining. In: Fifth Annual Airneth Conference, Den Haag, 14 April 2011. Hooper, P., 2005. The environment for Southeast Asia’s new and evolving airlines. J. Air Transp. Manage. 11, 335 347. International Air Transport Association (IATA), 2008. IATA Economics Briefing No 9: Air Travel Demand. International Air Transport Association (IATA), Montreal. Malighetti, P., Paleari, S., Redondi, R., 2008. Connectivity of the European airport network: “self-help hubbing” and business implications. J. Air Transp. Manage. 14, 53 65. Moreira, M.E., O’Connell, J.F., Williams, G., 2011. The viability of long-haul, low-cost business models. J. Air Transp. Stud. 2 (1), 69 91. Morrell, P., 2008. Can long-haul low-cost airlines be successful? Res. Transp. Econ. 24 (1), 61 67. O’Connell, J.F., Williams, G., 2005. Passengers’ perceptions of low cost airlines and full service carriers: a case study involving Ryanair, Aer Lingus, Air Asia and Malaysia Airlines. J. Air Transp. Manage. 11, 259 272. Official Airline Guide (OAG), 2016. Market Analysis. Reed Travel Group, Dunstable. Pels, E., 2008. Airline network competition: full-service airlines, low-cost airlines and long-haul markets. Res. Transp. Econ. 24 (1), 68 74. Redondi, R., Malighetti, P., Paleari, S., 2011. Hub competition and travel times in the worldwide airport network. J. Transp. Geogr. 19, 1260 1271. Redondi, R., Malighetti, P., Paleari, S., 2012. De-hubbing of airports and their recovery patterns. J. Air Transp. Manage. 18, 1 4. Reynolds-Feighan, A., 2001. Traffic distribution in low-cost and full-service carrier networks. J. Air Transp. Manage. 7, 265 275. Sabre AirVision Market Intelligence (MI), 2016. Data Based on Market Information Data Tapes (MIDT). Sabre, Southlake, TX. Wensveen, J.G., Leick, R., 2009. The long-haul low-cost carrier: a unique business model. J. Air Transp. Manage. 15, 127 133. Wilken, D., Berster, P., Gelhausen, M.C., 2016. Analysis of demand structures on intercontinental routes to and from Europe with a view to identifying potential for new low-cost services. J. Air Transp. Manage. 56B, 79 90. Williams, G., 2001. Will Europe’s charter carriers be replaced by “no-frills” scheduled airlines? J. Air Transp. Manage. 7, 277 286. The World Bank, 2016. World Development Indicators. Washington, D.C.

Part II

Models for assessing mitigation strategies This part deals with the methodological background of the forecast model, which is applied in Part III. We start in Chapter 5, General strategies for mitigating airport capacity constraints, with the discussion of various strategies for mitigating airport capacity constraints and their relevance for the forecast. Chapter 6, Modelling future air passenger demand, presents the air passenger demand forecast model, followed by models for current and future airport capacity and aircraft size development, for example passengers per aircraft, in Chapters 7 9.

105

Chapter 5

General strategies for mitigating airport capacity constraints As described in Chapter 4, Constrained and under-utilised airports, global air traffic is very concentrated on a relatively small number of important airports the majority of which are facing capacity problems at present or will in the near future. In fact, the constraint analysis has shown that 137 airports have peak hour problems occurring in up to 5% of all operating hours of the year, while 35 of the most important airports, such as London Heathrow and Beijing, are more or less severely constrained over several or even many hours per day. Global air traffic is expected to continue to grow in the long run, although with a pace that differs greatly between Asia and the Middle East, on the one hand, and Europe and North America, on the other. While in Asia the demand has only begun to grow during the last few decades and is growing rapidly, demand in North America is more mature with a high level of propensity to fly, and demand development shows signs of saturation with a relatively low growth tendency compared to other markets. Demand in Europe is still growing, but with decreasing growth rates. These differing traffic growth expectations imply that future capacity problems will be less severe at major airports in North America than at Asian airports, such as Delhi, Jakarta, Mumbai and others. Due to further demand growth, on the one hand, and political difficulties with the population living in the vicinity of airports and opposing the development of new airport capacity, on the other, problems of overcoming airport constraints are most likely to be aggravated in the future, primarily in Europe.

5.1

Overview: typology of mitigation measures

Solutions to the capacity problem will vary from airport to airport and from region to region, depending on the severity and type of the capacity bottleneck, the financial situation of the airport owner and region or state and the regulatory framework governing in the region or state of the airport. A whole range of technological investment and non-investment options does theoretically exist. Furthermore, a spectrum of demand and supply management measures may be applied, ranging from pure administrative measures, such Airport Capacity Constraints and Strategies for Mitigation. DOI: https://doi.org/10.1016/B978-0-12-812657-8.00005-1 © 2020 Elsevier Inc. All rights reserved.

107

108

PART | II Models for assessing mitigation strategies

as regulations, to hybrid measures, like slot coordination with secondary slot trading, to market-based options, like congestion pricing schemes and primary slot trading. Therefore management measures, in particular, do not primarily aim at increasing the capacity, but rather at optimising traffic flows or increasing traffic volumes within given capacities, thus improving the utilisation of the existing infrastructure. Fig. 5.1 shows a typology of options for mitigating the negative effects of airport capacity constraints. With this typology, we do not pretend to propose an all-encompassing structure of potential measures but rather aim to state and describe relevant options which are applied or maybe practical candidates for future application. Some of the options listed have not yet been commonly applied, such as congestion pricing of aircraft movements, since landing fees are regulated in many countries on grounds of nondiscrimination of airport users and prevention of the abuse of market power of airports operators. Broadly, we can distinguish between investment and non-investment options. Investment options may be subdivided again into options of investing directly into airport infrastructure, such as runways in particular, and indirectly into airport-related infrastructure such as rail terminals at airports, thus connecting the airside with the regional and/or the intercity rail system. Connecting Mitigation measures

Infrastructure investment options

Direct investment

Indirect investment

– Additional airport – Additional airportinfrastructure: related infrastructure, – Runways for example rail – Taxiways terminal – Ramps – Terminals – High–speed rail system – ATC infrastructure

Non-investment options

Organisational options

Regulatory options

Pricing options

– Stronger use of off-peak times

– IATA slot coordination

– Congestion pricing

– Diversion to less congested airports – Use of bigger aircraft – Raising of load factors

– Exclusion of traffic segments – Perimeter rule

– Differential pricing of congested and non-congested airports – Primary and secondary slot trading

– Changing ATC rules for example reduction of separation minima in take-off and landing procedures

FIGURE 5.1 Options to mitigate negative effects of airport capacity constraints. IATA, International Air Transport Organization; ATC, air traffic control.

General strategies for mitigating airport capacity constraints Chapter | 5

109

the airport with the regional train network would change the modal split of passengers from their true journey origin to the airport and from the airport to the final destination, while direct access to intercity trains would both enlarge the catchment area of airports and reduce the demand for shortdistance flights. Typically, for airports with up to two or three runways, the option of directly investing in runway systems would clearly bring the greatest capacity gains of all options. However, in some regions of the world, such as in the London area, no enlargement of airport capacity has been realised in the last few decades, in spite of many political propositions, whereas other regions, such as China or the Middle East, have experienced rapid extensions of runway systems or the construction of new airports. In many other instances, in particular, in Europe, new facilities, especially runways, have been added only with great delays, caused by the strong opposition of the population living in the vicinity of airports. These delays have been in the order of up to 20 plus years, a time span in which air traffic may have doubled. In such situations, public authorities, airport operators, airlines and air navigation service providers have to look for measures which optimise the throughput of facilities rather than increase the capacity, by applying in broad terms non-investment options, like operational, regulatory or pricing options. Operational measures aim at better use of the airport infrastructure available and affect the planning and operations of aircraft services at airports. Examples are diverting flights to off-peak times and/or to nearby airports with less congestion, the deployment of aircraft with higher seat capacity in order to increase the number of passengers without having to raise flight frequency and raising load factors of flights. Regulatory measures have been introduced by public authorities to optimise traffic flows in capacity constrained conditions or to allow or prioritise certain types of traffic like scheduled flights, and exclude others, for instance, general aviation at main airports. A famous example of air traffic regulation at constrained airports is the widely applied International Air Transport Association (IATA) slot coordination, by which flight requests of airlines are allocated to slots (points in time for a complete take-off or landing procedure) following administrative priority rules, in particular the socalled grandfather right. Another example is the perimeter rule, applied at some airports, which excludes certain traffic segments such as international flights, with the aim of directing traffic segments to certain airports. The IATA slot coordination deals with capacity scarcity as a regulatory measure without applying market-based measures, although slots at constrained airports have a high market value and are traded, wherever allowed, at high prices among airlines, after they have been allocated in the first place (secondary slot trading). Pricing options, in contrast, aim to optimise and maximise traffic flows by allowing airlines to trade with slots and airport operators to charge different prices for aircraft operations depending on the

110

PART | II Models for assessing mitigation strategies

degree of constraint. Primary slot trading (without any preceding slot coordination regulation) and congestion pricing are, however, still theoretical options and not yet applied in air transportation. Airports charge airlines with weight- and passenger-related landing fees per flight. Since these fees have to be cost-related and adopted by government agencies in many countries, especially in Europe, airports often have no means to raise fees in relation to capacity scarcity. Nevertheless, regulatory and pricing options are not measures to directly increase airport capacity but rather to optimise traffic flows in constrained situations. However, there is one exception possibly future regulations in air traffic control will permit reduced separation minima of aircraft in the take-off and landing procedure. Such regulations would rely on technological progress of surveying aircraft more accurately in the controlled airspace, for example, by satellite navigation. Since new air traffic control rules of shorter separation minima do not yet exist and are not likely to be introduced in the near future, the existing rules are expected to remain and will, therefore, influence traffic flow capacity in the same way as in the past. In Part III of this book, we will look further into the present and future airport capacity situation and forecast, on a global scale, air traffic, airport capacities and mitigation effects of some selected measures on capacity constrained airports. In the following section we will discuss these measures in more detail.

5.2

Investment option: new runways

In an environment of continued air traffic growth, airport operators have an inherent interest in meeting the demand by providing sufficient capacity of all functional elements of the airport. While capacity enlargements of groundside facilities, terminals and airside facilities like ramps, aprons, buildings and areas for servicing aircraft belong more or less to the normal business of managing airports, adding new runways is, on the other hand, a rare event which often requires lengthy planning procedures with strong involvement of the public. Even if national politics may be favouring the provision of sufficient airport capacity, regional politics and the population living in the vicinity of airports are likely to oppose in Western states more than in Asia plans to increase airport capacity, since they fear more gaseous emissions and noise caused by the additional traffic which a new runway would attract. An investment in new runways normally brings the highest capacity gain, compared to all other options listed in Fig. 5.1. The costs of building a new runway are also much higher than those of the non-investment options, although building a new rail terminal near the air terminal of an airport may easily exceed the costs of a new runway. Nevertheless, the long delays in realising a new runway, which typically accompany the expansion process of

General strategies for mitigating airport capacity constraints Chapter | 5

111

airports in Europe and other countries, are often not caused by a lack of financial means, but by political and public resistance to endorse the capacity enhancement. In addition to public investment, in particular in the United States, private investment has been available to mitigate the infrastructure gap at many airports around the world. A predominant example of such a problem of delaying investments in airport infrastructure is the project of increasing the capacity of airports in the London area. Although London Heathrow airport has been operating at the capacity limit for decades, it was only in 2018 that the UK government reached a decision to back a third runway at that airport. The opening of the runway is foreseen for the year 2027. However, it is likely that the population of communities surrounding the airport and other interest groups will oppose the project and fight in courts against the realisation, with the result that the runway might not open until a later date. If built, the airport would increase its capacity by around 260,000 aircraft movements per year (Airports Commission, 2015) so that the traffic volume could reach a level of about 740,000 movements per year. The number of passengers could grow to around 130 million from about 76 million in 2017. The hourly capacity gain of the new runway would thus be in the order of 40 50 aircraft movements, from a coordinated capacity of 80 90 movements in 2018. The London Heathrow case may not be a typical example. Although long delays in realising new runway projects are common in Europe, they are, in contrast, rather short and thus of lesser importance in Asian countries, where a great number of new airports and airport extensions have been built in recent years. Based on official data sources of airport layout plans, such as Digital Aeronautical Flight Information Files (DAFIF) (2016), unofficial data like OurAirports (2019) and various internal data sources, we have looked into the development of new runways worldwide and in major world regions between 2008 and 2016 (see Table 5.1). Some of the original data of airports in world regions had to be complemented and corrected, since the data sources have listed series of airports, in particular in Asia, without information regarding the number of runways. According to the data available, the 4054 airports of the global air traffic network provided a capacity of around 5515 runways in 2016, with the great majority of airports equipped with a single runway. The overall average number of runways per airport is just 1.36. Since 2008, when the global network counted 3790 airports, 264 airports have been added to the network until 2016. In the same time span, network capacity was enlarged by around 405 runways, either at new airports or as airport extensions, most of them in Asia (almost 190) and North America (almost 160). The data available reveal that the number of runways in the Middle East has slightly decreased from 2008 to 2016, although traffic in that region has gone up by almost 80% and thus significantly more than the total traffic (17%). The reason is that some airports were listed in 2008 that are no longer listed in 2016, when other

TABLE 5.1 Runway extensions at airports of the global network between 2008 and 2016 [Official Airline Guide (OAG), 2016; Digital Aeronautical Flight Information Files (DAFIF), 2016; OurAirports, 2019]. All airports Number of airports

Number of flights

Number of runways

Region 2008

2016

Δ (%)

2008

2016

Δ (%)

2008

2016

Δ (%)

Africa

359

388

8.1

905,039

1,206,038

33.3

448

479

6.9

Asia

749

921

23.0

5,677,855

10,149,228

78.8

832

1024

23.1

Europe

677

685

1.2

7,771,604

8,234,691

6.0

922

931

1.0

Middle East

113

111

21.8

694,164

1,239,500

78.6

152

144

25.3

North America

971

1018

4.8

11,480,093

10,181,932

211.3

1614

1773

9.9

South America

522

530

1.5

2,664,643

3,312,908

24.3

598

606

1.3

Southwest Pacific

399

401

0.5

1,081,085

1,129,757

4.5

542

556

2.6

World

3790

4054

7.0

30,274,483

35,454,054

17.1

5108

5513

7.9

100 top-ranking airports Number of airports

Number of flights

Number of runways

Region 2008

2016

Δ (%)

2008

2016

Δ (%)

2008

2016

Δ (%)

Africa

1

1

0.0

100,183

104,098

3.9

2

2

0.0

Asia

18

29

61.1

2,147,244

4,497,591

109.5

32

59

84.4

Europe

31

26

216.1

3,887,681

3,658,637

25.9

80

71

211.3

Middle East

1

4

300.0

113,283

514,973

354.6

2

8

300.0

North America

41

33

219.5

7,189,299

6,179,896

214.0

155

136

212.3

South America

5

4

220.0

489,677

550,290

12.4

10

8

220.0

Southwest Pacific

3

3

0.0

320,781

372,995

16.3

7

7

0.0

World

100

100

0.0

14,248,148

15,878,480

11.4

288

291

1.0

Other airports Number of airports

Number of flights

Number of runways

Region 2008

2016

Δ (%)

2008

2016

Δ (%)

2008

2016

Δ (%)

Africa

358

387

8.1

804,856

1,101,940

36.9

446

477

7.0

Asia

731

892

22.0

3,530,611

5,651,637

60.1

800

965

20.6

Europe

646

659

2.0

3,883,923

4,576,054

17.8

842

860

2.1

Middle East

112

107

24.5

580,881

724,527

24.7

150

136

29.3

North America

930

985

5.9

4,290,794

4,002,036

26.7

1459

1637

12.2

South America

517

526

1.7

2,174,966

2,762,618

27.0

588

598

1.7

Southwest Pacific

396

398

0.5

760,304

756,762

20.5

535

549

2.6

World

3690

3954

7.2

16,026,335

19,575,574

22.1

4820

5222

8.3

General strategies for mitigating airport capacity constraints Chapter | 5

115

airports entered the list instead. Europe, the Middle East, South America and the Southwest Pacific are the world regions that have seen hardly any capacity growth in terms of additional runways, quite in contrast to Asia and North America. An interesting question is which category of airports has gained most of the new runway capacity. Have the major airports added more capacity than the secondary airports or did the latter and new airports add more runways? In discussing the global air traffic distribution in Chapter 4, Constrained and under-utilised airports, we have seen that traffic has been concentrated on a relatively small number of important airports while the great majority of airports handle rather small traffic volumes. In fact, in the year 2016 about 45% of the global air traffic was concentrated on the 100 top-ranking airports, which represent just 2.5% of all airports. We have, therefore, checked the runway extensions of the two groups of airports separately. The result is rather surprising insofar as the top 100 airports have increased their runway capacity only marginally by adding just three new runways, whereas the other 3954 smaller airports have extended the capacity by about 400 new runways. This group includes new airports as well. It should be added that the top 100 airports in 2016 are not identical to those in 2008; some airports of the 2008 list, the traffic of which grew well below average, were no longer among the top 100 airports in 2016, while others with a stronger traffic growth entered the 2016 list. The traffic concentration was such that each runway of the top 100 airports handled on average about 55,000 flights in 2016, whereas for the other 3954 airports the equivalent traffic volume was less than 4000 flights. Comparing the traffic growth between the two airport groups, we see that traffic at the top 100 airports grew by 11% between 2008 and 2016, while the traffic at all other airports increased twice as much by 22%. One would perhaps assume that traffic grew similarly in the two groups. The reason for the relatively low growth at the major 100 airports is that a capacity shortage at a great number of these airports hindered them from offering free slots, so that they were unable to participate in the general air traffic growth. We have identified 30 of the top 100 airports belonging to the capacity utilisation classes B, C and D (see Section 4.5), which are more or less severely capacity constrained. They handled a traffic volume of over 6.2 million flights, which represent nearly 40% of the traffic of the top 100 airports. In addition, there are more airports with capacity bottlenecks at peak times, belonging to the capacity utilisation class A. We estimate, therefore, that at about 40 of the top 100 airports the traffic making up nearly half of the total traffic at the top 100 airports cannot be enhanced due to lack of capacity. To further accommodate traffic at these airports, additional runways would be needed. The small number of new runways added at the major 100 airports is indicative of the difficult circumstances at these airports to enlarge the capacity by means of new airside infrastructure.

116

PART | II Models for assessing mitigation strategies

5.3 Rerouting traffic to under-utilised airports and/or using more off-peak hours The option to reroute traffic from constrained to non-constrained airports seems to be at first glance a relief measure easy to realise. A look at some important airports shows, however, that this measure has not been applied to a great extent. Major airports often do not only serve the traffic originating in their catchment area but are also used as points of transfer traffic, in particular by those air carriers which are concentrating operations there, like full-service network carriers. The hubbing function of these airports interrelates feeder services with other flights, especially long-distance and intercontinental flights. Because of this interrelationship, schedules of both types of flights are coordinated and airlines have no interest in negatively affecting the arrival and departure structure by taking out flights as they would immediately lose market segments. Rerouting traffic from hub airports to secondary airports is, therefore, more a theoretical than a practised measure. Major airports with congestion problems serve primarily origin destination traffic like that of low-cost carriers. Incoming and outgoing flights do not depend on each other, and some of them may be diverted to other airports nearby. Experience has shown, however, that guest airlines, in particular, are not willing to divert operations, since they want to serve as much as possible the demand from or into the catchment area of such airports. This may be less so with airports having a smaller catchment area, but then these airports are less likely congested. Even if rerouting of flights to less congested airports may be regarded as a solution to mitigating constraint problems the question is which institution would be in charge of asking airlines to reduce frequencies at congested airports and diverting flights to other airports. Airlines could, of course, voluntarily renounce serving an airport as planned, for instance, in the case of terminating flights in non-sustainable markets. Otherwise, this would be a very unlikely event, since an abandonment of flights would be equal to discontinuing serving lucrative markets. Anyway, there is no regulatory institution which could prevent airlines from continuing services if they have performed in the past, according to prevailing regulations. On the contrary, whenever the IATA-type slot coordination is applied as an official procedure at congested airports, as for instance in Europe according to EU-Regulation 95/93 at Level 3 airports (see Chapter 2: Concepts of capacity and methods of estimation), then incumbent airlines have a ‘grandfather right’ to plan the same number and type of services for the next season as are operated in the present season. A diversion to less congested airports would only be feasible in the case of a coordinated airport system in a metropolitan area, where the slot coordinator might ask airlines applying for new services to offer them at a secondary airport. Those airlines would, however, have the choice of denying such requests. Most likely they would opt for

General strategies for mitigating airport capacity constraints Chapter | 5

117

another measure that is offering flights with more seats. Altogether, in 2016 almost 200 Level 3 airports worldwide have been, by definition of the IATA Worldwide Slot Guidelines, congested, either in peak times or over longer time periods. A number of these airports are rather small airports with mainly touristic traffic in summertime, where typically the terminal capacity is insufficient for handling waiting and incoming passengers in some peak periods per week. The greater number, however, are major airports with often important hubbing functions, where demand over longer periods exceeds the capacity, that is runway capacity. With the exception of the US airports the United States do not apply the IATA-type slot coordination all top-ranking airports with capacity problems are slot coordinated. We may conclude that the option to divert flights from congested to uncongested airports nearby may be seen as a ‘nolens volens measure’ which requires some kind of compulsory action from the regulatory side; it is typically not a voluntary airline option. Of a similar nature is the temporal diversion of flights to off-peak periods at partly congested airports. Only in exceptional cases do airlines have an interest in accepting shifting flights from their originally scheduled time. At hub airports, in particular, they need coordinated slots because of the interdependency of flight arrivals and departures. At Level 3 airports, however, slot coordinators have most likely applied the measure of filling up off-peak periods with flights by allocating new slot requests for unavailable peak time slots at off-peak times. It may well be that some airlines did not accept the diverted slot offer and withdrew, however, looking at hourly traffic patterns at congested airports we can observe that formerly off-peak periods with capacity reserves have been filled with flights over time, when traffic demand was growing and airlines were in need of additional slots. We cannot prove whether slot coordinators shifted arrival or departure times of slot requests for peak times to off-peak times, nevertheless at Level 3 airports off-peak times were filled with additional flights. However, there is a strong indication that slot coordination has been responsible for the new flight schedules. We have selected two examples of traffic patterns at London Gatwick and Frankfurt airport in different years to illustrate these schedule changes over time. Fig. 5.2 shows the daily traffic distribution at London Gatwick (LGW) during a peak week of the years 2000, 2008 and 2016. The annual traffic volume has grown from over 200,000 aircraft movements in 2000 to more than 233,000 in 2008 and further to nearly 274,000 movements in 2016. The annual traffic growth is mirrored in the change of the daily traffic pattern from 2000 to 2016. While in 2000 the hourly traffic varied around 40 aircraft movements, with a longer off-peak period in the afternoon, hourly traffic increased to about 40 45 movements in 2008, again with a pronounced variation, and then further to around 50 movements in 2016, with three peak times and traffic reaching values of 55 movements. Three off-peak times can

118

PART | II Models for assessing mitigation strategies

Aircraft movements per hour

60 50 40 30 20 10 0

0

6 12 18 0

6 12 18 0

Monday

Tuesday

6 12 18 0 Wednesday LGW 2000

6 12 18 0 Thursday LGW 2008

6 12 18 0 Friday

6 12 18 0 Saturday

6 12 18 Sunday

LGW 2016

FIGURE 5.2 Daily traffic pattern at London Gatwick (LGW) airport during a peak week in 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016].

be seen which are shorter and less pronounced than in former years. Along with the traffic growth, slots in off-peak periods have been used more and more. Since the demand for slots in London Gatwick has exceeded the capacity of the single runway for years, the airport is slot coordinated. There are only a few airports worldwide that reach hourly volumes on one runway as high as 50 aircraft movements and more in slot coordinated conditions. Frankfurt (FRA) airport is more of a major hub airport than London Gatwick with a dominant home carrier (Lufthansa). Due to the hubbing, the daily traffic pattern shows typical peaks during the day when either feeder flights or intercontinental and other flights with corresponding passengers arrive and depart in banks in order to keep the transfer time as short as possible. FRA was equipped with three runways in 2000 and 2008 and added an additional runway in 2011, which is used mainly for landings. Traffic grew from over 423,000 aircraft movements in 2000 to more than 469,000 in 2008 and went down slightly to around 455,000 movements in 2016. Fig. 5.3 shows the development of hourly traffic distribution at FRA airport during a peak week from 2000 to 2008 and to 2016. In 2008 the airport operated more at the capacity limit than in 2000, and again less in 2016 when the declared capacity had gone up with the new runway from about 80 to 100 aircraft movements per hour, however, at the price of a strict night curfew. Thanks to the capacity increase, the airport was able to build up a long peak in the morning and two additional peak periods in the afternoon. Whereas the peak traffic volumes could not go up from 2000 to 2008, due to capacity limits, the off-peak periods were partly filled with additional flights resulting from traffic growth in that time span. In the following period to

General strategies for mitigating airport capacity constraints Chapter | 5

119

Aircraft movements per hour

120 100 80 60 40 20 0

0

6 12 18 0

6 12 18 0

Monday

Tuesday

6 12 18 0 Wednesday FRA 2000

6 12 18 0 Thursday FRA 2008

6 12 18 0 Friday

6 12 18 0 Saturday

6 12 18 Sunday

FRA 2016

FIGURE 5.3 Daily traffic pattern at Frankfurt (FRA) airport during a peak week in 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016].

2016 annual traffic slightly decreased, however, hourly traffic in peak times increased, whereas traffic in two off-peak times in the afternoon decreased sharply, thus allowing for a hub-based banking of flight arrivals and departures during the day.

5.4

Raising seat capacity and load factor per flight

If, on the one hand, the overall demand for air transport grows, but airport capacity is, on the other hand, no longer available at congested airports, we would assume that airlines offer flights with more seats in order to cope with the demand. Analysis of developments of frequencies and average seat capacity at congested and not yet congested airports has shown that the hypothesis of bigger aircraft in congested situations is valid in most instances but not at all airports. Using bigger aircraft and aircraft with higher seat density are measures that airlines use to varying degrees depending on factors like level of airport congestion, airline fleet, network structure, competition with other airlines and so on. We would assume that airlines in liberalised markets wanting to serve a growing market increase their capacity by offering more seats on existing as well as new routes. At congested airports airlines would do so by deploying bigger aircraft and at uncongested airports by first increasing the number of flights. Capacity constraints would prevent airlines from increasing frequencies, whereas at airports with capacity surplus airlines would prefer to offer more flights in order to better comply with the needs of travellers, in particular business travellers. Our analysis of the development of the average seat

120

PART | II Models for assessing mitigation strategies

capacity of flights offered should then differentiate between more and less congested and uncongested airports, respectively. We have, therefore, subdivided the total population of airports (4054 in 2016) into the 100 top-ranking and other airports, as done previously in Section 5.2 to examine runway extensions of airports worldwide. We have derived the average values of seat capacity per flight, the number of passengers per flight and of the load factor per flight for the years 2008 and 2016 (see Table 5.2). The values of seats and passengers per flight as shown in Table 5.2 are lower by a small margin than the true values, since the number of flights (see Table 5.1) includes in addition to passenger flights cargo flights, which should be excluded if possible when the number of passengers and seats are related to the number of flights. Cargo flights, however, form only a small fraction of all flights, and the Official Airline Guide database does not even include all cargo flights, in particular not ad hoc cargo flights, so that the calculated values are roughly correct but not exact. Since we are more interested in the development of seat capacity and load factor, the lack of accuracy is of minor importance. The average number of seats per flight has grown globally from 105 seats in 2000 to 118 seats in 2008 and continued to rise to 138 seats per flight in 2016, hence in 16 years by 33 seats or by 31% in relative terms. This development proves that airlines have deployed, in general, bigger aircraft into the market in order to satisfy the growing demand. As can be seen in Table 5.2 and Fig. 5.4, this option has been chosen at the 100 major airports as well as at all other airports. Average seat capacity at the 100 top-ranking airports has risen from 132 in 2008 to 156 seats in 2016 and at all other airports from 105 to 123 seats per flight. At the 100 top-ranking airports, of which about 40 airports have been more or less capacity constrained, flights offered in 2016 had thus 33 seats more than at all other airports. In terms of passenger throughput efficiency, this means that at the major airports 127 passengers were transported per flight, whereas at a great number of all other airports 98 passengers, and thus 29 passengers less, were on average onboard a flight. Nevertheless, at both classes of airports, at the 100 top-ranking airports with high traffic volumes and more or less severe capacity problems, and at all other airports, many of them without capacity constraints, the average number of seats per flight has increased by 20 seats or by 2.5 seats per year in eight years’ time span. And there is no sign of saturation in this development, as can be seen in Fig. 5.4. As may be expected, the absolute value of seat capacity and the development over time vary with the region and airport. The smallest aircraft in terms of seats per flight are operated in North America in 2008 flights offered only 94 seats and in 2016 109 seats whereas flights with the highest seat capacity can be found in the Middle East 167 seats in 2008 and 192 seats in 2016. If we look at the eight airports in the Middle East belonging to the 100 top-ranking airports then the average seat capacity rises to 231

TABLE 5.2 Average numbers of seats and passengers per flight and load factors of global air traffic for the years 2008 and 2016 [Official Airline Guide (OAG), 2016; Sabre AirVision Market Intelligence (MI), 2016 ]. All airports Seats per flight

Passengers per flight

Load factor

Region 2008

2016

Δ (%)

2008

2016

Δ (%)

2008 (%)

2016 (%)

Δ (%)

Africa

117

120

2.6

79

89

12.6

67.4

74.0

9.7

Asia

159

163

2.6

110

133

20.7

69.3

81.6

17.6

Europe

126

147

17.0

91

119

31.1

72.0

80.8

12.1

Middle East

167

192

15.3

122

138

13.1

73.3

71.9

21.9

North America

94

109

16.4

67

89

32.8

71.8

82.0

14.1

South America

104

120

15.5

75

98

29.9

72.3

81.3

12.4

Southwest Pacific

105

124

17.4

79

97

23.9

74.5

78.6

5.5

World

118

138

17.2

84

111

32.6

71.3

80.7

13.1

100 top-ranking airports Seats per flight

Passengers per flight

Load factor

Region 2008

2016

Δ (%)

2008

2016

Δ (%)

2008 (%)

2016 (%)

Δ (%)

Africa

131

141

7.7

86

109

27.4

65.5

77.5

18.3

Asia

190

183

23.6

133

149

12.4

69.9

81.5

16.6

Europe

140

164

17.2

100

133

32.8

71.4

80.9

13.3

Middle East

213

231

8.4

161

177

10.2

75.3

76.5

1.6

North America

109

126

15.4

79

105

32.2

72.7

83.3

14.6

South America

119

146

22.2

86

118

37.0

72.1

80.9

12.2

Southwest Pacific

156

169

8.1

120

133

10.4

77.1

78.8

2.2

World

132

156

18.2

95

127

34.2

71.8

81.6

13.6

Other airports Seats per flight

Passengers per flight

Load factor

Region 2008

2016

Δ (%)

2008

2016

Δ (%)

Africa

116

118

2.4

78

87

11.3

67.7

73.6

8.7

Asia

139

146

4.9

96

119

24.4

68.9

81.6

18.5

Europe

112

134

19.7

81

108

32.5

72.8

80.6

10.7

Middle East

158

165

4.4

115

111

23.5

72.8

67.3

27.5

North America

68

83

22.3

47

66

38.9

69.5

79.0

13.6

South America

100

115

14.4

73

94

28.7

72.3

81.4

12.5

Southwest Pacific

84

102

21.0

61

80

31.0

72.5

78.5

8.2

World

105

123

17.4

74

98

32.3

70.7

79.7

12.7

2008 (%)

2016 (%)

Δ (%)

124

PART | II Models for assessing mitigation strategies

180 160

Seats per flight

140 120 100 80 60 40 20 0

2008

2009

2010

All airports

2011

2012

Other airports

2013

2014

2015

2016

100 top-ranking airports

FIGURE 5.4 Development of average seat capacity per flight in the global network of airports, of the 100 top-ranking airports and of all other airports from 2008 to 2016 [Official Airline Guide (OAG), 2016; Sabre AirVision Market Intelligence (MI), 2016].

seats, 75 seats more than on average at all 100 top-ranking airports. The growth in seat capacity has been quite different in world regions as well. While on average the number of seats offered per flight has gone up by 17.2% from 2008 to 2016, Africa and Asia have experienced almost no growth. At the same time, the traffic in Asia has grown by almost 80%, much more than the total traffic, which has increased by just 17.1%. Based on the linear growth of average seat capacity per flight over the period of 2008 16, as shown in Fig. 5.4, we could assume a continuation of this trend into the future. The overall development hides, however, the variation in regions and at individual airports. As there are some regions without the growth of average seat capacity and other regions with stronger than average growth, there are airports where flights have been operated with rather a constant aircraft size, while at most airports average seat capacity of flights has increased over time. A former study of the development of flight seat capacity (Berster et al., 2015) has shown that in almost 80% of constrained airports, as well as unconstrained airports, airlines have increased the seat capacity of flights. Reasons other than airport congestion played a decisive role in increasing the average seat capacity at unconstrained airports. If sustainable levels of load factor and frequency are achieved and demand continues to grow, airlines have an economic interest in scheduling aircraft with higher seat capacity at lower unit costs rather than increasing frequency, in constrained airport conditions even more than in unconstrained

General strategies for mitigating airport capacity constraints Chapter | 5

125

conditions. There has been a clear tendency to employ bigger aircraft types on longer routes. We have seen that average flight distances have gone up at both congested and uncongested airports. An increase in average flight distance may, therefore, be regarded as a factor describing the tendency of employing aircraft with higher seat capacity at lower unit cost. Other factors may cause airlines to schedule otherwise, depending, for instance, on local conditions, aircraft availability and airspace and airline regulation. For forecasting seat capacity of flights on a network level, a model has been developed which will be described in Chapter 9, Modelling future development of the average number of passengers per flight. Table 5.2 shows a third passenger throughput efficiency indicator: average load factors. Increasing load factors to a possible maximum is a general objective of airlines’ fleet, route and frequency planning, regardless of airport constraints. We have shown this efficiency indicator in connection with the two other indicators to demonstrate that the past development of load factors has already led to values exceeding 80% on average at all airports worldwide and at the 100 top-ranking airports and reaching 80% at all other airports (see Fig. 5.5). We can see that average load factors are rather similar in the two groups of airports, that is the 100 top-ranking airports and the many other airports with smaller traffic volumes. Fig. 5.5 also shows that the development of average network-wide load factors follows a linear trend with no indication of approaching a saturation value. Because of the absolute limit of 100%, an extrapolation of the trend must, however, approach

100% 90% 80%

Load factor

70% 60% 50% 40% 30% 20% 10% 0%

2008

2009

2010

All airports

2011

2012

Other airports

2013

2014

2015

2016

100 top-ranking airports

FIGURE 5.5 Development of average load factors in the global network of airports, of the 100 top-ranking airports and of all other airports from 2008 to 2016 [Official Airline Guide(OAG), 2016; Sabre AirVision Market Intelligence (MI), 2016].

126

PART | II Models for assessing mitigation strategies

saturation at some point in time, which we would assume to be near values of around 90% 95%. Low-cost carriers have been achieving average load factors of more than 80%, some airlines in peak months of almost 90%, thus reaching values which cannot be raised much higher, since load factors vary during days, weeks and seasons. If top load factors of 100% are reached in peak periods without denying too many passengers, then average load factors can hardly exceed values of 90% 95%. This means that the capacity reserve of free seats in flights has already been exhausted to a great extent. Nevertheless, full-service network carriers, in particular, have still some possibilities of raising load factors by some percentage points.

5.5 Case study: development of mitigation measures at three example airports: San Diego (SAN), London Heathrow (LHR) and Beijing (PEK) After our examination of diverse optional measures of mitigating capacity constraints at airports on a global scale, we concentrate in this section on the three example airports San Diego (SAN) in the United States, London Heathrow (LHR) in the United Kingdom and Beijing (PEK), the capital airport of China. Traffic and constraint characteristics of these airports were discussed in Sections 3.5 and 4.6. The three airports differ in runway capacity. SAN is a single runway airport with an hourly capacity varying between 48 aircraft movements in instrument flight rules (IFR) conditions and 57 in optimum conditions. LHR has two independent parallel runways which are used for environmental reasons in a segregated mode, one runway for take-offs and the other for landings. Since LHR is slot coordinated, the declared capacity determines the maximum slot offer. In 2016 the declared capacity reached values of up to 88 movements in some peak hours; traffic demand has been using the slot offer to the full degree during most hours of the day. PEK airport is the only airport among the three which had a substantial capacity growth by means of a new runway, which was added to the two existing runways in October 2007. With three independent parallel runways, PEK has a theoretical capacity about three times as high as SAN and 50% 70% higher than LHR. This capacity, however, cannot be used to the full degree, since limitations exist in using all three runways simultaneously and in the controlled airspace in the Beijing area so that the practical capacity equal to the declared capacity reaches values of up to 88 aircraft movements per hour according to the Air Traffic Management Bureau of the Civil Aviation Administration of China. The capacity utilisation of the example airports can be seen in Figs 5.6 5.8. As described in Section 5.3, slot coordinators have applied the measure of diverting slot requests to off-peak periods and thus filling up these time

General strategies for mitigating airport capacity constraints Chapter | 5

127

50 Aircraft movements per hour

45 40 35 30 25 20 15 10 5 0

0

6 12 18 0

6 12 18 0

6 12 18 0

6 12 18 0

Monday

Tuesday

Wednesday

Thursday

SAN 2000

SAN 2008

6 12 18 0 Friday

6 12 18 0

6 12 18

Saturday

Sunday

SAN 2016

FIGURE 5.6 Hourly traffic pattern at San Diego (SAN) airport during a peak week in 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016].

100 Aircraft movements per hour

90 80 70 60 50 40 30 20 10 0

0

6 12 18 0

6 12 18 0

Monday

Tuesday

6 12 18 0

6 12 18 0

Wednesday LHR 2000

Thursday

LHR 2008

6 12 18 0 Friday

6 12 18 0 Saturday

6 12 18 Sunday

LHR 2016

FIGURE 5.7 Hourly traffic pattern at London Heathrow (LHR) airport during a peak week in 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016] .

spans with traffic at Level 3 airports. SAN airport is not a slot coordinated airport and, as we can see in Fig. 5.6, the traffic pattern at SAN airport is characterised by high peaks reaching near-capacity levels with 45 aircraft movements. However, it also has off-peak periods with lower traffic volumes following the peaks. The hourly traffic in the peak week has slightly increased from 2000 to 2016; however, the annual traffic has not grown during that 16-year period (see Fig. 3.13). A limiting movement cap, as a

128

PART | II Models for assessing mitigation strategies

100 Aircraft movements per hour

90 80 70 60 50 40 30 20 10 0

0

6 12 18 0

6 12 18 0

Monday

Tuesday

6 12 18 0 Wednesday

PEK 2000

6 12 18 0 Thursday

PEK 2008

6 12 18 0 Friday

6 12 18 0 Saturday

6 12 18 Sunday

PEK 2016

FIGURE 5.8 Hourly traffic pattern at Beijing (PEK) airport during a peak week in 2000, 2008 and 2016 [Official Airline Guide (OAG), 2016].

declared capacity would be, does not seem to have been reached yet in most operating hours. On the contrary, SAN still has a capacity reserve of up to ten aircraft movements per hour before the airport would reach the technical capacity of 57 movements in optimum conditions. Such a movement cap in the form of a declared capacity has existed for many years at LHR airport. Fig. 5.7 shows the hourly traffic pattern of LHR during a peak week in 2000, 2008 and 2016; as can be seen, over many daytime hours traffic volumes reach the limit of the declared capacity of up to 88 movements. Traffic variation is much smaller than at SAN airport, since the declared capacity has almost been fully used during the day, at least in 2008 and 2016. Off-peak periods, which still existed in 2000, were filled up with new flights, some of which were certainly requested for nearby peak periods. PEK airport has experienced a different development of traffic as is shown in Fig. 5.8. Due to the strong traffic growth from less than 200,000 aircraft movements in 2000 to over 400,000 in 2008, hourly peak volume during a peak week went up from around 40 movements to 80 movements. The new runway opened in 2007 allowed the peak hour traffic to double from 2000 to 2008. However, this was not so in the following period to 2016, when the annual traffic continued to grow to over 600,000 movements without any further capacity growth. Peak hour volumes increased slightly to 88 movements and hourly traffic variation decreased, an indication that off-peak periods were filled by slot coordination with additional flights in a similar way to the case at LHR. The global analysis has shown that airlines use the option of deploying aircraft with more seats widely in order to cope with growing passenger numbers at both constrained and unconstrained airports. As we have seen,

General strategies for mitigating airport capacity constraints Chapter | 5

129

250

Seats per flight

200

150

100

50

0

2008

2009

2010 LHR

2011

2012 SAN

2013

2014

2015

2016

PEK

FIGURE 5.9 Development of average seat capacity per flight at SAN, PEK and LHR airports from 2008 to 2016 [Official Airline Guide (OAG), 2016].

there are economic reasons for offering flights with higher seat capacities, besides the fact that at constrained airports the possibilities for increasing frequencies are limited or no longer exist. The trend of increasing seat capacity has been linear over more than 15 years; we may assume therefore that with further demand growth airlines will continue to mitigate the negative effects of capacity constraints with bigger aircraft. Here, we will look at the developments at the example airports. As Fig. 5.9 shows, seat capacity per flight has been rising more or less linearly from 2008 to 2016 at SAN and PEK airports, at SAN from 123 to 145 and at PEK from 183 to 195 seats per flight. In 2008 LHR airport already had an average seat capacity of 193 seats, which grew to 199 seats in 2012 and remained constant until 2016. It seems that at LHR a seat capacity of around 200 seats per flight forms an upper threshold, given the fleet composition at that airport. It might, however, well be that the further development will not stop there, and flights with more seats will be offered to and from LHR, depending on the demand growth and the duration of the status quo of airport capacity. With greater uncertainty regarding LHR, we may conclude that at the three example airports average seat capacity per flight will also continue to rise. Reaching saturation levels in the near future is rather unlikely if the economics of flight operation and capacity constraints in future will also positively influence seat capacity. Load factors, as a further measure of increasing passenger throughput without raising flight frequency, have already reached rather high levels of more than 80% at all airports and major airports worldwide. This is also true for PEK and SAN airport, as can be seen in Fig. 5.10. The average load factor of flights at LHR has also risen in the past

130

PART | II Models for assessing mitigation strategies

100% 90% 80%

Load factor

70% 60% 50% 40% 30% 20% 10% 0%

2008

2009

2010 LHR

2011

2012

2013

SAN

2014

2015

2016

PEK

FIGURE 5.10 Development of average load factor per flight at SAN, PEK and LHR airports from 2008 to 2016 [Official Airline Guide (OAG), 2016; Sabre AirVision Market Intelligence (MI), 2016].

and has reached a level of 80% in 2016, whereas SAN and PEK airports had already surpassed 80% in the year 2012. As has been discussed earlier, these values may still increase in the future; however, there seems to be a natural limit at around 90% 95%, depending on the type of market. Short-haul business traveller routes, for instance, have a higher demand variation during a day than typical touristic routes, with the consequence that maximum average load factors may be lower in the first instance than in the second.

5.6

Conclusion

Solutions to specific capacity problems vary from airport to airport, depending on the actual situation and capacity constraint. In general, we can distinguish between investment and non-investment measures, whereby the first category encompasses direct investment options, such as new runways or terminals, and indirect investments, such as new rail terminals near airports, and the latter group involves demand and supply management options, ranging from administrative or regulatory measures to hybrid to pure marketbased options. The typology of options to mitigate the negative effects of airport capacity constraints as given in Fig. 5.1 includes a selection of measures which are practiced or may be candidates for future application. We have discussed in particular: G G

adding new runways as investment option, rerouting flights to less utilised airports nearby,

General strategies for mitigating airport capacity constraints Chapter | 5 G G G

131

diverting flights to off-peak hours, raising seat capacity of flights, and raising load factors of flights.

Enlarging airport capacity by means of a new runway or even by building a new airport is the most effective way of creating new capacity for additional flights. This measure is, on the other hand, the most controversial one to realise since the neighbouring population and regional politics are often at least in Western-type states opposed to such projects for environmental reasons. The most famous example of this kind is the long planning process of adding new runway capacity in the London area. After decades of proposing various capacity extensions, a political decision has been reached in favour of a new runway at London Heathrow airport. Up to the year 2019 this runway had not been realised. We have seen that the global network of 4054 airports in 2016 has been enlarged by about 400 runways since 2008, the majority of them, however, in the network of 3954 secondary airports. The top 100 airports in terms of traffic volume have seen almost no runway capacity extension, although these airports handle about 45% of the total traffic, in many instances in severe capacity constraint circumstances. The fact that these airports were not in a position with a few exceptions to add capacity is indicative of the difficult political situation regarding airport expansions in many states. We have discussed the measure of rerouting flights to secondary airports nearby and concluded that this option is not widely used, in particular not at hub airports, because of the interrelationship of incoming and outgoing flights. Operators of touristic services might choose such bypass routes, however, only as a secondary choice. Of a similar nature is the temporal diversion of flights to off-peak periods at partly congested airports. Only in exceptional cases do airlines have an interest in accepting shifting flights from their originally scheduled time. This measure has been applied by slot coordinators at Level 3 airports when airlines requested slots at peak times but these were not available, because incumbents held them, and slot coordinators proposed alternative slots at off-peak times. The filling-up option will also be used in future whenever slot coordinated airports still have off-peak periods with free slots for new entrant and incumbent airlines. In our forecasts, too, we will apply this measure implicitly by assigning forecast traffic volumes to airports until capacity is reached. In contrast to the flight rerouting and diverting options, raising seat capacity per flight has been the most effective measure among the non-investment options to keep pace with growing bottlenecks at airports. The average number of seats per flight has grown globally from 105 seats in 2000 to 138 seats per flight in 2016, hence in 16 years by 33 seats or by 31% in relative terms. This option has been chosen at the 100 major airports as well as at all other airports. Average seat capacity at the 100 top-ranking airports has risen by 24 seats to 156 seats from 2008 to 2016 and at all other airports by 18 seats to

132

PART | II Models for assessing mitigation strategies

123 seats per flight. At the 100 top-ranking airports, of which about 40 airports have been more or less capacity constrained, flights offered in 2016 had thus 33 seats more than at all other airports. In terms of passenger throughput efficiency, this means that at the major airports 127 passengers were transported per flight, whereas at the great number of all other airports 98 passengers, and thus 29 passengers less, were on average onboard a flight. As can be seen in Fig. 5.4, there has been no sign of saturation in the growth of the number of seats offered per flight. Besides the fact that a lack of available slots hinders airlines from increasing frequencies at congested airports, it is an economic interest which leads airlines to operate bigger aircraft with lower unit costs, both at constrained and unconstrained airports. All airlines have an economic interest in raising the load factors of flights. In fact, they have already achieved rather high levels of passenger occupancy rates with average values of 80% and more. Average load factors have risen in the past rather linearly without approximating a saturation level. Because of the absolute limit of 100%, however, in the future they will approach saturation, which we assume to be in the order of 90% 95%. This means that average load factors will still rise, however, not by more than 10% 15%. Low-cost carriers will reach maximum levels probably earlier than fullservice network carriers, because the latter serve less homogeneous passenger markets with greater variation of daily traffic patterns. The analysis of mitigation measures has shown that two measures new runways as an investment option and raising seat capacity per flight as an operational option have proven to be more effective than the other measures. In discussing future mitigation measures (see Part III), we will put emphasis on these options rather than on rerouting flights to less congested airports and diverting flights to off-peak periods at the same airport. Forecasting the capacity enhancement by means of new runways and further growth of seat capacity will be model driven and will not rely on diverging subjective opinions regarding possible runway extensions, thus ensuring consistency of input assumptions.

References Airports Commission, 2015. Airports Commission: Final Report, July 2015. Berster, P., Gelhausen, M.C., Wilken, D., 2015. Is increasing aircraft size common practice at congested airports? J. Air Transp. Manage. 46, 40 48. Digital Aeronautical Flight Information Files (DAFIF), 2016 US Government, Washington D.C. Official Airline Guide (OAG), 2016. Market Analysis. Reed Travel Group, Dunstable. OurAirports, 2019. http://ourairports.com/airports (accessed 20 February 2019) Sabre AirVision Market Intelligence (MI), 2016. Data Based on Market Information Data Tapes (MIDT). Sabre, Southlake, TX.

Chapter 6

Modelling future air passenger demand Typically, one of the main objectives of producing air transport forecasts is to predict future air passenger demand volume. Therefore in this chapter, we present an approach for estimating air passenger origin-destination (OD) flows and total air passenger flows, including transfer passengers, between countries as well as airports. The model is based on the fundamental theory of the gravity law. The results of the model form the basis for the forecasts in Part III of this book. Here, we are looking at air passenger demand from a perspective that does not include airport capacity constraints and their effects on demand volume and structure. Airport capacity and their limits are the topics of Chapter 7, Modelling future airport capacity and capacity utilisation, and Chapter 8, Modelling future airport capacity enlargements and limits, while modelling aircraft size, that is passengers per flight, is presented in Chapter 9, Modelling future development of the average number of passengers per flight. Finally, we integrate all the models presented in Part II, so that we can present an integrated forecast in Part III and discuss the mitigation strategies for limited airport capacity.

6.1

Background

Whilst variables relating to distance, population and gross domestic product (GDP) are rather common in gravity models, we expected further insights by including an airfare variable, which has been computed on the basis of Sabre AirVision Market Intelligence (MI) data. This includes, among other items, passenger flows and airfares between airports by airline. For better discrimination between different types of origin and destination, for example tourist destinations, we have included variables, such as tourism receipts and expenditures and population density. Furthermore, we have employed a Poisson pseudo maximum likelihood (PPML) estimator to produce better and more reliable forecast results, thereby enhancing the out-of-sample results, that is forecast efficacy. We begin this chapter with a brief literature review of gravity models in transport and economics and, before we turn to model parameter estimation and a test application, we present a brief section of Airport Capacity Constraints and Strategies for Mitigation. DOI: https://doi.org/10.1016/B978-0-12-812657-8.00006-3 © 2020 Elsevier Inc. All rights reserved.

133

134

PART | II Models for assessing mitigation strategies

theoretical results that underline the importance of considering a PPML estimator for gravity models. The test application comprises the typical 20 years’ forecasts, such as those of Airbus and Boeing; however, this is not the actual model application of Part III of the book. So far, the gravity model has been employed in a wide range of disciplines in regional science (e.g. Mikkonen and Luoma, 1999), transportation (e.g. Evans, 1976) and trade (e.g. Bergstrand, 1985; Linnemann, 1966; Tinbergen, 1962). One of the first gravity models employed in air transport research was developed by Harvey (1951) to analyse airline traffic patterns in the United States. An overview of selected gravity models in air transport research can be found in Grosche et al. (2007) and Tusi and Fung (2016). In particular, the scope of work of Grosche et al. (2007) has some common ground with this work. They presented two gravity models to estimate the air passenger volume of city pairs without currently existing non-stop air services. Looking at the bigger picture, the gravity model has been applied to many different research objects in air transport research. For example Tusi and Fung (2016) analysed passenger flows at Hong Kong (HKG) airport and focused their research work on a single airport, whereas Matsumoto (2004) and Shen (2004) based their gravity models more upon network analysis. Matsumoto (2004) estimated a gravity model for passenger and cargo flows between a distinct number of large conurbations, such as Tokyo, London, Paris and New York. Shen (2004) estimated a gravity model to analyse intercity airline passenger flows in a 25node US network. Bhadra and Kee (2008) employed a gravity model to analyse demand characteristics, that is fare and income elasticities of the US OD market over time. Endo (2007) developed a gravity model to analyse the impact of the bilateral aviation policy between the United States and Japan on passenger air transport. Hazledine (2009) estimated a gravity model to analyse the border effects in international air travel. The gravity models listed have been estimated by the traditional ordinary least squares (OLS) method in their log-linearised form. Table 6.1 displays

TABLE 6.1 Selected gravity models for origin-destination air passenger demand (Gelhausen et al., 2018). References

Number of variables

R2 (%)

Endo (2007)

4
9

43
69

Harvey (1987)

6

76

Hazledine (2009)

9
11

53
56

Matsumoto (2004)

7
19

39
45

Tusi and Fung (2016)

13

74

Modelling future air passenger demand Chapter | 6

135

various models in terms of their number of parameters and R2 based on the estimation data set (Gelhausen et al., 2018). If the source paper comprises various models, for example in the case of different market segments, ranges of the values are given. The number of variables and R2 vary considerably and are determined by, for example the scope of work, problem structure and data availability. However, there are two caveats: first, R2 is typically computed on the basis of the log-linearised model. Second, R2 is based upon the estimation data set, which does not really assess the out-of-sample forecast efficacy. Therefore we have chosen a different approach, based upon Carson et al. (2011) and Gelhausen et al. (2018), to test the forecast efficacy of the model on 10% of the data sample that has not been used for parameter estimation before.

6.2

Model theory

In a very general form the stochastic gravity equation can be written as (Silva and Tenreyro, 2006) yOD 5 expðα0 Þ L xi;OD βi 1 εOD

ð6:1Þ

yOD 5 expðα0 Þ L xi;OD βi ηOD

ð6:2Þ

εOD expðα0 ÞLi xi;OD βi

ð6:3Þ

i

respectively,

i

with ηOD 5 1 1

The dependent variable yOD is the flow to be modelled between origin ‘O’ and destination ‘D’, expðα0 Þ represents a constant factor and xi;OD are explanatory variables. α0 and β i are the coefficients to be estimated. Here,

xi;OD 5 0 and εOD and η represent error terms with E ε OD
  OD E ηOD
xi;OD 5 1, respectively, which are assumed to be statistically independent of the explanatory variables (Silva and Tenreyro, 2006). Thus, the gravity model is basically a constant elasticity model, such as the Cobb
Douglas production function. Therefore we can create a forecast by multiplying the base year values with the corresponding growth factors. There have been many applications of the gravity model to explain regional or economic phenomena. However, obtaining consistent estimators of the model coefficients of the gravity model has always been an issue (see Goldberger, 1968; Manning and Mullahy, 2001); nevertheless, in many

136

PART | II Models for assessing mitigation strategies

cases, the OLS method has been employed to obtain coefficient estimators of a log-linearised version of (6.2): X     ln yOD 5 α0 1 β i xi;OD 1 ln ηOD

ð6:4Þ

i

However, due to Jensen’s inequality, EðlnðyÞÞ 6¼ lnðEðyÞÞ, the expected value of the logarithm of a variable is different to the logarithm of its expected value (Silva and Tenreyro, 2006). Thus, estimated model coefficients are biased if the gravity model is estimated via OLS in its loglinearised form, unless the error term has certain distributional properties that are rather unlikely to hold. Silva and Tenreyro (2006) showed that εOD 5 expðα0 ÞLi xi;OD βi υOD has to be met to obtain consistent coefficient estimates. Here, υOD is a random variable statistically independent of the explanatory variables. In this case, ηOD equals 1 1υOD  and 
thus is statistically independent of the regressors. As a result, E ln ηOD
xi;OD is a constant with a value of 0. For example, the estimated distance coefficient of a gravity model is typically too low (Siliverstovs and Schumacher, 2009). In connection with air travel, this leads to an overestimation of international long-distance travel. On the other hand, Silva and Tenreyro (2006) proposed a PPML estimator, based upon a Poisson distributed error term, to obtain unbiased coefficient estimates of the gravity model. Here, the model is estimated in its original multiplicative form. The Poisson distribution is characterised by the equality of conditional conditional variance, that is

  mean  and E yOD
xi;OD ~ V yOD
xi;OD . Therefore the conditional mean is a function of a number of explanatory variables: ! X
 E yOD
xi;OD 5 exp α0 1 β i xi;OD ð6:5Þ i

Thus the original multiplicative form of the gravity model (6.1) can be formulated as a Poisson regression model: ! X
  
β i ln xi;OD E yOD xi;OD 5 exp α0 1 ð6:6Þ i

The coefficients are estimated by the method of maximum likelihood. However, the structure of heteroscedasticity, that is the different variability between subsets of the data, implied by the Poisson distribution is rather restrictive, thus we employed the overdispersion tests developed by Cameron and Trivedi (1990) to check for its appropriateness. Here, overdispersion means the presence of greater variability of the data compared to a given statistical model, in this case the Poisson distribution. There are many more possible specifications of heteroscedasticity; however, without further

Modelling future air passenger demand Chapter | 6

137

information on the structure of heteroscedasticity, it seems plausible to give the same weight to all observations, as the PPML estimator does (Silva and Tenreyro, 2006). On the one hand, observations with a larger conditional mean have a larger variance (Silva and Tenreyro, 2006). On the other, larger countries often provide better data quality in the case of trade data and the like (Frankel, 1997; Frankel and Wei, 1993). Nevertheless, for the coefficients to be estimated by PPML as unbiased ones, the data does not have to be Poisson distributed at all (hence the word ‘pseudo’ in PPML). Here, all that is needed is the correct specification of the conditional mean (Gourieroux et al., 1984), although the more the data resembles a Poisson distribution, the more efficient the coefficient estimates are. Regarding the data for model parameter estimation and testing, we have chosen the year 2014 as the reference year. The demand data, that is OD flow per country pair, the number of airports per country and average airfares, have been retrieved from Sabre AirVision Market Intelligence (MI) (2016), whilst socio-economic data, such as GDP, population and tourism expenditures/receipts, have been taken from the World Development Indicators (WDI) of The World Bank (2014). Total migrant stock by origin and destination has been retrieved from the United Nations (UN) (2015b) and is a measure of dependencies between two countries due to international migration. Here, international migrants are defined as people who live in a different country to that where they were born. The full data set comprises 14,642 country pairs with corresponding OD flows, of which 13,178 data sets were used for model estimation and 1464 (10%) were employed for model testing. Most variables (see Table 6.2) are more or less self-explanatory; however, some variables may need further explanation. ‘Distance’ and ‘total airfare’ both represent average values between two countries of flow origin and destination. Raw data, on an airport and airline level, is retrieved from Sabre MI and is weighted by OD demand size of corresponding airport pairs to obtain ‘average’ values between two countries. Whilst the concept of average distances and, especially, average airfares between two countries may seem to be rather inaccurate, the approach works quite well, as we can see in the estimation results section. ‘Country ties’ is a dummy variable that takes a value of 1, if there is a profound relationship between two countries, for example because of former colonial or cultural ties, or being a member of a common political community, such as the Commonwealth or French overseas territories. Data has been retrieved by our own analyses of various sources. The dummy variables ‘domestic’ and ‘continental’ take values of 1 in the case of domestic and continental flights, respectively, and otherwise 0. ‘Passenger-km per rail-km (domestic)’ and ‘population density (domestic)’ take a value of greater than 0 only for domestic passenger
OD flows and 0 otherwise. These variables stand for the ‘need’ for domestic flights and their actual substitution potential by other modes of travel, for example train. The

138

PART | II Models for assessing mitigation strategies

TABLE 6.2 Data employed for model estimation, model testing and example forecast 2035. Variable name

Data source

Data description

Category

Continental

Sabre MI

Value of 1, if it is a flight within one of the seven world regions (Africa, Asia, Europe, Middle East, North America, South America, Southwest Pacific), 0 otherwise

Fixed

Country ties

Own analysis

Value of 1, if there are significant ties between countries, for example the Commonwealth or former colonies

Semifixed

Distance

Sabre MI

Weighted mean flight distance in kilometres between countries. Weighted mean is based on airport pairs with OD demand as weights

Fixed

Domestic

Sabre MI

Value of 1, if it is a domestic flight, 0 otherwise

Fixed

GDP per capita (destination)

WDI/ OECD

GDP per capita in USD of the destination country

Variable

GDP per capita (origin)

WDI/ OECD

GDP per capita in USD of the origin country

Variable

Migration

UN

Total migrant stock by origin and destination

Semifixed

Number of airports (destination)

Sabre MI

Number of airports of the destination country

Semifixed

Number of airports (origin)

Sabre MI

Number of airports of the origin country

Semifixed

Passengers-km per rail-km (domestic)

WDI

Value of million passenger-km per total rail-km for domestic passenger flows, 0 otherwise

Semifixed

Population (destination)

WDI/ UN

Number of people of the destination country

Variable

Population (origin)

WDI/ UN

Number of people of the origin country

Variable

Population density (domestic)

WDI

Value of population density of a country for domestic passenger flows, 0 otherwise

Semifixed (Continued )

Modelling future air passenger demand Chapter | 6

139

TABLE 6.2 (Continued) Variable name

Data source

Data description

Category

Total airfare

Sabre MI

Weighted mean total airfare in USD between countries. Weighted mean is based on average total airfares between airports with OD demand as weights

Variable

Tourism expenditures (origin)

WDI

Tourism expenditures in USD of the origin country

Semifixed

Tourism receipts (destination)

WDI

Tourism receipts in USD of the destination country

Semifixed

MI, Market Intelligence; OD, origin-destination; OECD, Organisation for Economic Co-operation and Development; UN, United Nations; USD, US dollar; WDI, World Development Indicators.

United States, Canada and Australia, in particular, are examples of countries with a high propensity for domestic flights. Furthermore, the input variables have been subdivided into three categories: G

G

G

Fixed: this category comprises input variables, which are fixed over time; for example distance between countries, or whether a flight is a domestic or continental flight. These characteristics do not change in the future. Variable: this category comprises input variables, which are likely to change over time, such as GDP and airfares. These variables are especially important for forecast purposes. Semi-fixed: This category comprises input variables, which can basically change over time, but do so only very slowly or have a high likelihood of no significant change over time, that is the degree of change makes no significant contribution to the forecast. This category is more or less the middle ground between the first two, however, much closer to the first than the second category. Examples include ‘migration’, ‘passenger-km per rail-km’ and ‘country ties’. Nevertheless, these variables can be beneficial for specialised scenario analyses of issues, such as structural changes, like promoting high-speed trains in some countries. Furthermore, fixed and semi-fixed variables serve the purpose of differentiating between ‘types’ of countries and airport pairs, respectively.

For an example forecast of passenger flows for the year 2035, economic data, that means the GDP forecast, was retrieved from the OECD (2017) and the population data forecast from the United Nations (UN) (2015a). The Airbus Global Market Forecast (2016) and the Boeing Current Market

140

PART | II Models for assessing mitigation strategies

Outlook (2016) have been taken as reference forecasts for comparison purposes. Table 6.2 summarises the data that has been used to estimate and test the models and to produce the example forecast for 2035. With the exception of tourism expenditures and receipts, we focused on input variables of the second category. As we will see later in this chapter, the impact of varying tourism expenditures and receipts on the forecast results is rather small, and reliable forecasts for this type of data are hard to obtain. Nevertheless, this kind of data is important for identifying countries where tourism plays a major role, both inbound and outbound. Fig. 6.1 shows the cumulative distribution of global country
country OD passenger flows for the year 2014, based upon our own computations from Sabre AirVision Market Intelligence (MI) (2014). Global OD demand at the country level is highly concentrated; the top ten OD flows, which exclusively consist of domestic travel, account for more than 50% of global OD demand. The largest flow, US domestic, accounts for almost 18% of the global OD demand. As a result, the Gini coefficient also has a high value of 0.9715. For the top ten flows, Fig. 6.1 displays whether they belong to the estimation (E) or test (T) data set. As mentioned, we have taken a 10% sample from the full data set for the purposes of model testing. This comprised every tenth data record of a sample that is ranked by OD passenger volume (ODP). Therefore the largest OD flow in the test data set is Canada domestic, which is much smaller than US domestic.

100% Gini coefficient: 0.9715

Share of global OD passenger volume 2014

90% 80% 70% 60% 50% 40% 30%

Mexico domestic (E) Canada domestic (T) Turkey domestic (E) Australia domestic (E) India domestic (E) Indonesia domestic (E) Brazil domestic (E) Japan domestic (E ) China domestic (E)

20% US domestic (E)

10% 0% 0%

10%

20%

30% 40% 50% 60% 70% Share of OD passenger flows worldwide

80%

90%

100%

FIGURE 6.1 Lorenz curve and Gini coefficient of global country
country OD demand [Sabre AirVision Market Intelligence (MI), 2014].

141

Modelling future air passenger demand Chapter | 6

TABLE 6.3 Key statistics of estimation and test data set [Sabre AirVision Market Intelligence (MI), 2014]. Estimation data

Test data

Ratio

Mean

186,662

79,383

0.43

Standard deviation

5404,174

865,304

0.16

Minimum

0.68

El Salvador
Dem. Rep. Congo

0.92

Afghanistan
Jamaica

1.35

Maximum

461,189,820

US domestic

30,270,504

Canada domestic

0.07

Hence, we have compiled a table of key statistics of the estimation and test data sets, as shown in Table 6.3. Because of the high concentration of global ODP on a rather small number of flows, the mean, standard deviation, minimum and maximum values are much lower in the test data set as compared to the estimation data set. This is almost inevitable and important to know when we later interpret the results of model estimation and testing.

6.3

Model estimation and testing

The coefficient estimates of the gravity equation used here refer to the functional form:     ð6:7Þ E yOD jxi;OD 5 expðα0 Þ L exp αi zi;OD L xj;OD βj L xk;OD β k zk;OD i

j

k

where yOD is the annual passenger flow between origin ‘O’ and destination ‘D’. The first multiplicand is the model constant (variable 1 in Table 6.4), the second refers to the dummy variables (zi;OD ) ‘domestic’, ‘continental’ and ‘country ties’ (variables 2
4), that is dummies which can take only the values 0 and 1. The third relates to continuous variables xj;OD , such as GDP per capita, population and total airfare (variables 5
15). Finally, the fourth multiplicand applies to continuous variables xk;OD . However, they take values greater than 0 only in particular cases and, thus, are combined with a dummy variable zk;OD . This applies to the two variables ‘passenger-km per rail-km (domestic)’ and ‘population density (domestic)’ (variables 16
17). Coefficients to estimate are α0 ; αi ; β j and β k with corresponding indices’ ranges. Coefficients are estimated by log-linearisation of the gravity equation and applying OLS in the OLS approach as well as applying maximum

TABLE 6.4 Estimation results of the ordinary least squares (OLS) and Poisson pseudo maximum likelihood (PPML) approach. OLS

PPML

No.

Variable

Coefficient

Standard error

p-value

Coefficient

Standard error

p-value

1

Constant

214.44461

0.37479

0.00000

28.00390

0.00268

0.00000

2

Domestic

2.48589

0.17201

0.00000

1.02496

0.00193

0.00000

3

Continental

0.00000




0.00000

0.09919

0.00022

0.00000

4

Country ties

3.01515

0.08860

0.00000

2.29920

0.00024

0.00000

5

Distance

20.62874

0.03126

0.00000

20.02307

0.00017

0.00000

6

Total airfare

21.18108

0.03928

0.00000

21.11258

0.00022

0.00000

7

GDP per capita (origin)

0.36591

0.02713

0.00000

0.44865

0.00017

0.00000

8

GDP per capita (destination)

0.44137

0.01792

0.00000

0.22858

0.00011

0.00000

9

Population (origin)

0.32521

0.02239

0.00000

0.35845

0.00015

0.00000

10

Population (destination)

0.52188

0.01327

0.00000

0.26970

0.00010

0.00000

11

Tourism expenditures (origin)

0.34429

0.02254

0.00000

0.08142

0.00013

0.00000

12

Tourism receipts (destination)

0.23082

0.01257

0.00000

0.24743

0.00009

0.00000

13

Number of airports (origin)

0.23094

0.01349

0.00000

0.11970

0.00009

0.00000

14

Number of airports (destination)

0.16442

0.01386

0.00000

0.13854

0.00009

0.00000

15

Migration

0.10384

0.00494

0.00000

0.04065

0.00002

0.00000

16

Passengers-km per rail-km (domestic)

20.54470

0.12864

0.00000

20.81636

0.00066

0.00000

17

Population density (domestic)

0.00000




0.00000

20.05032

0.00046

0.00000

2

R (log-linear model): 77.98%

2

McFadden pseudo-R : 99.24% Overdispersion test g 5 μ(i): 1.581 Overdispersion test g 5 μ(i)2: 0.841

The sum of

(7
10)

1.65437

1.30537

144

PART | II Models for assessing mitigation strategies

likelihood estimation to the original multiplicative gravity equation in the PPML model. Table 6.4 displays the estimated model coefficients for the OLS as well as the PPML approach. All coefficient estimates included are highly significant (p-value ,0.001) and of the expected sign. The ‘continental’ and ‘population density (domestic)’ variables of the OLS model are not significant at the 10% significance level and of the wrong sign, respectively. However, while all coefficient estimates included in the final model setup are highly significant, their standard errors are much lower in the PPML model, by a factor of between 90 (‘domestic’) and almost 370 (‘country ties’), the average factor being around 175. The R2 of the OLS approach is 77.98%, based upon the log-linearised version of the gravity model as typically reported in the literature. Compared to the models of Table 6.1, this is a rather satisfactory value of the R2. McFadden’s pseudo-R2 of the PPML approach is 99.24%. However, McFadden’s pseudo-R2 is not the same as R2 and is defined as (Hensher et al., 2005) pseudo-R2 5 1 2

LL ðestimated modelÞ LL ðbase modelÞ

ð6:8Þ

LL is the value of the log-likelihood function. The reason for using two different statistics for evaluating model fit is that one model is estimated by using OLS, while the other model is estimated using a maximum likelihood method, PPML. However, Domencich and McFadden (1975) have established an approximate relationship between McFadden’s pseudo-R2 and R2 of OLS: 30% pseudo-R2 equals around 65% R2, 40% pseudo-R2 is around 78% R2, 50% pseudo-R2 is about 90% R2 and 60% pseudo-R2 corresponds to more than 95% R2. To test for overdispersion, we employed the regression-based tests developed by Cameron and Trivedi (1990). Here, the null hypothesis   ð6:9Þ H 0 :Var yi 5 μi ðmean-variance equalityÞ is tested against the alternative hypothesis     ð6:10Þ H 1 :Var yi 5 μi 1 αg μi ðoverdispersionÞ   We have specified g μi to be equal to μi and μ2i , to test for different forms of overdispersion. The test statistics report values of 1.581 and 0.841, respectively, thus failing to reject the null hypothesis of no overdispersion. The critical value from the chi-squared table for one degree of freedom is 3.84 at the 5% significance level (Econometric Software, 2007). This is in line with the empirical simulations of King (1988) that the coefficient estimates of the Poisson regression are consistent and somewhat efficient, especially in large data samples. Furthermore, Table 6.4 reports the sum of coefficients of GDP per capita and population (both for origin and

*1012

Modelling future air passenger demand Chapter | 6

145

3,000,000 2,597,193 2,500,000

2,000,000

1,500,000

1,000,000

500,000 238,165 0

6848 In-sample residual sum of squares OLS

145

Out-of-sample residual sum of squares PPML

FIGURE 6.2 Residual sum of squares of the OLS and PPML approaches. OLS, ordinary least squares; PPML, Poisson pseudo maximum likelihood.

destination), which serves as an indicator of how GDP-responsive the model is. Here the OLS model is slightly more GDP-responsive than the PPML approach (1.65 vs 1.31). Before discussing the estimated elasticities of both models, we take a closer look at forecast efficacy, evaluated by means of the test data set. Fig. 6.2 displays the in-sample and out-of-sample residual sum of squares (RSS) of the OLS and PPML approaches. The values are very high, due to the large sample sizes, (especially the estimation data set) and the fact that both approaches are evaluated by means of the multiplicative form of the gravity model. For comparison, in its log-linearised form, the OLS model produces, for the RSS, a value of 32,221.22 for the estimation data set. In contrast, in both cases, the estimation and test data sets, the PPML model produces a significantly lower RSS and, thus, performs much better insample and out-of-sample than the OLS model. Figs 6.3 and 6.4 show the in-sample and out-of-sample mean absolute forecast error and the standard deviation of the absolute forecast error for both approaches; the interpretation of the results is basically the same as in the case of Fig. 6.2. The PPML model produces much better results than the OLS approach. Finally, Fig. 6.5 displays the corresponding ratios compiled from Figs 6.2 to 6.4. The PPML model performs better than the OLS approach in any case, especially out-of-sample. Whilst the RSS of the PPML model is lower by a factor of about eleven (1/0.09) in-sample, it increases to a value of 50 (1/ 0.02) out-of-sample. The mean absolute forecast error of the PPML model is

146 *103

PART | II Models for assessing mitigation strategies 300

285

250

200

150

144

138

100

50

39

0

In-sample mean absolute forecast error OLS

Out-of-sample mean absolute forecast error PPML

*103

FIGURE 6.3 Mean absolute forecast error. OLS, ordinary least squares; PPML, Poisson pseudo maximum likelihood.

16,000 14,000

14,036

12,000 10,000 8000 6000 4000 2000 0

4249 2158 313 In-sample standard deviation of absolute Out-of-sample standard deviation of absolute forecast error forecast error OLS PPML

FIGURE 6.4 Standard deviation of the forecast error. OLS, ordinary least squares; PPML, Poisson pseudo maximum likelihood.

lower by a factor of two (1/0.48) in-sample and almost four (1/0.27) out-ofsample. Finally, the standard deviation of the absolute forecast error of the PPML model is more than three times (1/0.30) lower in-sample and about seven times (1/0.14) lower out-of-sample. As a result, we can conclude that the PPML approach has a much better out-of-sample forecast efficacy.

Modelling future air passenger demand Chapter | 6

147

0.6

0.5

0.48

0.4 0.30

0.3

0.27

0.2 0.14

0.1

0.09 0.02

0.0

In-sample residual sum of squares (PPML/OLS)

In-sample Out-of-sample Out-of-sample In-sample mean Out-of-sample standard standard residual sum of absolute forecast mean absolute deviation of deviation of squares error (PPML/OLS) forecast error (PPML/OLS) absolute forecast absolute forecast (PPML/OLS) error (PPML/OLS) error (PPML/OLS)

FIGURE 6.5 Ratio of forecast efficacy evaluation statistics (PPML/OLS). OLS, ordinary least squares; PPML, Poisson pseudo maximum likelihood.

However, compared to various other models (Table 6.1), even the OLS model performs rather well in terms of R2, but, because of log-linearisation, R2 of gravity models should be taken with a pinch of salt. The out-of-sample forecast efficacy of the OLS approach is typically lower than the in-sample statistics of the log-linearised version might suggest. Fig. 6.6 visualises the model elasticities of the OLS and PPML approaches. Distance, in particular, plays a bigger role in the OLS approach than in the PPML model. On the other hand, ‘domestic’ factors are more important in the PPML approach. As a result, the OLS model is much more distance-responsive. Tourism expenditures, and the number of airports and migration are more important in the OLS model than in the PPML approach. Tourism receipts are more or less equally important in both models. However, in the example forecast presented in this chapter we will focus on GDP per capita, population and airfare. The remaining variables tend to be more stable over time and mainly serve to discriminate between different types of origins and destinations, for example leisure versus business destinations and large and thinly populated countries with a less developed railway network versus small and densely populated countries with a highly developed railway network. Fig. 6.7 compares the values of the constants as well as domestic, continental and country ties dummy variables of the OLS and PPML approaches. The absolute values of these variables (with the exception of the continental variable, which is not significant in the OLS model) are significantly greater

148

PART | II Models for assessing mitigation strategies

0.8 0.6 0.4 0.2 0.0 –0.2 –0.4 –0.6 –0.8 –1.0 –1.2 –1.4 OLS

PPML

FIGURE 6.6 Comparison of the estimated model elasticities of the OLS and PPML approaches. OLS, ordinary least squares; PPML, Poisson pseudo maximum likelihood. 4 2 0

Constant

Domestic

Continental

Country ties

–2 –4 –6 –8 –10 –12 –14 –16 OLS

PPML

FIGURE 6.7 Comparison of the estimated constants, domestic, continental and country ties dummy variables of the OLS and PPML approaches. OLS, ordinary least squares; PPML, Poisson pseudo maximum likelihood.

in the OLS model when compared to the PPML approach. We interpret these findings as follows: the OLS model penalises the increasing distance much more than the PPML approach. This is partly offset by the large absolute value of the constant and a high value of the ‘domestic’ and ‘country ties’

Modelling future air passenger demand Chapter | 6

149

dummy variables to account for domestic traffic and special relationships between countries, especially in the United States and China. An advantage of the gravity model is the constant elasticity property, which means that growth factors can be employed to produce a forecast. The base year value multiplied by the growth factor equals the forecast value. For example, a GDP growth rate of 2% per year (CAGR) over a 20-year time horizon and a GDP elasticity of 1.5 equals a growth factor of 1.0320 5 1.81, thus, the forecast value is 1.81 times the base value. The advantage of applying growth factors is that potential erroneous estimates of base year flows have no effect on forecast values. Next we need to compute total passenger volume (TP) growth rates, including transfer passengers, from ODP growth rates. Fig. 6.8 illustrates OD and total passenger volume growth rates for the period 2003
16. According to Sabre MI, ODP increased between 2003 and 2016 on average by 6.1% per year (2011
16: 9.6%) and TP, including transfer passengers, grew on average by 5.4% per year (2011
16: 9.1%). However, whilst being more volatile in recent years, the difference between these two growth rates is more or less stable over time, as Fig. 6.8 shows. Furthermore, as discussed in Chapter 4, Constrained and under-utilised airports, of Part I of this book, the concentration of flights on airports worldwide has more or less stagnated over the period 2000
16. Thus the long-term difference in average annual growth rates of ODP and TP has been about 0.5%
0.7% points on a global scale.

Annual growth rate of OD and total passenger volume

25%

20%

15%

10%

5%

0%

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

–5%

–10%

Annual growth rate of OD passenger volume (in %), AGROD Annual growth rate total passenger volume (in %), AGRTP AGRTP - AGROD

FIGURE 6.8 Comparison of annual growth rates of global OD passenger and total passenger volume (including transfer passengers) for the period 2003
16 [Sabre AirVision Market Intelligence (MI), 2016].

150

PART | II Models for assessing mitigation strategies

Modelling airport choice based upon individual passenger decisions (e.g. Gelhausen, 2011; Gelhausen et al., 2008 for German case studies) is a rather complex and data-intensive endeavour, especially if access mode, flight route, airline or destination choice are also included. Therefore the majority of models and case studies are limited to a small number of airports of a particular region, for example the San Francisco Bay area, because of good data availability (see, e.g. Harvey, 1987; Pels et al., 2001, 2003; Basar and Bhat, 2004; Hess and Polak, 2005, 2006). Expanding this approach to a global perspective with more than 4000 airports seems to be a more or less insurmountable task on grounds of data availability alone. Therefore we have chosen a more parsimonious and less sophisticated model for transforming growth rates of ODP into growth rates of TP. Based upon Sabre MI data of almost 4000 airports worldwide in 2015, we linked TP to ODP and ODP share (ODPS) in a nonlinear way. Therefore the equation, which we call a variable gravity model because of the variable exponent of ODP, can be written as TP 5 1:01 ODP1:0420:13ODPS

ð6:11Þ

where ‘TP’ is the total passenger volume, ‘ODP’ is the OD passenger volume and ‘ODPS’ is OD passenger volume share. The ‘ODP’ is defined as the number of passengers flying from airport A to B on non-stop and stopover routes, whereby airport A is the one of the trip origin and airport B is the one of the trip destination. The ‘TP’ is, in contrast, the number of passengers on board all direct flights between these two airports. The ‘ODPS’ is the ratio of ODP to TP between these two airports. The ODP elasticity of TP is defined as @TP ODP 3 5 1:04 2 0:13 ODPS @ODP TP

ð6:12Þ

As a result, the rather simple functional relationship is not a drawback, as the estimation results in Table 6.5 underline. The model has been estimated by PPML. The ODP elasticity is not constant but depends on the ODPS, hence, we have a gravity model with a variable exponent. The higher the ODPS, the lower the TP growth rate. The coefficients are all highly significant (p-value ,0.001) and model fit is very good (McFadden pseudoR2 5 99.96%). However, the values of the overdispersion tests are not as good as in the case of the more complex OD demand model, but still satisfactory. Fig. 6.9 illustrates the model results graphically and shows how the ODP and TP are interrelated. If the share of ODP is below 32.1%, then TP will grow faster than ODP and vice versa. The ODP and TP models can be employed on different spatial levels. These include the following: G G G G

Global Regional Country Airport

Modelling future air passenger demand Chapter | 6

151

TABLE 6.5 Estimation results of the total passenger forecast model. No.

Variable

Coefficient

Standard error

p-value

1

Constant

1.01238

0.00116

0.00000

2

OD passenger volume

1.04142

0.00013

0.00000

3

OD passenger volume share

20.13310

0.00006

0.00000

McFadden pseudo-R2

99.96%

Overdispersion test g 5 μ(i)

5.065

Overdispersion test g 5 μ(i)

8.571

2

1.05

1.00

0.95

0.90

96%

91%

86%

81%

76%

71%

66%

61%

56%

51%

46%

41%

36%

31%

26%

21%

16%

11%

0.80

6%

0.85

1%

Total passenger volume growth/OD passenger volume growth

OD, origin-destination.

OD passenger volume/Total passenger volume

FIGURE 6.9 OD passenger volume elasticity of total passenger volume as a function of OD passenger volume share. OD, origin-destination.

At the regional and airport level the main limiting factor for applying the ODP flow model is the availability of forecast data of the input factors, such as GDP and population. Therefore typically, this model can be applied on a country level (or on a sub-country level for large countries, such as the United States or Russia), and the TP model on airport level, with the ODPS being the differentiating factor between individual airports. Thus our basic setup for Part III of this book is, first, to run the ODP model on the country level and then the TP model on the airport level, which we have found to work very well. Forecast ODP growth rates for each country pair are the

152

PART | II Models for assessing mitigation strategies Population (destination) 0.27% GDP per capita (origin) 0.45%

GDP per capita (destination) 0.23%

Total airfare –1.11%

Population (origin) 0.36%

OD passenger demand

Tourism expenditures (origin) 0.08%

Population density (domestic) –0.05%

Tourism receipts (destination) 0.25%

Total passenger demand FIGURE 6.10 Variable input factor elasticities of the total passenger demand model.

basis for the forecast TP growth rates for each airport pair, which are applied to the corresponding passenger volumes of the year 2016 to obtain forecast TP for each airport pair for the years 2030 and 2040. However, in the next section, both models are run on the global level, as we are primarily only interested in global forecast values. Here, we found no significant difference between running the models on a lower level and aggregating the results up to the global level and directly running the models on the global level. To prepare for the next section, Fig. 6.10 illustrates the impact of the input factors on ODP, which underlie the forecasts. For example, if airfares rise by 1%, ODP declines by 1.11%. On the other hand, if GDP per capita increases in the origin country by 1%, then ODP demand rises by 0.45%. Total passenger demand growth depends on the ODPS. If ODPS is less than 32.1%, then TP growth is marginally higher than OD demand growth; in the case that ODPS is higher than 32.1%, TP growth is marginally lower than OD demand growth. After modelling OD demand volume and TP demand volume (on already existing flight connections), we turn to the topic of new direct flight services between airport pairs where currently no non-stop flight exists. Here, we looked at non-stop connections that are viable in the long run given a particular flight frequency between two airports. Of course, many direct flight services open and close each year, but here we have taken a long-term view. Furthermore, we have already developed a similar model to assess the

Modelling future air passenger demand Chapter | 6

153

potential for intercontinental low-cost flights (Wilken et al., 2016) and the model presented in this book is based upon this approach. However, we have improved the model in several respects. First, we have enlarged the model’s scope. Therefore it comprises flights of all distances worldwide and, thus, is not only limited to long-haul travel from Europe. Second, we have employed a PPML estimator for model estimation. The purpose of this model is to forecast the number of air passengers flying directly on a link between two airports, assuming the existence of direct services in future. The forecast is based upon the following explanatory variables of the model: G G

G

Annual flight volume between two airports to model flight frequency. OD passenger volume between two airports, which includes both passengers flying non-stop and those passengers taking a stopover flight. Flight distance in kilometres between two airports.

Certainly, for our purposes, the OD passenger volume variable is the main driver of new direct connections between airports in the future. However, service frequency is also important to make a non-stop service more attractive from the passengers’ perspective as compared to a stopover flight. For example, passengers might be more likely to take a stopover flight, which is served twice daily compared to a non-stop flight twice a week. In turn, given a particular OD demand, flight frequency depends on aircraft size, that is passengers per aircraft, if load factors are economically viable. Therefore based on our forecast of OD passenger volume, we want to find out which stopover connections might be viable for non-stop flights in the future and what share of OD passenger volume choose the direct service. Therefore we have identified different types of routes between airports: G

G

G

G

‘Hub routes’ (Hub), defined as having a share of less than 50% OD demand between two airports compared to all passengers. In this case the number of transfer passengers at that airport pair exceeds the number of OD travellers taking a non-stop flight between those two airports. ‘Point-to-point (P2P) routes’, defined as having a share of between 95% and 105% OD demand between two airports compared to all passengers. These are routes where the majority of OD passengers are typically travelling non-stop. ‘Low-frequency (LF) routes’, defined as having a share of more than 150% OD demand between two airports compared to all passengers. These are routes where the majority of OD passengers takes a stopover flight. Some routes do not belong to any of the aforementioned categories, as they have an OD passenger volume share of 50% to less than 95% and more than 105% up to 150%, respectively.

154

PART | II Models for assessing mitigation strategies

We have tested various route definitions empirically and have found that these categories produce good results in terms of model fit and parameter significance. However, ultimately, the definitions remain more or less arbitrary. The model has been estimated by PPML. The dependent variable is the ‘TP’ on direct flights between two airports. Table 6.6 presents the estimation results. ‘Hub’, ‘P2P’ and ‘LF’ are dummy variables that take a value of 1, if the route belongs to the category, and 0 otherwise. All coefficients show the expected sign and are highly significant (p-value ,0.01). McFadden’s pseudo-R2 of the model is 98.3%. The values of the overdispersion tests are very good; hence, the PPML estimator is fine for model estimation. All model elasticities are positive, although less than 1. Passenger volume flying non-stop is rather inelastic with regard to supply and demand characteristics, such as an increase of annual flights, OD demand and flight distance. Increasing OD demand between two airports by 1% increases the number of non-stop air passengers of that OD demand between those two airports by only 0.628%. This is because stopover flights on alternative routes are typically cheaper than non-stop flights. Increasing annual flight volume by 1% only raises the number of non-stop travellers on a route by 0.385%. Flight distance has a rather small effect on passenger volume on a route, thus, the elasticity is 0.132%. However, there are large differences in passengers on a route if we look at route categories. Ceteris paribus, a hub route has almost twice as many passengers as a P2P route (exp(0.6442)/exp (20.0562) 5 2.01). An LF route has only about 56% compared to a P2P

TABLE 6.6 Estimation results of the new direct connections model. No.

Variable

Coefficient

Standard error

p-value

1

Constant

0.86632

0.00029

0.00000

2

Hub

0.64420

0.00007

0.00000

3

P2P

20.05615

0.00005

0.00000

4

LF

20.64248

0.00020

0.00000

5

OD passenger volume

0.62771

0.00004

0.00000

6

Annual flight volume

0.38505

0.00005

0.00000

7

Flight distance

0.13150

0.00003

0.00000

2

98.31%

McFadden pseudo-R

Overdispersion test g 5 μ(i)

1.430

Overdispersion test g 5 μ(i)

0.000

2

LF, low frequency; OD, origin-destination; P2P, point-to-point.

Modelling future air passenger demand Chapter | 6

155

OD passenger demand volume forecast

Existing routes

Non-stop flights model

New routes

Total passenger demand volume and flight network structure forecast

FIGURE 6.11 The air passenger demand model.

route. Altogether, these results show the importance of airline network strategies and, here especially, of consolidating traffic at hubs in long-haul air travel (Alderighi et al., 2005; Hooper, 2005). Therefore the route categories and dummy variables are important to account for airline network strategy effects in estimating the number of passengers on a route between two airports. Finally, Fig. 6.11 summarises the air passenger demand model graphically. First, for each airport pair, OD and TP are forecast. Then, each airport pair, including existing as well as potentially new direct services, is checked in view of viable non-stop connections.

6.4

Model application: comparing different forecasts

For this brief case study, we have chosen GDP and population forecasts, which are available to the public. However, for the more elaborated forecasts of Part III of this book, we needed more detailed data. Therefore in this chapter, forecast data has been retrieved from OECD (2017) for global real GDP growth from 2016 until 2035. The medium variant of the UN World Population Prospects [United Nations (UN), 2015a] has been chosen as a forecast of global population growth for the period from 2016 to 2035. Thus, we have taken a value of 3.2% per annum (p.a.) for global real GDP growth and 0.092% p.a. for global population growth. In the matter of real airfares, we distinguish between three scenarios: G G G

Constant real airfares 1% decline p.a. of real airfares 2% decline p.a. of real airfares

A long-term decline of real airfares seems plausible, due to the growth of the low-cost carrier segment and legacy carriers pushing their own low-cost subsidiaries. Furthermore, there is market potential in the long-haul segment

156

PART | II Models for assessing mitigation strategies

350

14

300

12

250

10

200

8

150

6

100

4

50

2

0

2002

2004

2006

2008

OD average base fare (USD)

2010

2012

2014

2016

Average yield (US cent/km)

OD average base fare (USD)

(e.g. Wilken et al., 2016). Even nominally stagnant airfares mean declining real airfares, due to inflation. Fig. 6.12 shows the development of the OD average base fare, that is the average fare excluding taxes and other charges, in USD and the average yield in US cents per km for the period 2002
16, as total airfares, including taxes and other charges, are only available from 2014 to 2016 [Sabre AirVision Market Intelligence (MI), 2016]. Between 2002 and 2016 the nominal OD average base fare and average yield declined by 0.7% per year [compound annual growth rate (CAGR)]. The highest nominal values for the OD average base fare and average yield were reached in 2012. For this period, CAGR was 3.6% and 3.7%, respectively. Thereafter, from 2012 to 2016, as well as the years 2009 and 2010, OD average base fare and average yield declined rapidly by 10.6% and 10.8% per year on average, respectively, accompanied by a large decline in jet fuel prices and increasing lowcost competition. The period with the largest CAGR is 2002
08. Here, we find values of 4.7% and 3.5% for the OD average base fare and the average yield, respectively. Whilst airfares are volatile over time, the long-term assumption of stagnating or slightly declining nominal airfares seems to be realistic.

0

Average yield (US cent/km)

FIGURE 6.12 Development of OD average base fare in USD and yield in US cent worldwide for the period 2002
16 [Sabre AirVision Market Intelligence (MI), 2016]. OD, origindestination.

Modelling future air passenger demand Chapter | 6

50% 40%

3.0

30% 2.5

20% 10%

2.0

0% 1.5

–10% –20%

1.0

Year-to-year price change

US kerosene-type jet fuel retail sales by refiners in USD/gallon

3.5

157

–30% 0.5

0.0

–40%

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 US kerosene-type jet fuel retail sales by refiners in USD/gallon

–50%

Year-to-year price change

FIGURE 6.13 Development of US kerosene-type jet fuel retail sales by refiners USD/gallon (US Energy Information Administration, 2017).

Fig. 6.13 displays the development of US kerosene-type jet fuel retail sales by refiners in USD/gallon and the year-to-year price changes in percentages. There was a large decline in 2009 and, since 2012, jet fuel prices have continued to decline. Furthermore, jet fuel prices and airfares are highly correlated. The value of the correlation coefficient for the data sample is 0.9; hence, jet fuel prices and airfares move more or less in the same direction. However, whilst the OD average base fare and average yield declined by 0.7% per year during the period 2002
16, jet fuel prices increased by 4.4% per year on average, despite large fluctuations. On the other hand, during the 2012
16 period, jet fuel prices fell on average by 19% per year, whilst the OD average base fare declined by 11% per year. Figs 6.12 and 6.13 illustrate that jet fuel prices are much more volatile than airfares but generally tend to move in the same direction. Against this background the scenario assumptions seem realistic. The average inflation rate of consumer prices in the Euro area was 2.2% between 1995 and 2015 and 4.4% worldwide (The World Bank, 2015). However, airfares could also nominally increase, but less than the inflation rate, which also leads to a depreciation of real airfares. Still, these assumptions remain more or less arbitrary; however, from our point of view, a further decline of real airfares is rather likely.

158

PART | II Models for assessing mitigation strategies

7% 5.9% 5.5%

Annual growth rate (CAGR)

6% 5%

4.0%

4%

5.1% 4.8%

4.6% 4.3%

4.5%

3.9% 3.7%

3.4% 3.2%

3%

2.7% 2.6%

2% 1% 0%

Airbus

Boeing

OLS (constant real air fares)

PPML (constant real air fares)

CAGR total passenger volume

OLS (1% PPML (1% OLS (2% PPML (2% real air fares real air fares real air fares real air fares decline p.a.) decline p.a.) decline p.a.) decline p.a.) CAGR OD passenger volume

FIGURE 6.14 Comparison of different forecast scenarios in terms of airfare development with Airbus Global Market Forecast and Boeing Current Market Outlook (CAGR, 2016
35). CAGR, compound annual growth rate.

Fig. 6.14 shows the three different scenarios of the annual global total passenger demand volume growth rate forecast for the period from 2016 to 2035 and the corresponding forecasts of Airbus (2016) and Boeing (2016). Forecast values for Airbus are 4.5% p.a. and 4.0% p.a. for Boeing. Depending on the scenario chosen, the OLS model produces forecast values between 3.2% p.a. for constant real airfares and 5.5% p.a. for a 2% decline of real airfares. The PPML approach produces somewhat lower forecast values, which are in a range of 2.6% p.a. and 4.8% p.a. Overall, a decline of real airfares seems to be more likely than constant real airfares, especially compared to the Airbus and Boeing forecasts. Furthermore, Fig. 6.14 shows the corresponding ODP growth rate forecasts (excluding Airbus and Boeing, since they are not available to the public). As already explained earlier in this chapter, they are slightly higher and are in a range of 2.7% and 5.1% for the PPML model, depending on the assumptions about the future development of airfares. Correspondingly, the OLS model produces values that are about 0.7%
0.8% points higher. However, there is a caveat: as much as forecast results depend on forecast methodology, input data and (model) assumptions play a very important role as well. Therefore we have aimed for as much transparency as possible.

Modelling future air passenger demand Chapter | 6

159

References Alderighi, M., Cento, A., Nijkamp, P., Rietveld, P., 2005. Network competition
the coexistence of hub-and-spoke and point-to-point systems. J. Air Transp. Manage. 11, 328
334. Airbus, 2016. Global Market Forecast 2016
2035. Airbus, Blagnac. Basar, G., Bhat, C., 2004. A parameterized consideration set model for airport choice: an application to the San Francisco Bay area. Transp. Res., B 38, 889
904. Bergstrand, J.H., 1985. The gravity equitation in international trade: some microeconomic foundations and empirical evidence. Rev. Econ. Stat. 67, 474
481. Bhadra, D., Kee, J., 2008. Structure and dynamics of the core US air travel markets: a basic empirical analysis of domestic passenger demand. J. Air Transp. Manage. 14, 27
39. Boeing, 2016. Current Market Outlook 2016
2035. Boeing, Seattle, WA. Cameron, A.C., Trivedi, P.K., 1990. Regression-based tests for overdispersion in the Poisson model. J. Econometrics 46, 347
364. Carson, R., Cenesizoglu, T., Parker, R., 2011. Forecasting (aggregate) demand for US commercial air travel. Int. J. Forecast. 27 (3), 923
941. Domencich, T.A., McFadden, D., 1975. Urban Travel Demand - A Behavioral Analysis. Elsevier, New York. Econometric Software, 2007. LIMDEP Version 9.0 and NLOGIT Version 4.0 Reference Guides. Plainview. Endo, N., 2007. International trade in air transport services: Penetration of foreign airlines into Japan under the bilateral aviation policies of the US and Japan. J. Air Transp. Manage. 13, 285
292. Evans, S.P., 1976. Derivation and analysis of some models for combining trip distribution and assignment. Transp. Res. 10, 37
57. Frankel, J., 1997. Regional Trading Blocs in the World Economic System (Institute for International Economics). Peterson Institute for International Economics, Washington, DC. Frankel, J., Wei, S., 1993. Trade blocs and currency blocs. In: NBER Working Paper No. 4335. Gelhausen, M.C., 2011. Modelling the effects of capacity constraints on air travellers’ airport choice. J. Air Transp. Manage. 17 (2), 116
119. Gelhausen, M.C., Berster, P., Wilken, D., 2008. Airport choice in Germany and the impact of high-speed intercity train access: the case of the Cologne region. J. Airport Manage. 2, 355
370. Gelhausen, M.C., Berster, P., Wilken, D., 2018. A new direct demand model of long-term forecasting air passengers and air transport movements at German airports. J. Air Transp. Manage. 71, 140
152. Goldberger, A., 1968. The interpretation and estimation of Cobb-Douglas functions. Econometrica 36 (3/4), 464
472. Gourieroux, C., Monfort, A., Trognon, A., 1984. Pseudo maximum likelihood methods: Applications to Poisson models. Econometrica 52, 701
720. Grosche, T., Rothlauf, F., Heinzl, A., 2007. Gravity models for airline passenger volume estimation. J. Air Transp. Manage. 13, 175
183. Harvey, D., 1951. Airline passenger traffic patterns within the United States. J. Air Law Commerce 18, 157
165. Harvey, G., 1987. Airport choice in a multiple airport region. Transp. Res., A 21 (6), 439
449. Hazledine, T., 2009. Border effects for domestic and international Canadian passenger air travel. J. Air Transp. Manage. 15, 7
13.

160

PART | II Models for assessing mitigation strategies

Hensher, D.A., Rose, J.M., Greene, W.H., 2005. Applied Choice Analysis - A Primer. Cambridge. Hess, S., Polak, J.W., 2005. Mixed Logit modelling of airport choice in multi-airport regions. J. Air Transp. Manage. 11 (2), 59
68. Hess, S., Polak, J.W., 2006. Airport, airline and access mode choice in the San Francisco Bay area. Pap. Reg. Sci. 85 (4), 543
567. Hooper, P., 2005. The environment for Southeast Asia’s new and evolving airlines. J. Air Transp. Manage. 11, 335
347. King, G., 1988. Statistical models for political science event counts: Bias in conventional procedures and evidence for the exponential Poisson Regression model. Am. J. Pol. Sci. 32 (3), 838
863. Linnemann, H., 1966. An Econometric Study of International Trade Flows. North-Holland Publishing Company, Amsterdam. Manning, W.G., Mullahy, J., 2001. Estimating log models: to transform or not to transform? J. Health Economics 20, 461
494. Matsumoto, H., 2004. International urban systems and air passenger and cargo flows: Some calculations. J. Air Transp. Manage. 10, 241
249. Mikkonen, K., Luoma, M., 1999. The parameters of the gravity model are changing
how and why? J. Transp. Geogr. 7, 277
283. Organisation for Economic Co-operation and Development (OECD), 2017. GDP Long-Term Forecast (Indicator). https://doi.org/10.1787/d927bc18-en. Retrieved May 2, 2017. Pels, E., Nijkamp, P., Rietveld, P., 2001. Airport and airline choice in a multi-airport region: an empirical analysis for the San Francisco bay area. Reg. Stud. 35 (1), 1
9. Pels, E., Nijkamp, P., Rietveld, P., 2003. Access to and competition between airports: a case study for the San Francisco Bay area. Transp. Res., A 37, 71
83. Sabre AirVision Market Intelligence (MI), 2016. Data Based on Market Information Data Tapes (MIDT). Sabre, Southlake. Shen, G., 2004. Reverse-fitting the gravity model to inter-city airline passenger flows by an algebraic simplification. J. Transp. Geogr. 12, 219
234. Siliverstovs, B., Schuhmacher, D., 2009. Estimating gravity equations: to log or not to log? Empirical Econ. 36 (3), 645
699. Silva, J.M.C.S., Tenreyro, S., 2006. The log of gravity. Rev. Econ. Stat. 88 (4), 641
658. The World Bank, 2014. World Development Indicators. Washington, DC. The World Bank, 2015. World Development Indicators. Washington, DC. Tinbergen, J., 1962. Shaping the World Economy. Twentieth Century Fund, New York. Tusi, W.H.K., Fung, M.K.Y., 2016. Analysing passenger network changes: the case of Hong Kong. J. Air Transp. Manage. 50, 1
11. United Nations (UN), 2015a. World Population Prospects: The 2015 Revision, DVD ed. United Nations (UN), Department of Economic and Social Affairs, Population Division. United Nations (UN), 2015b. Trends in International Migrant Stock: Migrants by Destination and Origin (United Nations Database, POP/DB/MIG/Stock/Rev.2015), Department of Economic and Social Affairs. US Energy Information Administration, 2017. US Kerosene-Type Jet Fuel Retail Sales by Refiners. Retrieved May 8, 2017. Wilken, D., Berster, P., Gelhausen, M.C., 2016. Analysis of demand structures on intercontinental routes to and from Europe with a view to identifying potential for new low-cost services. J. Air Transp. Manage. 56B, 79
90.

Chapter 7

Modelling future airport capacity and capacity utilisation For assessing the effects of limited airport capacity on global air traffic, we ideally need to assign capacity values to each airport, at least to those airports which may reach their capacity limit before the forecast horizon. This is an enormous task and a major difference to pure demand forecasts, which often neglect potentially negative effects of limited airport capacity. Naturally, a global capacity assessment cannot be conducted with the same precision as a detailed capacity study for a particular airport. Here, we have detailed information about many influencing factors, such as the runway and terminal layout, fleet mix, weather and terrain and air traffic management. It would be an insurmountable task to gather and process such an amount of information for all airports worldwide, and to make projections about their development for the next 20 30 years. Even limiting this approach to only such airports that may reach their capacity limit before the forecast horizon, and collecting capacity relevant information for each airport, generates a sample that is most likely too large for a detailed airport capacity assessment within a reasonable time frame. Furthermore, while a number of airports will reach their capacity limit before the forecast horizon, some airports may be successfully enlarged, typically by one or more new runways. Therefore we not only need a method that is able to assess current airport capacity on a global level, but also one that is able to assess future airport capacity. Whether adding a new runway is a realistic option for enlarging airport capacity will be the topic of Chapter 8, Modelling future airport capacity enlargements and limits. As a result, we need a method that produces robust results with input information which is generally available. We focused therefore on the runway system, as this typically limits overall airport capacity in the long term. As we will see later in more detail, it is not necessary (and hardly possible on a global level) for the model to produce current and future capacity values that are as ‘exact’ as possible. Future development of global air traffic depends among other things on the availability of future airport capacity, as we allow for capacity enlargements in our model. Therefore if the airport capacity model under- or overestimates current airport capacities, future Airport Capacity Constraints and Strategies for Mitigation. DOI: https://doi.org/10.1016/B978-0-12-812657-8.00007-5 © 2020 Elsevier Inc. All rights reserved.

161

162

PART | II Models for assessing mitigation strategies

airport capacity enlargements will be easier or harder to realise, respectively, until the forecast horizon. One major factor which determines the probability of a successful airport capacity enlargement is the number of aircraft movements that the airport has to handle. The more aircraft movements, the harder airport capacity enlargements become. Thus under- or overestimating airport capacity means more or fewer steps of increasing capacity, respectively. Any inaccuracies in assessing airport capacity that still remain typically play only a minor role in the strategic planning process. Nevertheless, for our generic airport capacity model we aimed for an as-high-as-possible degree of accuracy, but we do not pretend it is as accurate as detailed airport capacity analyses for specific airports.

7.1

Background

In Chapter 3, Capacity utilisation at airports worldwide, we introduced the analytical tool of air traffic ranking curves and the capacity utilisation index (CUI) concept. The CUI is an indicator to describe capacity utilisation of airports by relating the average hourly air traffic volume to the 5% peak hour volume. However, this relationship only works for airports in a meaningful way if an airport is operating close to its capacity during the 5% peak hour. As we have seen, this is not the case for many smaller airports which do not reach traffic volumes near capacity during peak hours. For such airports a value of practical capacity has to be found in another way for computing the capacity utilisation of today and for the future. Another main benefit of the CUI concept is the deduction of a so-called annual airport capacity or annual service volume. These two terms are used interchangeably in this book. This is in contrast to a naı¨ve approach, whereby the annual capacity is calculated as maximum hourly capacity multiplied by the number of operating hours of the airport per day multiplied by the number of days per year, whereby the CUI takes the demand side into account. For example, there is less demand during night hours or weekends; thus the technical capacity will not be fully used because of a temporary lack of demand. This is even the case for highly utilised airports such as London Heathrow (LHR). However, it is difficult to define universally applicable CUI values that describe maximum airport capacity utilisation considering the demand side. While such values vary from airport to airport and even from runway system to runway system, maximum CUI values are in a range from 0.75 to 0.86 and might be even lower for airports comprising of many runways in a complex layout. Asian airports, for example, tend to have more belly-shaped ranking curves compared to their US and European counterparts, resulting in slightly higher CUI values, while very large airports with a high portion of domestic traffic such as Atlanta Hartsfield Jackson (ATL) achieve more pronounced peaking as the annual air traffic ranking curves clearly indicate. However, while the CUI is a valuable tool for analysing traffic structures at airports and deriving a

Modelling future airport capacity and capacity utilisation Chapter | 7

163

meaningful concept of annual airport capacity or service volume, the large number of airports operating well below their capacity during peak hours makes it difficult to develop a robust and simple econometric model of generic airport capacity on the basis of the CUI concept. As we have already outlined before, the two requirements of the model to be robust as well as simple to use are a conditio sine qua non.

7.2

Model theory and parameter estimation

We have chosen an approach for estimating future airport capacity which is based on data envelopment analysis (DEA) and regression analysis. DEA is a non-parametric empirical method of operations research to estimate production frontiers employing linear programming techniques. This approach was first described in Charnes et al. (1978). A good overview of DEA can also be found in Cooper et al. (2007). Today, DEA is a standard tool for efficiency analyses and benchmarking of so-called decision-making units (DMUs). DEA allows us to compare DMUs which differ in their input and output structure. Examples of such DMUs comprise hospitals, energy production or cost-/profit-centres of large organisations. A major advantage of DEA is that there is no need to explicitly describe the production function, not even the functional form; hence, it is a non-parametric method. Instead, the production function is unveiled by empirical data. However, therein lies a major weakness of DEA: the results depend very much on the DMUs and the selection of inputs and outputs upon which the DEA is based. Increasing the number of inputs and outputs results in an increase of the number of efficient DMUs, and it is difficult to select between different DEA setups (Berg, 2010). Thus we have chosen an effective DEA setup. We have taken two inputs and one output and a constant returns to scale output-oriented DEA formulation. Instead of measuring the efficiency of DMUs, we analyse airport capacity utilisation: which airport achieves what output given a particular input structure? This is analogous to measuring efficiency. Inputs are the number of runways [Digital Aeronautical Flight Information Files (DAFIF), 2016] and the number of operating hours of an airport [Official Airline Guide (OAG), 2016]. The number of runways is simply added with the exception that pure independent runway systems are counted as single runway systems with the appropriate fraction of their annual aircraft movements. For example, London Heathrow (LHR) is represented as an airport with one runway and half the aircraft movements. Beijing (PEK) is also counted as an airport with one runway but only a third of its annual aircraft movements because of three independent runways. Operating hours of an airport are defined as hours with more than five aircraft movements in the case of larger airports to reflect, for example, night flight restrictions. In the case of small airports, all hours with at least one

164

PART | II Models for assessing mitigation strategies

aircraft movement are counted as operating hours, as these hours contribute to a significant part of their total air traffic volume. The output is given by the number of annual aircraft movements of an airport [Official Airline Guide (OAG), 2016] or an appropriate fraction in the case of independent runways. The idea behind the ‘fraction’ approach is that we make the simplifying assumption that the capacity of an independent runway system equals the capacity of a single runway multiplied by the number of runways. This model assumption enables us to use the information in the data more efficiently by imposing constraints based on prior knowledge. It should be added that airports with two independent runways have a slightly higher capacity per runway than single runway airports, since often aircraft can be assigned to each runway so as to optimise the runway utilisation, for instance, by sequencing following aircraft by their weight. To conduct DEA, we employed the Robust Data Envelopment Analysis (DEA) for R package (rDEA) (Simm and Besstremyannaya, 2016). The output/input structure, that is the ratio of the number of annual aircraft movements per runway and of average number of aircraft movements per operating hour, is assumed to be constant over time for each airport. This ratio varies between airports and describes their production characteristics. For example, Chicago O’Hare (ORD), an airport with a complex system of eight runways and rather small aircraft, with on average of 90 passengers per flight, achieves on average about 109 aircraft movements per operating hour, but only about 107,000 annual aircraft movements per runway. Chicago O’Hare has a high average hourly throughput due to pronounced peaking characteristics, the high number of runways and rather small aircraft, but because of its complex runway system, it has a rather small annual throughput per runway. On the other hand, London Heathrow, an airport with two independent runways and much larger aircraft, with on average of 159 passengers per flight, achieves on average about 72 aircraft movements per operating hour, but around 238,000 annual aircraft movements per runway. Hourly throughput is limited by the rather low number of runways and larger aircraft, but the independent runway system allows for a high annual throughput per runway, even though the airport is used primarily during the daytime and not at night because of a curfew from 22 p.m. to 5 a.m. These examples show that there is a great deal of implicit information in the data which makes this problem well suited for DEA. Furthermore, the assumption of a constant input structure per airport for the forecast period seems to be reasonable. Before we discuss the results of the DEA in more detail, Fig. 7.1 gives an overview of the generic airport capacity model estimation of which the DEA is a part. The first step is the aforementioned DEA to estimate current airport capacity for a sample of airports. For this, we have chosen the largest 200 airports in terms of 2016 aircraft movements, because they belong to the

Modelling future airport capacity and capacity utilisation Chapter | 7

165

Performing data envelopment analysis (DEA) of the 200 largest airports in terms of aircraft movements

Current capacity utilisation per airport

Computing average number of aircraft movements per runway and operating hour at highest possible capacity utilisation

Mean values per runway class at highest possible capacity utilisation

Estimating functional relationship between number of runways and average number of aircraft movements per runway and operating hour at highest possible capacity utilisation FIGURE 7.1 Estimation of the generic airport capacity model.

large number of airports already suffering from a lack of capacity or are likely to do so within the next 20 30 years, as described in Chapter 3, Capacity utilisation at airports worldwide, and Chapter 4, Constrained and under-utilised airports. Furthermore, including smaller airports does not change the efficient frontier which we want to estimate. In a second step, we calculated the average number of aircraft movements per runway and operating hour at the highest possible level of capacity utilisation for each airport. This was deduced from the annual distribution of hourly volumes as given by the volume ranking curves of those airports which have been operating at their capacity limit for many years, such as London Heathrow. The CUI at these airports reaches a maximum of about 0.86, depending on the complexity of operations at multi-runway airports. The so-called maximum average number of aircraft movements stands, thus, for a practical or sustainable capacity which can be achieved over several hours and may be exceeded in peak hours such as the 5% peak hour. The hourly capacity of a runway, which may serve as a short-time capacity of say one hour, is thus higher than the maximum average number of aircraft movements. In fact, this capacity value is defined by the ratio of the maximum average number of aircraft movements and the highest possible CUI. In the case of a single runway or two independent parallel runways the

166

PART | II Models for assessing mitigation strategies

capacity is thus 15% 20% higher than the maximum average number of aircraft movements. This defined capacity also represents a practical capacity, but without any indication of the average delay per aircraft movement, as is normally the case when we speak of the practical capacity in contrast to the theoretical capacity (see Chapter 2: Concepts of capacity and methods of estimation). As has been mentioned, comparable delay statistics are not available for a global capacity analysis. To compute the maximum average number of aircraft movements, we first calculated the annual service volume per airport by dividing the actual number of annual aircraft movements by the capacity utilisation as estimated by the DEA from step one (see Fig. 7.2). Then, the annual service volume is divided by the number of runways and operating hours of the airport. Based on this, mean values per airport class are calculated. Airport classes are defined by the number of runways, whereby airports with two runways are further subdivided into those with parallel independent, parallel dependent and crossing runways. Two runways are defined as independent if the two parallel runways are not closer than 1500 m. The last step is to perform a regression analysis based upon the mean values from step two. Thereby, it is possible to generalise the results from step two and extrapolate to runway systems with more than eight runways. So far, there are only three airports with six runways [Denver (DEN), Detroit Metropolitan Wayne County (DTW) and Amsterdam Schiphol (AMS)], one with seven runways [Dallas/Fort Worth (DFW)] and one with eight runways [Chicago O’Hare (ORD)]. Thus, performing the regression analysis on top of 300,000

London Gatwick (LGW): 0% capacity reserve

Annual aircraft movements per runway

100% LGW

250,000 Jakarta SoekarnoHatta (CGK): 19% capacity reserve

200,000

PEK

97% LGW and 3% ATL London Heathrow (LHR): 9% capacity reserve *

83% LGW and 17% ATL

Dubai (DXB): 10% capacity reserve

HKG

31% LGW and 69% ATL

SIN

Atlanta HartsfieldJackson (ATL): 0% capacity reserve

Los Angeles (LAX): 13% capacity reserve

150,000

100% ATL

100,000 Chicago O'Hare (ORD): 10% capacity reserve DEN

50,000

0

DFW

AMS

* Excluding annual movement cap of 480,000

0

20

40

60

80

100

Average aircraft movements per operating hour

FIGURE 7.2 Efficient or capacity frontier of the 200 sample airports.

120

140

Modelling future airport capacity and capacity utilisation Chapter | 7

167

the mean values yields more robust results, as we will see later in this chapter in more detail. Fig. 7.2 shows the efficient frontier, or what we call a capacity frontier, of the 200 sample airports for estimating the actual capacity utilisation. The number of annual aircraft movements represents the output and the number of runways and operating hours are the inputs for the DEA model. Typically, the two axes of a DEA diagram with one output and two inputs are defined as ‘input 1/output’ and ‘input 2/output’, resulting in an efficient frontier which is orientated to the origin of the diagram (see, e.g., Cooper et al., 2007). However, for better interpretation of the model results we have chosen to invert the axes’ definition, that is output to input, which means ‘annual aircraft movements per runway’ and ‘average aircraft movements per operating hour’. Such a modification of a standard DEA diagram results in a nonlinear middle part of the efficient frontier in Fig. 7.2. This affects the presentation of the model results but has no effect on their computation. The capacity frontier is defined by London Gatwick (LGW, one runway) and Atlanta Hartsfield Jackson (ATL, five runways) airports, which are the reference airports. London Gatwick has the highest number of annual aircraft movements per runway, and ATL has the highest average aircraft movements per hour. Therefore these two airports have 100% capacity utilisation in the sense that, according to the DEA employed, the actual traffic volume represents the maximum throughput among the airports of the sample of the 200 biggest airports. The capacity utilisation of the remaining 198 airports is computed on the basis of virtual reference airports on the capacity frontier which have the same input structure. This is illustrated in Fig. 7.2 by projections from the origin through the airport considered onto the capacity frontier. The ratio of the dashed part of the arrow to the full arrow, that is the dashed plus the solid part, is equal to the capacity utilisation of the airport. Virtual reference airports are a linear combination of the airports that form the capacity frontier: LGW and ATL. For illustration purposes, we have highlighted five example airports: G

G

Jakarta Soekarno Hatta (CGK), an airport with two independent runways which is benchmarked against 100% LGW. From this comparison, CGK has a capacity reserve of 19%. London Heathrow (LHR), an airport with two independent runways which is benchmarked against a linear combination of 97% LGW and 3% ATL. From this comparison, LHR has a capacity reserve of 9%, if we do not account for the annual movement cap of 480,000. However, due to the movement cap, LHR is currently virtually fully utilised. Nevertheless, LHR potentially could handle more flights. LHR managers have proposed to raise the cap to about 500,000 movements per year (Evening Standard, 2016).

168 G

G

G

PART | II Models for assessing mitigation strategies

Dubai (DXB) airport, an airport with two dependent parallel runways which is benchmarked against a linear combination of 83% LGW and 17% ATL. From this comparison, DXB has a capacity reserve of 10%. Los Angeles (LAX) airport, an airport with four runways which is benchmarked against a linear combination of 31% LGW and 69% ATL. From this comparison, LAX has a capacity reserve of 13%. Chicago O’Hare (ORD) airport, an airport with eight runways which is benchmarked against 100% ATL. From this comparison, ORD has a capacity reserve of 10%.

Fig. 7.2 displays a cone which is defined by a dashed line from the origin to LGW and ATL, respectively. Most airports are within the cone, which basically means that the reference airports cover a wide range of output/input combinations. This is helpful for identifying virtual reference airports that are combinations of LGW and ATL. However, there are some airports a little outside the cone. Beijing (PEK), Hong Kong (HKG) and Singapore Changi (SIN) airports are examples that are to the left of the cone and have a high value of annual aircraft movements per runway compared to their value of average aircraft movements per operating hour. This is a result of their above-average number of operating hours (more than 8700 hours per year) and rather small number of runways (two for HKG and SIN and three for PEK). On the other hand, Amsterdam Schiphol (AMS), Denver (DEN) and Dallas/Fort Worth (DFW) airports are examples that are to the right of the cone and have a low value of annual aircraft movements per runway compared to their value of average aircraft movements per operating hour. This is a result of their below average number of operating hours (less than 7200 hours per year for DEN and DFW and less than 7400 per year for AMS) and rather high number of runways (six for AMS and DEN and seven for DFW). The approach bears a resemblance to the capacity envelopes discussed in Chapter 2, Concepts of capacity and methods of estimation (Gilbo, 1993, and Fig. 2.2 for Munich airport). However, while the capacity envelopes are based upon output measures, that is arrivals and departures, our DEA approach employs relative performance indicators of annual aircraft movements per runway and average aircraft movements per operating hour. Furthermore, capacity envelopes help to assess hourly capacity and the relationship with the arrival/departure mix, while our model aims more at the long-term capacity and focuses on the average hour volume at full capacity utilisation and a balanced arrival/departure mix. Table 7.1 shows the top 50 airports ranked by the degree of capacity utilisation derived from the DEA. As has been mentioned, based on the capacity utilisation we can eventually calculate an annual service volume per airport (column 9) and a maximum average capacity per runway per operating hour (column 10), that is at the highest possible annual capacity

TABLE 7.1 Results from the data envelopment analysis (DEA) for the top 50 airports in terms of 2016 capacity utilisation. 1

2

3

4

5

6

7

8

9

10

11

12

No.

IATA code

Airport name

ACM 2016

CUI 2016

OH 2016

RWYs 2016

CU DEA2

eASV (DEA)

MAC/RWY/ OH (DEA)

CU DEA3

Cols. 8 11

1

ATL

Atlanta Hartsfield Jackson

875,211

0.67

7223

5

100

875,211

24.2

100

0

2

LGW

London Gatwick

273,606

0.69

7334

1

100

273,606

37.1

100

0

3

MEX

Mexico City Benito Juarez

406,022

0.68

8083

2

93

436,778

26.9

93

0

4

LHR

London Heathrow

476,226

0.82

6645

2

91

523,302

39.3

91

0

5

ORD

Chicago O’Hare

855,194

0.63

7811

8

90

946,134

15.1

90

0

6

DXB

Dubai

403,542

0.82

8784

2

90

447,947

25.5

90

0

7

LGA

New York LaGuardia

372,399

0.74

6827

2

90

415,506

30.4

99

29

8

LAX

Los Angeles

635,265

0.69

7753

4

87

733,198

23.6

89

22

9

SAW

Istanbul Sabiha Gokcen

219,326

0.63

7161

1

81

270,751

37.6

81

0

10

CGK

Jakarta Soekarno Hatta

440,707

0.71

8024

2

81

547,212

34.0

100

219

11

IST

Istanbul Ataturk

457,446

0.76

8734

3

76

598,129

22.8

76

0

12

DUB

Dublin

201,368

0.67

7122

1

75

270,049

37.7

75

0 (Continued )

TABLE 7.1 (Continued) 1

2

3

4

5

6

7

8

9

10

11

12

No.

IATA code

Airport name

ACM 2016

CUI 2016

OH 2016

RWYs 2016

CU DEA2

eASV (DEA)

MAC/RWY/ OH (DEA)

CU DEA3

Cols. 8 11

13

DFW

Dallas/Fort Worth

646,079

0.68

7189

7

74

872,443

17.3

74

0

14

PEK

Beijing

606,105

0.79

8757

3

74

820,818

31.2

74

0

15

HKG

Hong Kong

399,489

0.69

8739

2

73

547,212

31.3

73

0

16

CAN

Guangzhou Baiyun

430,241

0.74

8374

3

73

592,467

23.5

73

0

17

XMN

Xiamen Gaoqi

194,030

0.80

6950

1

73

267,205

38.2

73

0

18

CTU

Chengdu Shuangliu

310,453

0.86

7653

2

72

430,010

27.9

72

0

19

CLT

Charlotte Douglas

511,794

0.75

6664

4

72

714,935

26.7

72

21

20

MUC

Munich

374,718

0.64

6842

2

71

529,674

38.7

71

0

21

EWR

Newark Liberty

400,757

0.72

7337

3

70

575,353

26.0

73

24

22

FUK

Fukuoka

167,574

0.85

5490

1

69

242,227

44.1

69

0

23

BOM

Mumbai Chhatrapati Shivaji

307,071

0.77

8694

2

69

446,555

25.7

69

0

24

SEA

Seattle Tacoma

395,526

0.68

7387

3

69

576,423

25.9

72

24

25

PHX

Phoenix Sky Harbor

381,678

0.69

7129

3

67

571,770

26.6

70

23

26

SAN

San Diego

173,266

0.71

6589

1

66

260,870

39.5

82

216

27

SIN

Singapore Changi

357,945

0.69

8751

2

65

547,212

31.3

65

0

28

BLR

Bengaluru Kempegowda

176,468

0.67

8442

1

64

273,606

32.3

80

216

29

DEL

Delhi Indira Gandhi

383,259

0.70

8784

3

64

598,905

22.7

64

0

30

CSX

Changsha Huanghua

166,711

0.78

6639

1

64

262,093

39.2

64

0

31

FRA

Frankfurt

454,775

0.69

6749

4

63

716,242

26.5

63

0

32

DEN

Denver

545,318

0.59

7173

6

63

868,642

20.1

63

0

33

TAO

Qingdao Liuting

167,266

0.81

6919

1

63

267,187

38.2

63

0

34

SUB

Juanda

162,080

0.66

6607

1

62

261,603

39.3

77

215

35

CDG

Paris Charles de Gaulle

448,506

0.67

7225

4

62

724,247

25.0

62

0

36

URC

¨ ru¨mqi Diwopu U

165,285

0.74

6935

1

62

266,905

38.3

62

0

37

HND

Tokyo Haneda

450,207

0.64

7730

4

61

733,002

23.6

61

0

38

BKK

Bangkok Suvarnabhumi

331,898

0.71

8760

2

61

547,212

31.2

61

0

39

KMG

Kunming Changshui

328,953

0.82

7287

2

60

544,909

37.3

60

0

40

KUL

Kuala Lumpur

356,414

0.70

8378

3

60

592,694

23.5

60

0 (Continued )

TABLE 7.1 (Continued) 1

2

3

4

5

6

7

8

9

10

11

12

No.

IATA code

Airport name

ACM 2016

CUI 2016

OH 2016

RWYs 2016

CU DEA2

eASV (DEA)

MAC/RWY/ OH (DEA)

CU DEA3

Cols. 8 11

41

CKG

Chongqing Jiangbei

255,462

0.83

7369

2

60

425,531

28.6

60

0

42

LIM

Lima Jorge Chavez

164,080

0.60

7837

1

60

273,606

34.3

75

215

43

JFK

New York John F. Kennedy

440,398

0.70

7924

4

60

736,280

23.1

60

0

44

STN

London Stansted

154,785

0.59

6508

1

59

260,224

39.6

59

0

45

MNL

Manila Ninoy Aquino

261,597

0.72

8398

2

59

442,076

26.2

59

0

46

SVO

Moscow Sheremetyevo

258,837

0.69

8681

2

58

446,434

25.7

58

0

47

SFO

San Francisco

419,364

0.68

7540

4

57

729,765

24.1

59

21

48

GVA

Geneva

148,053

0.66

6405

1

57

257,814

40.1

57

0

49

SZX

Shenzhen Bao’an

314,042

0.79

8095

2

57

547,212

33.6

57

0

50

SYD

Sydney Kingsford Smith

317,815

0.68

6222

3

57

556,976

29.6

57

0

362,892

0.71

7533

2.5

70

513,323

30.2

72

22

Ø

Cols. 8 11, difference between the values of columns 8 and 11; CU, capacity utilisation in %; CUI, capacity utilisation index; DEA, data envelopment analysis; DEA2, DEA with two inputs; DEA3, DEA with three inputs; MAC/RWY/OH, maximum average capacity per runway per operating hour; OH, operating hours; RWYs, runways.

Modelling future airport capacity and capacity utilisation Chapter | 7

173

utilisation. The values of column 10 form the input for estimating the generic airport capacity model in step three of Fig. 7.1. Before we proceed, we want to discuss briefly an alternative DEA setup. As Berg (2010) pointed out, it is difficult to decide between different DEA setups. We have seen in Chapter 3, Capacity utilisation at airports worldwide, that slot coordination is an influencing factor for modelling the relationship between 5% peak hour volume and average hour volume. Thus slot coordination affects airport capacity and capacity utilisation. We have to check, therefore, whether including slot coordination improves our model significantly. Including more input variables leads to a further discrimination between airports and an increase in the number of airports which have 100% capacity utilisation. This ultimately leads to lower capacity values in the generic airport capacity model for a given runway system. Slot coordination has been considered in two categories in the alternative DEA setup (IATA, 2016): G G

Airport is not Level 3 slot coordinated Airport is Level 3 slot coordinated

Columns 11 and 12 of Table 7.1 present the results. Column 11 shows the new capacity utilisation values and column 12 the difference in capacity utilisation between the original and the alternative DEA setup. The alternative DEA setup yields three airports with 100% capacity utilisation: G G G

LGW (capacity utilisation in the original DEA setup: 100%) ATL (capacity utilisation in the original DEA setup: 100%) Jakarta Soekarno Hatta (CGK, capacity utilisation in the original DEA setup: 81%)

There is a significant increase of capacity utilisation, greater than 10% points, at Lima Jorge Chavez (LIM, 115% points), Juanda (SUB, 115% points), San Diego (SAN, 116% points), Bengaluru Kempegowda (BLR, 116% points) and CGK (119% points) airports. This is a result of a different capacity frontier, which is now defined by LGW, ATL and CGK. The major difference between the original and the alternative DEA setup is a lower annual service volume for a number of airports with one or two runways. This value drops from between 262,000 and 274,000 aircraft movements to between 210,000 and 220,000 aircraft movements per runway. From our point of view, these values seem to be rather low. As the difference in capacity utilisation is otherwise very small, we have chosen the original setup for the capacity forecast. However, a model assessment with actual airport data will be carried out at the end of this chapter. Nevertheless, this shows how data-sensitive DEA is. Including more inputs or outputs increases the number of efficient DMUs. We further checked for setups with variable returns to scale or splitting the dataset completely according to slot coordination status but this leads to increasing numbers of airports which were supposed to operate at their capacity limit. The results

174

PART | II Models for assessing mitigation strategies

became, thus, more unreasonable when checking key statistics such as 5% peak hour volume, CUI and annual aircraft movements. Considering air traffic control in the DEA, in other words, whether instrument flight rules as for example in Europe or visual flight rules as in the United States normally apply, yields similar results in terms of annual service volume. Therefore adding features such as inputs and outputs or more flexible production frontiers have to be carried out in a meaningful way, and even then, more complexity is not always better. Nevertheless, this clearly illustrates how difficult the choice is between different setups. For our case, it is essential for the setup to remain effective and to choose a smart problem representation, such as the ‘fractional’ approach, to use the information in the dataset as efficiently as possible and to keep problem complexity as low as possible. Table 7.2 displays the airport class means of the maximum average number of aircraft movements per runway and operating hour and their variance based upon the results of the DEA (see Table 7.1, column 10, all 200 airports), which is the basis for step three of the generic airport capacity model (see Fig. 7.1), that is the regression analysis. As expected, the maximum average number of aircraft movements per runway and operating hour— representing a practical capacity which can be maintained over several hours—declines gradually with increasing numbers of runways. The variance tends to decline with increasing numbers of runways as well but is only available for airports of up to six runways. The reason for the latter is that

TABLE 7.2 Maximum average aircraft movements per runway and operating hour by airport class. Variance (σi2)

Number of runways

Maximum average aircraft movements per runway and operating hour based on DEA

1

37.9

9.18

2, independent

37.9

9.18

2, dependent

29.8

20.99

2, crossing

29.8

5.72

3

26.3

6.18

4

25.3

2.50

5

23.3

3.07

6

20.1

4.48

7

17.3

8

15.1

DEA, data envelopment analysis.

Modelling future airport capacity and capacity utilisation Chapter | 7

175

Maximum average aircraft movements per runway and operating hour

40

35 30 25

y = 47.708 – 11.439x0.5 R 2 = 97.36%

20 15

10 5 0

0

1

2

3

4

5

6

7

8

9

Number of runways per airport

FIGURE 7.3 Regressing maximum average aircraft movements per runway and operating hour on the number of runways.

there is only one airport with seven and one with eight runways. The airport class of six runways consists of three airports (DEN, AMS and DTW). Fig. 7.3 displays the values of Table 7.2 in a chart which visualises the functional relationship between maximum average aircraft movements per runway and operating hour and the runway system of an airport as represented by the number of runways. In Fig. 7.3 the value for airports with two runways is an average of dependent and crossing systems, while the capacity value per runway for two independent runways is equal to the value for single runway airports. The regression line in Fig. 7.3 has been estimated by ordinary least squares (see, e.g., Greene, 2011). For airports with two dependent and crossing runways the two original data points were considered for model estimation, so that we have in total nine data points. Table 7.3 presents the estimation results of the generic airport capacity model. The square root of the number of runways per airport is the independent variable. We have taken this figure because the maximum average aircraft movements per runway and operating hour declines slightly less than linearly as depicted in Fig. 7.3. The model has been estimated on class mean values; nevertheless, both variables are highly significant. R2 is very high with 97.36%. However, this should not be overemphasised as this is a rather parsimonious model approach. The model allows us to calculate annual service volumes for a generic airport with a given number of runways and operating hours. The annual service volume is thereby the product of the hourly capacity per runway

176

PART | II Models for assessing mitigation strategies

TABLE 7.3 Estimation results of the generic airport capacity model with square root transformation. No.

Variable

Coefficient

1

Constant

2

Square root of number of runways R2

Standard error

p-value

47.70753

1.63136

0.00000

2 11.43949

0.76903

0.00001

97.36%

(column 3 of Table 7.4), the number of runways and the number of operating hours per year. We call the application of the model in this way the base mode, since the model yields the same results of annual airport capacity for airports with the same number of runways (in the case of two runways differentiated by independent and dependent or crossing runways) and operating hours. Table 7.4 provides the base case capacities for runways of airports with up to eight runways (but not limited to) and 18 operating hours per day year-round (column 4 of Table 7.4). Furthermore, column 5 displays the relative capacity gain by adding one runway at a time. As expected, adding more and more runways to an airport yields typically decreasing capacity gains, and significant capacity gains in terms of annual service volume can be only achieved with up to five to six runways if operating hours are held constant (this is not the case for peak hour volumes, as we will see later in this chapter). However, in the next chapter, we will show that expanding an airport gradually from one or two runways to up to six runways or more is, for our analyses in Part III of the book, of lesser importance, because such an enlargement will be in many cases an insurmountable task. Column 6 displays the annual service volume per runway and airport class. Again, this value declines with increasing numbers of runways. There are only five airports with six or more runways in the data sample of 200 airports. Thus a closer analysis of their modelled capacity values seems appropriate. Dallas/ Fort Worth (DFW), an airport with seven runways, had 646,079 aircraft movements in 2016 and on average about 20 operating hours per day (Table 7.1). Based on the DEA (Table 7.1 and Fig. 7.1), the annual service volume was estimated to be 872,443 aircraft movements, thus a capacity utilisation of 74%. A generic airport with seven runways and 20 operating hours per day has an annual service volume of 802,134 3 20/18 5 891,260 aircraft movements (based on Table 7.4 and corrected for operating hours). Chicago O’Hare (ORD), an airport with eight runways, handled 855,194 aircraft movements in 2016 during on average about 21 operating hours per day (Table 7.1). Based on the DEA, we calculated an annual service volume

Modelling future airport capacity and capacity utilisation Chapter | 7

177

TABLE 7.4 Applying the generic capacity model in base mode. 1

2

3

4

5

6

Number of runways

MAACM/ RWY/OH (DEA)

MAACM/ RWY/OH (RA)

ASV (18/ 365)

Increase of ASV by adding one RWY (%)

ASV/ RWY (18/ 365)

1

37.9

36.3

238,281

2, independent

37.9

36.3

476,562

100

238,281

2, dependent

29.8

31.5

414,300

74

207,150

2, crossing

29.8

31.5

414,300

74

207,150

3

26.3

27.9

549,786

15a; 32b

183,262

4

25.3

24.8

652,494

19

163,124

5

23.3

22.1

726,906

11

145,381

6

20.1

19.7

776,046

7

129,341

7

17.3

17.4

802,134

3

114,591

8

15.1

15.4

806,888

1

100,861

238,281

ASV, annual service volume; DEA, data envelopment analysis; MAACM/RWY/OH, maximum average aircraft movements per runway per operating hour; RA, regression analysis; RWY, runway; 18/365, 18 hours 365 days a year. a Adding a third runway to an airport with two independent runways. b Adding a third runway to an airport with two dependent/crossing runways.

of 946,134 aircraft movements, thus a capacity utilisation of 90%. A generic airport with eight runways and 21 operating hours per day has an annual service volume of 806,888 3 21/18 5 941,369 aircraft movements (based on Table 7.4 and corrected for operating hours). Finally, based on the DEA, the average annual service volume of airports with six runways is estimated to be around 878,000 aircraft movements, and for the three airports with six runways ranges from almost 869,000 to more than 892,000. Average operating hours per day are nearly 20 hours. The generic airport with six runways of Table 7.4 results in an annual service volume of about 862,000 aircraft movements for 20 operating hours per day. Thus we conclude from the discussion in this chapter and the brief model assessment for the five largest airports in terms of their number of runways that the modelled values are sufficiently precise for a global or network encompassing analysis in Part III of the book. Fig. 7.4 illustrates the incremental mode of the generic capacity model, which is typically employed for already existing airports. After performing a

178

PART | II Models for assessing mitigation strategies

Perform data envelopment analysis (DEA) for airports under study

Current capacity utilisation

Calculate annual service volume for the highest possible capacity utilisation of the current runway system

Estimate of currrent airport capacity

Calculate the capacity difference between the current and future runway system based on the generic capacity model

Estimate of capacity improvement

Add the generic capacity difference to the current annual service volume of the airports under study FIGURE 7.4 Applying the generic capacity model for forecasting in incremental mode.

DEA, only the capacity gain by adding one runway at a time is added to the estimated current capacity limit of an airport. Under the assumption of a constant input structure, as discussed earlier, the multiplier for the capacity gain can be retrieved from column 5 of Table 7.4. For example, upgrading a single runway airport to two independent runways yields a multiplier of 2 (100% capacity gain), while upgrading an airport with three runways to four runways has a multiplier of just 1.21 (21% capacity gain).

7.3 Model application: comparison of model results with actual traffic data The question we want to answer in the following section is how do the model results compare with actual peak hour capacities of airports? With that we want to present a deeper insight into the capabilities of the chosen approach.

Modelling future airport capacity and capacity utilisation Chapter | 7

179

This comparison has been carried out for those airports in Tables 2.2 and 2.3 of Chapter 2, Concepts of capacity and methods of estimation, which are among the 200 airports the analysis is based upon. As can been seen from these tables, airport capacity is typically presented as some kind of sustained or practical capacity (‘declared capacity’ if the airport is slot coordinated). The key concept of the model aims to identify the maximum average number of aircraft movements per runway and operating hour. Therefore we have tried to connect both concepts by means of the CUI. As described in Chapter 3, Capacity utilisation at airports worldwide, the CUI is defined by the ratio of average hourly traffic volume to the 5% peak hour volume. If the airport is at its capacity limit, the average hourly traffic volume is basically equal to the maximum average number of aircraft movements per runway and operating hour multiplied by the number of runways. Therefore dividing the latter metric by the CUI value yields a decent approximation of 5% peak hour volume and, thus, of practical capacity. However, the question remains which CUI value should be taken as the appropriate one when we want to estimate 5% peak hour volume? As can be seen from Table 7.1, the maximum attainable CUI value tends to decrease with the number of runways, in other words, with the complexity of the runway system of an airport. Fig. 7.5 displays the CUI values of those airports in Table 7.1, which have the highest CUI values within each airport capacity class. These

Maximum CUI value of top 50 airports in terms of DEA capacity utilisation

1 0.9 0.8

FUK

DXB

y = 0.8686 – 0.0319x R² = 89.44%

CLT IST

0.7

DFW ATL

0.6

AMS

ORD DEN

0.5

y = 0.8704 – 0.0337x R² = 81.51%

0.4 0.3 0.2 0.1 0

0

1

2

3

4

5

6

7

8

9

Number of runways FIGURE 7.5 Reference CUI values in relation to the number of runways (two runways applies to dependent and crossing runways, two independent runways is equal to single runway). CUI, capacity utilisation index.

180

PART | II Models for assessing mitigation strategies

represent the top 50 airports in terms of capacity utilisation of the sample of 200 airports for the DEA. Here, fully independent runway systems such as London Heathrow or Beijing airports are assigned to single runway airports, and the category of two-runway airports comprises only those with dependent and crossing runways. As can be seen, there is a distinct decrease of the maximum CUI value with an increasing number of runways. The value of Denver (DEN) airport seems to be rather low; therefore we have chosen the CUI value of Amsterdam Schiphol (AMS) as an alternative, in spite of the fact that AMS is ranked 60th in terms of capacity utilisation. The approach chosen ensures that the maximum CUI value for each generic airport can be found as given by the number of runways, and airports with high CUI values, but sub-maximum peak hour volumes, that is a lack of capacity utilisation, can be avoided. Nevertheless, this procedure is only as good as the data that is employed, and there are, in general, only a few airports with five or more runways worldwide. Therefore the results should be interpreted with caution and should not be overemphasised. Fig. 7.5 shows two linear regression lines. The lower dashed line is the original regression (with Denver airport, called the DEN model), while the upper solid line is the alternative model (with Amsterdam Schiphol airport, called the AMS model). Table 7.5 displays detailed results of both regressions. While both models perform very well in terms of goodness-of-fit (R2 . 80% in both cases), and the estimated coefficients are significant (p-value ,0.01 for the DEN model and even ,0.001 for the AMS model), we have chosen to retain the AMS model for further analysis, as it is superior in terms of R2 and significance of coefficient estimates. Besides, the regression results tend to support the hypothesis that the CUI value for

TABLE 7.5 Estimation results of the model ‘capacity utilisation index (CUI) by number of runways’. Model

No.

Variable

AMS

1

Constant

2

Number of runways

Coefficient

Standard error

p-value

0.86857

0.02260

0.00000

2 0.03190

0.00448

0.00038

R2

89.44% DEN

1

Constant

2

Number of runways

0.87036

0.03308

0.00000

2 0.03369

0.00655

0.00213 81.51%

Modelling future airport capacity and capacity utilisation Chapter | 7

181

TABLE 7.6 Modelled maximum 5% peak hour volume for comparison purposes (two runways applies to dependent and crossing runways, two independent runways is equal to two times single runway values). Number of runways per airport

Reference CUI

Maximum average aircraft movements per operating hour

Modelled maximum 5% peak hour volume

1

0.84

36.3

43.3

2

0.80

63.1

78.4

3

0.77

83.7

108.3

4

0.74

99.3

134.0

5

0.71

110.6

156.0

6

0.68

118.1

174.4

7

0.65

122.1

189.2

8

0.61

122.8

200.2

CUI, capacity utilisation index.

Denver airport may be an outlier, which we have corrected by taking Amsterdam Schiphol instead. Based upon the estimated AMS model, we derive so-called reference CUI values for each airport capacity type as given by the number of runways (see Table 7.6). These reference CUI values represent maximum values of airports by the number of runways and decrease with an increasing number of runways. Values start at 84% for single runway airports and airports with independent runway systems, such as London Heathrow or Beijing airports, and reach a value of 61% for airports with eight runways, such as Chicago O’Hare. Furthermore, as a result of the model approach, values for airports with more than eight runways can be computed as well. This is particularly useful in a forecast scenario, where adding runway capacity seems to be appropriate. This will be the topic of Chapter 8, Modelling future airport capacity enlargements and limits. Based upon the maximum or reference CUI values and maximum average number of aircraft movements per operating hour, we can calculate the modelled maximum 5% peak hour volumes by airport capacity class, which serve as a proxy for the hourly capacity of the runway system, as already discussed in Chapter 2, Concepts of capacity and methods of estimation, and Chapter 3, Capacity utilisation at airports worldwide, in this book. Thereby, it is possible to conduct a comparison of the model results of the DEA approach and the capacity values of Tables 2.2 and 2.3 of Chapter 2,

182

PART | II Models for assessing mitigation strategies

Concepts of capacity and methods of estimation [excluding Memphis airport (MEM), as it is mainly a cargo airport]. Tables 7.7 and 7.8 show the comparison results. For this purpose, we have added three columns to Tables 2.2 and 2.3. One for modelled 5% peak hour volumes and another for the current capacity utilisation as obtained from the DEA (Table 7.1 and Fig. 7.2). The last column is a more or less subjective evaluation of the results. To calculate the 5% peak hour volume for each airport, we divided the estimated annual airport capacity, which we have obtained from the DEA, by both the number of operating hours and the reference CUI value. Comparing observed with estimated volumes and capacity utilisation rates, we have grouped modelled 5% peak hour volumes and capacity utilisation based on the DEA in Tables 7.7 and 7.8 according to the following rather ‘soft’ rules: G

G

G

Good: Modelled 5% peak hour volumes are more or less within the capacity ranges and close to the declared capacity, respectively. Sufficient: Here, modelled 5% peak hour volume is significantly higher than the capacity ranges and declared capacity. However, these values are still acceptable, as capacity utilisation is also rather low. Therefore we assume that there are still long-term capacity reserves which the airport can use in the future. Low capacity ranges and declared capacity ranges may be the result of issues such as a lack of terminal or apron capacity, environmental constraints or just a current lack of demand. Examples include airports such as San Francisco (SFO, four crossing runways) and Du¨sseldorf (DUS, two dependent runways, with administrative restrictions). Modest: In the case of Denver airport (DEN), modelled 5% peak hour volume is significantly lower than compared to the capacity ranges of Table 7.7, even if we take the rather low capacity utilisation (63%) into account. As we have already seen in Fig. 7.5, the actual CUI value of DEN is rather low, which results in a larger spread between 5% peak hour and average hour volume.

Overall, the model produces quite reasonable 5% peak hour volume values compared to the capacity ranges of Table 7.7 and declared capacities of Table 7.8. Here, we have to consider that the 5% peak hour volume is a good proxy of hourly airport capacity, but not the highest peak hour volume an airport can handle; the 5% peak hour volume is rather a good representation of the practical hourly capacity (see Chapter 2: Concepts of capacity and methods of estimation). Furthermore, peak hour volumes show a high degree of variation, depending on which peak hour we choose, as we have seen from the analysis of traffic ranking curves. This variation increases with decreasing CUI values, as the traffic ranking curve becomes steeper. This is a particularly important point for large airports with a high number of

TABLE 7.7 Comparison of the modelled 5% peak hour volumes with the hourly capacity rate ranges of the airports of Table 2.2. Airport identifier and name

ATL

Atlanta Hartsfield Jackson

Aircraft operations (arrivals and departures) per hour Visual

Marginal

Instrument

216 226 (AP)

201 208 (AP)

175 190 (AP)

206 (DP)

166 169 (LIMC AP)

219 222 (DP)

Modelled 5% peak hour volume

Capacity utilisation based on DEA (%)

Result evaluation

170

100

Good

168

41

Sufficient

183 186 (DP)

168 179 (LIMC DP) BOS

Boston Logan

116 125

109 112

84 86

BWI

Baltimore Washington

68 80

64 80

62 64

104

39

Sufficient

CLT

Charlotte Douglas

176 182

161 162

138 147

144

72

Good

DCA

Ronald Reagan Washington

69 72

69 72

54 64

109

53

Sufficient

DEN

Denver

262 266 (AP)

224 279

224 243

178

63

Modest

266 298 (DP) DFW

Dallas/Fort Worth

226 264

194 245

170

187

74

Good

DTW

Detroit Metropolitan Wayne County

178 184

163 164

136

178

43

Good (Continued )

TABLE 7.7 (Continued) Airport identifier and name

Aircraft operations (arrivals and departures) per hour Visual

Marginal

Instrument

EWR

94 99 (AP)

76 84

Newark Liberty

Modelled 5% peak hour volume

Capacity utilisation based on DEA (%)

Result evaluation

68 70

101

70

Good

94 100 (DP) FLL

Fort Lauderdale Hollywood

74 82

66 72

56 66

93

44

Sufficient

HNL

Honolulu

117 120

91 105

60 77

135

23

Sufficient

IAD

Washington Dulles

150 159 (AP)

112 120 (AP)

108 111 (AP)

142

32

Good

156 164 (DP)

136 145 (DP)

125 132 (DP)

IAH

Houston George Bush

172 199

152 180

144 151

170

52

Good

JFK

New York John F. Kennedy

84 87 (AP)

85 86

74 84

125

60

Sufficient

90 93 (DP)

LAS

Las Vegas McCarran

122 128

106 111

78 83

134

48

Sufficient

LAX

Los Angeles

167 176

147 153

133 143

127

87

Good

LGA

New York LaGuardia

80 86

76 77

74 76

76

90

Good

MCO

Orlando

160 171

148 161

144

135

41

Good

MDW

Chicago Midway

64 84

64 74

52 70

170

23

Sufficient

MIA

Miami

132 150

132 148

100 104

122

47

Good

MSP

Minneapolis Saint Paul

156 167

142 151

114 141

136

52

Good

ORD

Chicago O’Hare

214 225

194 200

168 178

197

90

Good

PHL

Philadelphia

120 126

94 96

84 88

135

50

Sufficient

PHX

Phoenix Sky Harbor

138 145

108 109

96 101

103

67

Good

SAN

San Diego

48 57

48 52

48

47

66

Good

SEA

Seattle Tacoma

100 112

86 100

76 78

100

69

Good

SFO

San Francisco

100 110

90 93

70 72

130

57

Sufficient

SLC

Salt Lake City

148 150

138 140

114 120

106

42

Good

TPA

Tampa

113 150

95 115

90 95

105

27

Good

AP, arrival priority configuration; DEA, data envelopment analysis; DP, departure priority configuration; LIMC, low instrument.

186

PART | II Models for assessing mitigation strategies

TABLE 7.8 Comparison of the modelled 5% peak hour volumes with the declared capacities of the airports of Table 2.3. Airport (IATA code)

Declared capacity (aircraft movements per hour)

Modelled 5% peak hour volume

Capacity utilisation based on DEA (%)

Result evaluation

Single runway airports London Stansted (STN)

36 50

47

59

Good

Dublin (DUB)

38 46

45

75

Good

Stuttgart (STR)

42

48

38

Sufficient

London Gatwick (LGW)

38 50

44

100

Good

Geneva (GVA)

36

48

57

Sufficient

Airports with two intersecting runways Hamburg (HAM)

48

81

35

Sufficient

Warsaw (WAW)

38 42

79

33

Sufficient

Perth (PER)

38

72

24

Sufficient

Lisbon (LIS)

35 40

75

42

Sufficient

Airports with two close parallel runways Dubai (DXB)

57 65

63

90

Good

Manchester (MAN)

44 56

74

43

Sufficient

Berlin Tegel (TXL)

52

81

45

Sufficient

Nice (NCE)

40 50

81

41

Sufficient

43 45

79

50

Sufficient

a

Du¨sseldorf (DUS)

Airports with two independent parallel runways Munich (MUC)

90

92

71

Good

London Heathrow (LHR)a

79 89

94

91

Good

Oslo (OSL)

76 80

94

44

Good

Palma de Mallorca (PMI)

66

95

33

Sufficient

a Du¨sseldorf capacity is limited by administrative rules to a single runway capacity, London Heathrow has an annual movement cap of 480,000.

Modelling future airport capacity and capacity utilisation Chapter | 7

187

runways, as in these cases, CUI values tend to be lower, and thus the variation of peak hour volumes increases. When comparing modelled 5% peak hour volume with declared capacities of Level 3 airports, we see a rather good approximation with single runway airports and airports with two independent runways, however, rather high estimates for airports with intersecting and two close parallel runways. It seems that declared capacities in these cases represent the rather cautious results of capacity finding discussions in coordination committees, as can be seen by comparing declared capacities with those of single runways. The values of airports with intersecting runways are barely higher and even the values of airports with two close parallel runways do not differ greatly from single runway airports. Annual airport capacities as a means of annual service volume are computed on the basis of maximum average hourly volume (per airport or per runway for a given number of runways) in our approach. Since traffic ranking curves typically run much flatter around the average hourly volume of an airport than around peak hour volumes, even for lower degrees of capacity utilisation, we expect the model to perform better than Tables 7.7 and 7.8 suggest.

7.4

Conclusion

While a detailed airport capacity assessment produces more precise results, but lacks forecast and large-scale application capabilities, the purpose of this chapter was to present a robust approach to compute annual capacities as annual service volumes of airports worldwide that have an adequate precision. Thus, the approach presented is highly problem-specific and cannot be a substitute for a detailed airport capacity assessment in airport-specific studies. However, obtaining annual service volume of airports worldwide is a prerequisite for reflecting limited airport capacity in a global air traffic forecast. The next step is to consider how airport capacity evolves in future, which is the topic of the next chapter. The chosen approach is based upon key ideas of Chapter 3, Capacity utilisation at airports worldwide, and Chapter 4, Constrained and under-utilised airports, such as ranking traffic curves, 5% peak hour and average hourly volume, which lead to the concept of the CUI. If an airport operates close to its maximum technical capacity during peak hours, the CUI can serve as an indicator of what share of that capacity (approximated by the 5% peak hour volume) can be utilised realistically over the course of a year on average, taking the demand side and complexity factors, such as air traffic control and interactions between runways into account. Utilisation rates have been observed to lie between 59% and 86%, depending on the runway system. As a result, the CUI is a link between a rather technical view of airport capacity,

188

PART | II Models for assessing mitigation strategies

that is peak hour volume, and the demand side of air traffic and complexity factors, that is average hourly volume. However, we had to extend these ideas, because in many cases, 5% peak hour volumes have been found to be well below peak hour capacity. Furthermore, the maximum value of the CUI tends to depend on the runway system. This makes classical regression analysis difficult. Therefore we expanded our approach by incorporating DEA and conducted a regression analysis based upon the DEA results. In contrast to the CUI, capacity utilisation rates in the DEA are based on average hourly traffic volumes per airport and runway, respectively; hence, values of highest possible capacity utilisation can be reached. Nevertheless, 5% peak hour volumes can be modelled additionally for comparison purposes. This model approach enables us to compute current and future annual service volumes of airports worldwide. To assess the model, we have calculated modelled 5% peak hour volumes and compared them with capacity rates of US airports and declared capacities of mainly European airports. Here, the model performed satisfactorily. Furthermore, while the analysis of Chapter 3, Capacity utilisation at airports worldwide, and Chapter 4, Constrained and under-utilised airports, is about the current traffic situation, we can now forecast 5% peak hour volume and average hourly volume in a situation of the highest possible capacity utilisation.

References Berg, S., 2010. Water Utility Benchmarking: Measurement, Methodology, and Performance Incentives. International Water Association. Charnes, A., Cooper, W.W., Rhodes, E., 1978. Measuring the efficiency of decision-making units. Eur. J. Oper. Res. 2 (6), 429 444. Cooper, W.W., Seiford, L.M., Tone, K., 2007. Data Envelopment Analysis—A Comprehensive Text With Models, Applications, References and DEA-Solver Software. Springer, New York. Digital Aeronautical Flight Information Files (DAFIF), 2016. US Government, Washington D.C. Evening Standard, 2016. Heathrow Airport Bosses Propose Raising Cap on Flights to 500,000 a Year. Evening Standard, 9/29/2016. ,https://www.standard.co.uk/news/politics/heathrowbosses-propose-raising-cap-on-flights-to-500000-a-year-a3357071.html#comments. (accessed 09.07.18.). Gilbo, E., 1993. Airport capacity: representation, estimation, optimisation. IEEE Trans. Control Syst. Technol. 1 (3), 144 154. Greene, W.H., 2011. Econometric Analysis. Pearson Education, Harlow. IATA, 2016. Worldwide Slot Guidelines (WSG)—Annex 11.6. Montreal. Official Airline Guide (OAG), 2016. Market Analysis. Reed Travel Group, Dunstable. Simm, J., Besstremyannaya, G., 2016. Robust Data Envelopment Analysis (DEA) for R. CRAN. ,https://cran.r-project.org/web/packages/rDEA/rDEA.pdf..

Chapter 8

Modelling future airport capacity enlargements and limits Global air traffic has grown substantially in the past and the pace of growth has only been interrupted by oil and financial crises, terrorism and wars (see Figure 1.1). Between 1991 and 2016, the number of aircraft movements increased by around 140% and reached a volume of about 35.5 million aircraft departures in 2016. This means an annual average growth per year of almost 3.6% (compounded annual growth rate—CAGR). The number of air passengers has grown even stronger. The number of air passengers grew from 1.1 to 3.8 billion, an increase of 234%, which corresponds to an average rate of about 4.9% per year [International Civil Aviation Organization (ICAO), 2016b]. The long-term forecasts of aircraft manufacturers, such as Airbus and Boeing, as well as institutions, such as ICAO and Eurocontrol, differ only marginally and basically see a continuation of the past growth for the future. According to Boeing’s Current Market Outlook (Boeing, 2016), annual passenger volume is forecast to grow by a CAGR of 4.0% between 2016 and 2035. Airbus (2016) forecasts a 4.5% CAGR of annual passenger volume for the same time period. The ICAO (2016a) forecasts a 4.6% CAGR of revenue passenger kilometres for the period 2012
32 and 4.5% for 2012
42. Finally, Eurocontrol (2013) discusses different future scenarios and forecasts CAGR for instrument flight rules (IFR) flights between 0.8% and 2.6% in the Eurocontrol Statistical Reference Area. In a more recent study, Eurocontrol (2018) considers capacity constraints and forecasts an average annual growth of IFR flights between 0.5% and 2.7% between 2017 and 2040. In the most likely scenario, one of regulation and growth, they forecast an average annual growth rate of 1.9% for IFR flights. Eurocontrol arrives at the conclusion that there is a capacity gap of 1.5 million flights in their most likely scenario and even a capacity gap of 3.7 million flights in their global growth scenario, which is characterised by strong economic development. Because of the capacity shortage, between 8% and 16% of air passenger demand, that is, between 160 and 360 million people, cannot be served. If we compare the past development with the results of the aforementioned forecasts, we find a high degree of conformity. This essentially means Airport Capacity Constraints and Strategies for Mitigation. DOI: https://doi.org/10.1016/B978-0-12-812657-8.00008-7 © 2020 Elsevier Inc. All rights reserved.

189

190

PART | II Models for assessing mitigation strategies

that on a global scale the past growth of air traffic development is expected to continue in the future. However, we have to take a note of the fact that all these forecasts are to different degrees only demand forecasts. In other words, they are based on the implicit hypothesis of more or less sufficient airport capacity to serve the forecast demand. However, it is increasingly difficult to enlarge the capacity of an airport substantially, for example, by adding new runways, especially in economically well-developed countries. In such countries a time-consuming plan approval procedure is typically employed with the involvement of the public, which is more or less opposed to airport enlargements because of the increase of noise and air pollution. Therefore the purpose of this chapter is to present an econometric model to forecast runway expansion delays to take care of the important issue of potential airport capacity shortage in demand forecasting. The focus of the model is on runway capacity, because in the long run this is the most critical bottleneck in airport expansion plans and typically requires lengthy plan approval procedures in many countries. It is not rare for runway expansion plans at large airports, for example, Frankfurt, to take more than ten years until they are completed (Wissenschaftlicher Beirat beim Bundesminister fu¨r Verkehr, Bau und Stadtentwicklung, 2011). This chapter closes with a brief case study of analyses of the effects of limited airport capacity on future growth of flight volume worldwide.

8.1

Background

The starting point of the model is that if there is a current or future capacity gap at an airport, we need to analyse whether adding new runways is possible in time, with regard to the demand development, and, if realisation is not possible within an appropriate time, how long this process may be delayed or if capacity expansion is possible at all. This analysis is conducted for each airport and each new runway at an airport. The model that has been developed is based on the idea that there is a particular degree of opposition to airport expansion from the population surrounding the airport. This depends on factors such as noise annoyance, welfare level, economic opportunities, participation level and intermodal substitution. The degree of opposition may range from almost no opposition to such an intense opposition that airport expansion is virtually impossible. As a result, the estimated model enables us to estimate the probability of realisation of a new runway which can be transformed into an expected value of delay.

8.2

Model theory and parameter estimation

The approach used is a probabilistic one based on Markov chains (Markov, 2006) and discrete choice theory (McFadden, 1974). The Markov chain comprises two states that an airport can take (Fig. 8.1), which are as follows:

Modelling future airport capacity enlargements and limits Chapter | 8

191

Forecast demand0

5

0.

1

2

4

6

8–

–

10

Expected runway expansion delay in years

15.3%

–

15

e

or

M

an

20

th

FIGURE 8.5 Distribution of delayed runway expansions in the most likely scenario until 2040 (case study with a uniform CAGR of 3.5%). CAGR, compounded annual growth rate.

Adding a new runway is on average delayed by 13.9 years with a standard deviation of 25.9 years. The high value of the standard deviation illustrates the uneven distribution of delays. There are 23 runway expansions that are delayed by 15 years or even more, but about 68 runway expansions are delayed by not more than four years and, thus, are more or less negligible in a long-term forecast (Fig. 8.5).

199

Modelling future airport capacity enlargements and limits Chapter | 8

In an optimistic scenario, 12 more runways (5115 runways in total) are realised before 2040. The number of delayed runway expansions is 136, and thus 21 airport capacity enlargement projects are still in progress when reaching the forecast horizon. Average delay is 8.1 years with a standard deviation of 12.6 years. Limited airport capacity leads to a capacity gap of 11.69% of flights. As a result, the CAGR of air traffic, corrected for airport capacity shortage, is 2.97%. In a pessimistic scenario, only 95 new runways are finished before 2040. The number of delayed runway expansions decreases to 128, because the rather long delays tend to dampen the number of runway expansions that are realised or started but still in progress up to the forecast horizon of 2040. Average delay is 20.3 years with a standard deviation of 33.4 years. This leads to an overall capacity gap of 14.11% of aircraft movements compared to the unconstrained demand forecast, and the CAGR corrected for capacity constraints is 2.85%. Figs. 8.5
8.7 display the distribution of delayed runway expansions for the different scenarios in more detail. Between 45% and 55% of delayed runway expansions are in the category ‘less critical airports’ and capacity enhancements are delayed by up to four years. About 60
71 delays do not last more than four years and thus are of lesser importance for a long-term forecast. Most of these delays take place at small airports, for example, airports with a single runway that are upgraded to a two-runway system. On the other hand, between 12% and 25% of delayed runway expansions are in the

Number of delayed runway expansions

35 35.3%

12.5%

‘Important airports regarding constraints’

‘Heavily constrained airports’

52.2% 30

30

27

‘Less critical airports’

25

20

18

136 delayed runway expansions expected until 2040

15 12

12

10

9 7

9

7 5

5 5.1%

0

–

5 0.

>0

5.1%

22.1%

–1

–2

5

0.

1

19.9%

–4

2

8.8%

6.6%

–6

–8

10

4

13.2%

6

8–

Expected runway expansion delay in years

6.6%

3.7%

15

20

–

10

8.8%

–

15

e

or

M

an

20

th

FIGURE 8.6 Distribution of delayed runway expansions in the optimistic scenario until 2040 (case study with a uniform CAGR of 3.5%). CAGR, compounded annual growth rate.

200

PART | II Models for assessing mitigation strategies

45 46.9%

Number of delayed runway expansions

40

28.9%

39

‘Less critical airports’

35

24.2% ‘Heavily constrained airports’

‘Important airports regarding constraints’

30 26

25

128 delayed runway expansions expected until 2040

20

14

15

12

12

10

7

7 5

4

5

2 5.5%

0

–0

.5

>0

1.6%

0.

5

–1

9.4%

1

–2

30.5%

10.9%

9.4%

3.1%

5.5%

3.9%

4 2–

6 4–

8 6–

0 –1

5 –1

0 –2

8

Expected runway expansion delay in years

10

20.3%

15

e

or

M

an

20

th

FIGURE 8.7 Distribution of delayed runway expansions in the pessimistic scenario until 2040 (case study with a uniform CAGR of 3.5%). CAGR, compounded annual growth rate.

category ‘heavily constrained airports’ that comprises delays of 15 years and more. Of these, 12
26 delays are expected to last more than 20 years, and, as such, major runway enhancements are virtually impossible. This affects very large hubs such as London Heathrow, Chicago O’Hare, Frankfurt or Paris Charles de Gaulle (Table 8.4). Because of their high number of flights, they efficiently interconnect a large number of origin
destination pairs and therefore play an important role in the global flight network (e.g. Dennis, 1994; Veldhuis, 2013). However, these airports are also those that are prone to long-lasting capacity constraints. In this case study, with uniform traffic growth rates at all airports, the top 20 airports in terms of their share of unaccommodated flights in 2040 in the most likely scenario as a result of limited airport capacity account for about 13% of global flights in 2016 (Table 8.4). Most of these airports belong to the constraints category 3, in which adding significant new runway capacity is delayed by 15 years or more in the most likely scenario. Adding significant new runway capacity is especially difficult at airports which already have a high number of runways, such as ATL, ORD and DEN in the United States. As we have shown in Chapter 7, Modelling future airport capacity and capacity utilisation, capacity gains decrease with each new runway. In such cases, building a new airport seems to be a more effective option, such as the new airport Beijing Daxing (PKX) in China. Thus PEK is not among the airports mentioned in Table 8.4, because we expect PKX to handle the traffic that exceeds the capacity of PEK.

TABLE 8.4 Relative capacity gap of the top 20 airports (most likely scenario) worldwide in 2040 (case study with uniform CAGR). Number

Airport name

Country

IATA code

Share of unaccommodated flights Optimistic (%)

Most likely (%)

Pessimistic (%)

Constraints category (most likely scenario)

1

London Heathrow

United Kingdom

LHR

45.1

51.9

51.9

3

2

Atlanta Hartsfield
Jackson

United States

ATL

51.4

51.6

51.6

3

3

Chicago O’Hare

United States

ORD

51.5

51.5

51.5

3

4

Istanbul Sabiha Gokcen

Turkey

SAW

45.9

45.9

45.9

3

5

Jakarta Soekarno
Hatta International

Indonesia

CGK

45.6

45.6

45.6

3

6

Dublin

Ireland

DUB

41.3

41.3

41.3

3

7

Dallas/Fort Worth

United States

DFW

40.5

40.5

40.5

3

8

Fukuoka

Japan

FUK

0.0

36.7

36.7

3

9

Mumbai Chhatrapati Shivaji

India

BOM

36.3

36.3

36.3

3

10

New York LaGuardia

United States

LGA

22.0

34.6

51.1

3

11

Los Angeles

United States

LAX

32.3

34.2

35.7

3

12

Bengaluru Kempegowda

India

BLR

32.1

32.1

32.1

3

13

Delhi Indira Gandhi

India

DEL

31.6

31.6

31.6

3 (Continued )

TABLE 8.4 (Continued) Number

Airport name

Country

IATA code

Share of unaccommodated flights Optimistic (%)

Most likely (%)

Pessimistic (%)

Constraints category (most likely scenario)

14

Frankfurt

Germany

FRA

31.0

31.0

31.0

3

15

Hong Kong

Hong Kong

HKG

14.4

29.3

29.3

3

16

Juanda

Indonesia

SUB

29.3

29.3

29.3

3

17

Paris Charles de Gaulle

France

CDG

29.3

29.3

29.3

3

18

Munich

Germany

MUC

16.7

29.2

29.2

2

19

Tokyo Haneda

Japan

HND

28.7

28.7

28.7

3

20

Denver

United States

DEN

27.5

27.5

27.5

3

CAGR, compounded annual growth rate.

Modelling future airport capacity enlargements and limits Chapter | 8

203

At some airports, there is a big difference in the share of unaccommodated flights in 2040 between the three scenarios. For example, airports 3
7 and 12
14 show no variation, while airports 8, 10, 15 and 18 show a high degree of variation. Factors that influence the variation comprise the following: G

G

G

The region an airport belongs to in the model affects variation between scenarios. Airports 2, 5, 9, 12, 13 and 16 belong to region R3, while airports 8 and 10 belong to region R2, which has a higher standard error (Table 8.2). Very high or very low levels of airport capacity expansion delays may lead to the same results in terms of new runways irrespective of the scenario chosen (e.g. airports 9, 12 and 14), while intermediate levels of expansion delays may result in different outcomes (airports 8, 15 and 18). Very high levels of airport expansion delays tend to occur more often at large airports which are surrounded by a large population, while very low levels tend to occur more often at rather small airports where the surrounding population is limited. Furthermore, the expected date of airport expansion in relation to the forecast horizon may play a role. If the forecast horizon is 2040 and airport expansion is expected in the most likely scenario in 2038, a few years can make a difference in terms of new runways realised. This is the case for London Heathrow, where the number of new runways depends on the scenario chosen (between zero and one new runway up to 2040, however, very sensitive to the chosen forecast horizon).

Finally, we want to note again that the case study served primarily to prove that model results vary in a consistent way in relation to assumptions on model inputs and seem to be plausible. The results, however, cannot be interpreted as outcomes of a probable forecast, because of the simplifying assumption of uniform growth rates. The actual forecast will be presented in Part III of this book.

8.4

Conclusion

In this chapter, we have introduced a model which incorporates limited airport capacity in air transport forecasts. The key idea of the model is that delayed runway expansions are a result of opposition due to negative effects of such plans on the airport’s neighbourhood. Factors of the model are noise annoyance, level of welfare, economic opportunities, intermodal substitution possibilities and level of participation. A number of factors are modelled by auxiliary variables because of measurement problems. Because of the naturally high complexity and uncertainty that is typical for the task of forecasting capacity enhancement delays in the long term, we have included forecast confidence intervals to allow for different scenarios.

204

PART | II Models for assessing mitigation strategies

Under the assumption that future traffic growth for the next 24 years uniformly follows the same trend as in the past 25 years, there is a great need for additional airport capacity, especially at hub airports. However, expanding airport capacity is typically a problem for large airports which are close to large agglomerations such as London Heathrow, Chicago O’Hare, Frankfurt and Paris Charles de Gaulle. Furthermore, these airports are often major hubs in the global flight network and capacity constraints have far-reaching effects. On the other hand, limited airport capacity plays only a minor role at small airports or where large agglomerations are not nearby. Nevertheless, an important mitigation measure to overcome airport capacity constraints, at least partially, is to employ larger aircraft size. While aircraft size mainly depends on the business decisions of individual airlines, we can observe a trend towards larger aircraft at congested hubs. However, there are many factors that influence the deployment of aircraft by airlines, but airport congestion plays a role as has been shown in Chapter 5, General strategies for mitigating airport capacity constraints, and as we will see in the next chapter, too.

References Airbus, 2016. Global Market Forecast 2016
2035. Airbus, Blagnac. Ben-Akiva, M., Lerman, S.R., 1985. Discrete Choice Analysis. MIT Press, Cambridge, MA. Bhattacharyya, G.K., Johnson, R.A., 1977. Statistical Concepts and Methods. John Wiley & Sons, New York. Boeing, 2016. Current Market Outlook 2016
2035. Boeing, Seattle, WA. Bright, E.A., Coleman, P.R., King, A.L., Rose, A.N., 2008. LandScan 2007. Oak Ridge National Laboratory, Oak Ridge, TN. Dennis, N., 1994. Airline hub operations in Europe. Journal of Transport Geography 2 (4), 219
233. Domencich, T.A., McFadden, D., 1975. Urban Travel Demand
A Behavioral Analysis. Elsevier, New York. Eurocontrol, 2013. Challenges of Growth 2013
Task 4: European Air Traffic in 2035. Brussels. Eurocontrol, 2018. European Aviation in 2040
Challenges of Growth. Brussels, Belgium. International Civil Aviation Organization (ICAO), 2016a. Long-Term Traffic Forecasts
Passenger and Cargo. Montreal. International Civil Aviation Organization (ICAO), 2016b. ICAO Traffic Statistics. Montreal. Markov, A.A., 2006. An example of statistical investigation of the text Eugene Onegin concerning the connection of samples in chains. Sci. Context 19, 591
600. McFadden, D., 1974. Conditional logit analysis of qualitative choice behavior. In: Zarembka, P. (Ed.), Frontiers in Econometrics. Academic Press, New York. Official Airline Guide (OAG), 2007. Market Analysis. Reed Travel Group, Dunstable. The World Bank, 2007. World Development Indicators. Washington, DC. Veldhuis, J., 2013. Improving connectivity at hub airports. Journal of Airport Management 7 (2), 136
151. Wissenschaftlicher Beirat beim Bundesminister fu¨r Verkehr, Bau und Stadtentwicklung, 2011. Handlungsbedarf fu¨r Planung und Nutzung der Flughafeninfrastruktur in Deutschland. Zeitschrift fu¨r Verkehrswissenschaft 82 (2), 91
121.

Chapter 9

Modelling future development of the average number of passengers per flight 9.1

Background

The average aircraft size model is the link which relates flight capacity with the passenger capacity of an airport (Fig. 9.1). The main limiting factor of airport capacity is, in many instances, the capacity of the runway system, which is typically measured in flights per time period, for example one or several hours. However, eventually we want to know how the original passenger demand forecast (Chapter 6: Modelling future air passenger demand) is affected by limited airport capacity. Therefore we need to know how many passengers are transported on average per flight. Berster et al. (2015) analysed the question of whether aircraft size increases with airport congestion, that means high utilisation rates of airport capacity. Airport congestion has a positive effect on the development of aircraft size, but average flight length and regional factors also have a significant effect on average aircraft size at a particular airport. For example, everything else being equal, average aircraft size tends to be lower in North America compared to the rest of the world because of the higher share of high-frequency domestic flights. Furthermore, not only the congested airports but also the non-congested ones are affected by increasing aircraft size,

Airport capacity (flights)

Airport capacity (passengers)

Average aircraft size model (passengers per flight)

FIGURE 9.1 Transforming airport capacity from flight into passenger volume by the average aircraft size model. Airport Capacity Constraints and Strategies for Mitigation. DOI: https://doi.org/10.1016/B978-0-12-812657-8.00009-9 © 2020 Elsevier Inc. All rights reserved.

205

206

PART | II Models for assessing mitigation strategies

because of the hub-and-spoke structure of large parts of the air transport network. Therefore compared to the Berster et al. (2015) study, we have developed a more disaggregated approach for the case studies in Part III of this book. We employ a combination of data envelopment analysis (DEA) and regression analysis, the same as the method which we used in Chapter 7, Modelling future airport capacity and capacity utilisation, of this book for airport capacities. Key concepts of this approach are ‘passenger capacity potential’ and ‘passenger capacity potential utilisation’. These terms describe the maximum passenger volume that can realistically be transported on a particular route and its current utilisation rate. The current passenger capacity potential is calculated by a DEA and depends on the capacity utilisation of the two airports of a particular route, the number of flight movements and the flight length of the route. From this analysis, we can then calculate the current number of passengers per aircraft and the future number of passengers per aircraft given a particular utilisation rate of the passenger capacity potential.

9.2

Model theory and parameter estimation

The chosen model approach for forecasting future aircraft size, that is passengers per flight, is based on a DEA for estimating the current passenger capacity potential and its utilisation of a particular route. Based on this analysis, we performed linear and logistic regression analyses to estimate the future development of the passenger capacity potential and its utilisation (Fig. 9.2). The approach is very similar to the method of estimating airport Data envelopment analysis (DEA) of sample routes (Table 9.1)

Generalisation of findings from DEA

Logistic regression analysis on sample routes

Linear regression analysis on sample routes

(Table 9.2)

(Table 9.3)

Capacity potential utilisation development

Capacity potential development

Development of future number of passengers per flight (aircraft size)

FIGURE 9.2 Model for estimating future aircraft size (number of passengers per flight).

Modelling future development of the average number Chapter | 9

207

capacity in Chapter 7, Modelling future airport capacity and capacity utilisation; therefore the reader is referred to this chapter for a more detailed discussion of the DEA method. The DEA has been employed for routes between the largest 200 airports that have been the subject of the capacity analysis in Chapter 7, Modelling future airport capacity and capacity utilisation. Data is taken from Sabre AirVision Market Intelligence (MI) (2016). For illustration, Table 9.1 lists the top 50 routes in terms of passengers per aircraft. Output of the DEA is the passenger capacity potential utilisation (PCU, column 12) and, thus, the annual passenger capacity potential (PC, column 13), which can be calculated by means of the results of columns 8 (annual passenger volume, APV) and 12. As inputs for the DEA we have employed the following: G

G

G

The maximum airport capacity utilisation, which is the maximum value of annual capacity utilisation of the departure airport (CU DA, column 4) and the arrival airport (CU AA, column 7). In this case the airport with the more serious capacity constraints is assumed to be the major bottleneck in flight scheduling. However, a weighted average of both capacity utilisation values did not produce significantly different results. The annual flight volume on a particular route (AFV, column 9), to segregate between different types of flights, for example high-frequency domestic flights versus lower frequency long-distance flights, and different regions of the world, for example the United States versus Asia. The flight distance in km [Dist. (km), column 10] to account for shorthaul versus long-haul flights.

As a result of the DEA, there are four reference routes that were operating at their capacity potential limit and have more than 400 passengers per flight: G G G G

Cancun (CUN) to London Gatwick (LGW) Dubai (DXB) to New York JFK (JFK) London Gatwick (LGW) to Antalya (AYT) London Gatwick (LGW) to Dubai (DXB)

Not surprisingly, at least one airport of each of these four reference routes is at, or close to, its annual capacity limit. According to our capacity analysis in Chapter 7, Modelling future airport capacity and capacity utilisation, LGW has 100% capacity utilisation and DXB has 90% capacity utilisation. Furthermore, flight distances of these four routes tend to be rather long. LGW to AYT has a flight distance of about 2900 km and is the shortest of the routes, while the other three routes have flight distances of about 5500 km (LGW to DXB), almost 8000 km (CUN to LGW) and 11,000 km (DXB to JFK). Both factors
a high degree of airport capacity utilisation and a long flight distance
tend to produce high values of passengers per

TABLE 9.1 The top 50 routes in terms of passengers per aircraft in 2016. 1

2

3

4

5

6

7

8

9

10

11

12

13

14

No.

IATA code

DA

CU DA

IATA code

AA

CU AA

APV

AFV

Dist. (km)

P/F

PCU

PC

PC/F

1

CUN

Cancun

28

LGW

London Gatwick

100

216,665

421

7966

515

100

216,665

515

2

DXB

Dubai

90

JFK

New York JFK

60

563,836

1098

11,000

514

100

563,836

514

3

LGW

London Gatwick

100

CUN

Cancun

28

213,399

422

7966

506

98

217,177

515

4

YUL

Montreal

34

ORY

Paris Orly

42

46,650

93

5530

502

97

47,862

515

5

MAN

Manchester

43

AYT

Antalya

15

170,530

340

3129

502

98

173,379

510

6

DXB

Dubai

90

MEL

Melbourne

56

395,287

791

11,655

500

97

406,371

514

7

LGW

London Gatwick

100

AYT

Antalya

15

233,447

470

2906

497

100

233,447

497

8

MEL

Melbourne

56

DXB

Dubai

90

391,036

789

11,655

496

96

405,345

514

9

JFK

New York JFK

60

DXB

Dubai

90

533,025

1098

11,000

485

95

563,836

514

10

DXB

Dubai

90

LAX

Los Angeles

87

262,886

545

13,401

482

94

280,194

514

11

AYT

Antalya

15

LGW

London Gatwick

100

225,963

471

2906

480

97

233,844

496

12

LGW

London Gatwick

100

DXB

Dubai

90

536,608

1129

5472

475

100

536,608

475

13

LAX

Los Angeles

87

DXB

Dubai

90

257,030

545

13,401

472

92

280,194

514

14

GLA

Glasgow

34

AYT

Antalya

15

67,257

146

3357

461

90

75,131

515

15

MNL

Manila

59

JED

Jeddah

35

97,494

215

8594

453

88

110,648

515

16

AYT

Antalya

15

GLA

Glasgow

17 18

34

66,180

149

3357

444

86

76,669

515

AYT

Antalya

15

HEL

Helsinki

28

50,340

115

2618

438

85

59,184

515

DXB

Dubai

90

CPH

Copenhagen

45

168,498

387

4818

435

85

198,849

514

19

CDG

Paris CDG

62

AYT

Antalya

15

12,607

29

2641

435

84

14,925

515

20

FRA

Frankfurt

63

LAX

Los Angeles

87

217,387

504

9323

431

84

259,140

514

21

DXB

Dubai

90

SFO

San Francisco

57

157,669

366

13,021

431

84

188,360

515

22

BNE

Brisbane

46

DXB

Dubai

90

219,153

511

11,977

429

83

262,755

514

23

MIA

Miami

47

FRA

Frankfurt

63

156,822

366

7763

428

83

188,189

514

24

DXB

Dubai

90

BNE

Brisbane

46

218,754

511

11,977

428

83

262,755

514

25

AYT

Antalya

15

MAN

Manchester

43

144,427

340

3129

425

83

173,379

510

26

CPH

Copenhagen

45

DXB

Dubai

90

168,053

396

4818

424

83

203,459

514

27

DXB

Dubai

90

LGW

London Gatwick

100

475,858

1129

5472

421

89

536,608

475

28

HEL

Helsinki

28

AYT

Antalya

15

48,110

115

2618

418

81

59,184

515

29

LAX

Los Angeles

87

FRA

Frankfurt

63

210,132

504

9323

417

81

259,140

514

30

DXB

Dubai

90

SYD

Sydney

57

456,588

1096

12,046

417

81

562,810

514

31

SFO

San Francisco

57

DXB

Dubai

90

152,280

366

13,021

416

81

188,360

515

32

FRA

Frankfurt

63

MIA

Miami

47

151,864

366

7763

415

81

188,189

514

33

HEL

Helsinki

28

LPA

Gran Canaria

23

106,719

258

4688

414

81

132,532

514

34

BCN

Barcelona

52

DXB

Dubai

90

301,304

732

5174

412

83

362,433

495

35

DXB

Dubai

90

MAN

Manchester

43

450,567

1098

5654

410

87

516,174

470

(Continued )

TABLE 9.1 (Continued) 1

2

3

4

5

6

7

8

9

10

11

12

13

14

No.

IATA code

DA

CU DA

IATA code

AA

CU AA

APV

AFV

Dist. (km)

P/F

PCU

PC

PC/F

36

DUS

Du¨sseldorf

50

DXB

Dubai

90

299,262

732

5005

409

83

361,165

493

37

DOH

Doha

44

MNL

Manila

59

350,528

862

7287

407

85

414,107

480

38

FCO

Rome

55

DXB

Dubai

90

408,050

1006

4342

406

88

461,236

458

39

MAN

Manchester

43

DXB

Dubai

90

445,295

1098

5654

406

86

516,174

470

40

YYZ

Toronto

48

DXB

Dubai

90

126,564

314

11,080

403

78

161,598

515

41

DXB

Dubai

90

DUS

Du¨sseldorf

50

294,784

732

5005

403

82

361,165

493

42

SYD

Sydney

57

DXB

Dubai

90

440,661

1097

12,046

402

78

563,323

514

43

CUN

Cancun

28

SCL

Santiago

23

26,402

66

6312

400

78

33,966

515

44

YYZ

Toronto

48

JED

Jeddah

35

62,369

156

10,462

400

78

80,284

515

45

SYD

Sydney

57

DOH

Doha

44

121,644

305

12,377

399

78

156,854

514

46

AMS

Amsterdam

52

AYT

Antalya

15

149,755

376

2654

398

80

186,659

496

47

CDG

Paris CDG

62

DXB

Dubai

90

560,364

1407

5,237

398

91

612,895

436

48

MUC

Munich

71

DXB

Dubai

90

461,565

1166

4562

396

89

519,688

446

49

CUN

Cancun

28

MAD

Madrid

50

150,614

381

7950

395

77

195,784

514

50

DXB

Dubai

90

YYZ

Toronto

48

124,110

314

11,080

395

77

161,598

515

60

243,328

558

7444

437

87

276,482

503.7

Ø

59

AA, arrival airport name; AFV, annual flight volume; APV, annual passenger volume; CU, capacity utilisation in %; DA, departure airport name; Dist., distance; PC, estimated annual passenger capacity potential; PCU, passenger capacity potential utilisation in %; PC/F, estimated passenger capacity potential per flight; P/F, passengers per flight.

Modelling future development of the average number Chapter | 9

211

aircraft. As a result, average aircraft size, that is passengers per flight, is in a rather narrow range of 395
515. Finally, Table 9.1 displays the results of the DEA in columns 12
14: G

G

G

Column 12 displays the passenger capacity potential utilisation (PCU), which is 100% for the four reference routes. Based upon the passenger capacity potential utilisation, we can calculate the estimated annual passenger capacity potential (PC, column 13), which is the ratio of the current annual passenger volume and the value of the passenger capacity potential utilisation. For the four reference routes, current annual passenger volume equals estimated annual passenger capacity potential, since the passenger capacity potential utilisation has a value of 100%. Eventually, we can calculate the estimated passenger capacity potential per flight (PC/F, column 14), which is the ratio of the estimated annual passenger capacity potential and the current number of flights. This is the value we are looking for to forecast the development of average aircraft size per route; however, in many cases, this is a long-term value which may not be reached until the forecast horizon. Furthermore, this value may change as the route structure evolves, for example as a result of demand growth and the development of airport capacities. Thus, this is a dynamic system which needs some further modelling as illustrated by Fig. 9.2.

However, before we proceed, we take a closer look at the last row of Table 9.1. Average annual airport capacity utilisation of those airports is around 60%, while average maximum airport capacity utilisation is 75% (to be calculated from columns 4 and 7) and thus significantly higher, because the maximum values of the departure and arrival airports are retained for expressing maximum airport capacity utilisation. Nevertheless, the 50 largest routes in terms of passengers per flight are linking airports which have a rather high degree of capacity utilisation. Average number of passengers is more than 240,000 per year. Furthermore, average flight length is more than 7400 km, and flight volume is more than 550 flights per year. In many cases, these are long-distance routes which are served on average by ten flights per week. Average passenger capacity potential utilisation is high and takes a value of 87%. As a result the current value of the average number of passengers per flight of these 50 routes is 437. The capacity analyses of flight routes as conducted in this chapter is not as clear-cut as in the case of single airports described in Chapter 7, Modelling future airport capacity and capacity utilisation. The main reason for this being that two airports define a nonstop flight route. These two airports may be very different in terms of their capacity utilisation, as a large part of the global flight volume is still handled by the hub-and-spoke system. Here, the hub airports typically have a much higher degree of capacity utilisation compared to their spoke counterparts. In

212

PART | II Models for assessing mitigation strategies

Berster et al. (2015), increasing aircraft size was not only limited to airports which were operating at or close to their capacity limit but was found at non-congested airports as well. After calculating the current passenger capacity potential and its utilisation, we need to forecast their development until the forecast horizon. Various intermediate forecast steps may be included as illustrated by Fig. 9.3. Furthermore, depending on the forecast horizon, the passenger demand, and airport capacity forecast, the passenger capacity potential utilisation of most routes will be well below 100% as indicated by the example of Fig. 9.3. Therefore we need a model capable of modelling passenger capacity potential utilisation and its development. For this task we have chosen a logistic regression of the form yj 5

2

11e

1 P i

ð9:1Þ

β i 3 xij

here yj is the dependent variable and represents the passenger capacity potential utilisation of route j. Hence, xij is the value of attribute i for routes j and β i are the coefficients for attribute i. We have identified the natural logarithm of the annual passenger volume and the natural logarithm of the flight distance as significant explanatory variables (Table 9.2). As expected the utilisation of the passenger capacity potential tends to increase with passenger volume and the flight distance. This is an effect which can be observed from Table 9.1; however, the rather low McFadden

Passenger capacity potential utilisation

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

0

t

t+1 Time

…

T

FIGURE 9.3 Illustration of the development of the passenger capacity potential from time 0 until the forecast horizon T.

Modelling future development of the average number Chapter | 9

213

pseudo-R2 indicates that route-specific effects, which are not included fully in the model, play an important role as well. To adequately capture these effects, we employed the model for applications in incremental mode (see Chapter 7: Modelling future airport capacity and capacity utilisation); thus the current utilisation is used as a starting value, and the future utilisation develops according to the model elasticity of the number of passengers. The flight distance is assumed to be fixed for a given route. As passenger volume, flight volume or the airport capacity utilisation of a flight route vary, the passenger capacity potential changes as well. To capture this effect, we have developed a linear regression model so that these developments can be accounted for. Table 9.3 displays the estimation results. The dependent variable is the passenger capacity potential of a particular route, while the annual flight volume, the flight distance and the maximum airport capacity utilisation are the explanatory variables. The passenger capacity potential increases with the number of flights. Furthermore, increasing airport capacity utilisation of the more constrained airport typically leads to increasing aircraft size and, thus, a larger passenger capacity potential. Finally, increasing flight distance reduces passenger capacity potential slightly by about 1167 passengers per additional 1000 km

TABLE 9.2 Estimation results of the passenger capacity potential utilisation model. No.

Variable

Coefficient

Standard error

p-value

1

Constant

2 7.19996

0.34811

0.00000

2

ln(annual passenger volume)

0.38174

0.01989

0.00000

0.29702

0.02498

0.00000

3

ln(flight distance) McFadden pseudo-R2

12.62%

TABLE 9.3 Estimation results of the passenger capacity potential model. Standard error

p-value

41,913.00221

3420.55982

0.00000

Annual flight volume

232.28138

0.49830

0.00000

3

Flight distance

2 1.16669

0.36334

0.00133

4

Airport capacity utilisation

143,685.89611

5716.18065

0.00000

No.

Variable

1

Constant

2

R2

Coefficient

96.85%

214

PART | II Models for assessing mitigation strategies

of flight distance, as passenger demand typically decreases with increasing distance (see, e.g., Table 6.4). However, since flight distance is assumed to be fixed for a given route, this is of less interest. R2 is high, but for increased forecast accuracy, the model is still applied in incremental mode, so that the current passenger capacity potential serves as a starting value and then develops according to the model elasticities of annual passenger volume and the maximum airport capacity utilisation. Fig. 9.4 summarises the various steps of applying the aircraft size model. First, a DEA is performed for all airports under study to obtain current values of passenger capacity potential and its utilisation for each airport. In the next step, these values are updated on the basis of the passenger demand forecast by means of the passenger capacity potential utilisation model (Table 9.2) and the passenger capacity potential model (Table 9.3). Typically, both models need to be run multiple times to reach stable values of flight volume and maximum airport capacity utilisation. Finally, we can calculate the future number of passengers per flight. Here, we take a weighted average of the projected number of passengers per flight and the most recent DEA results, typically the base year of the forecast. The weights are calculated on the basis of the passenger volume for the base and the forecast year for each route. Thereby, we aim to smooth large fluctuations on particular routes where we find large increases or decreases of passenger volumes. While this approach does not significantly affect overall model fit in a positive or negative way, it produces more stable results with less volatility. Perform data envelopment analysis (DEA) for all routes under study

Current capacity potential and capacity potential utilisation

Update capacity potential based on passenger

Update capacity potential utilisation based on

forecast

passenger forecast

Forecast capacity potential

Forecast capacity potential utilisation

Calculate future number of passengers per flight (aircraft size)

FIGURE 9.4 Application of the aircraft size model (number of passengers per flight).

Modelling future development of the average number Chapter | 9

215

9.3 Model application: 2008
16 projection of aircraft size (passengers per flight) on 40 sample routes The purpose of this brief case study is to analyse a projection of aircraft size development of those 40 sample routes of Table 9.1, which were served in 2008 by non-stop flights. However, before moving on to aircraft size development at the route level, we want to look at aircraft size development at the airport level for four sample airports. Fig. 9.5 illustrates these developments from 2008 to 2016 for the following airports: G G G G

London Heathrow (LHR) New York JFK (JFK) Beijing (PEK) Frankfurt (FRA)

At LHR the average number of passengers per flight for all arriving or departing flights increased from 134 to 160, that is by about 19% between 2002 and 2016. As we have seen in Part I, LHR has been severely capacity constrained for some years, so that additional passenger demand can only be served by employing larger aircraft. While not reaching the same capacity utilisation levels as LHR, but still highly utilised, the same applies for JFK and FRA. Passengers per flight increased from 111 for JFK and 107 for FRA to 135 and 137, that is by 21% and 29%, respectively, which is even more than at LHR in relative terms. However, we have to consider that a fourth 180 160

Passengers per flight

140 120 100 80 60 40 20

0

2002

2008 LHR

JFK

PEK

2016 FRA

FIGURE 9.5 Development of the number of passengers per flight at London Heathrow (LHR), New York JFK (JFK), Beijing (PEK), and Frankfurt (FRA) airports between 2002 and 2016 [Sabre AirVision Market Intelligence (MI), 2016].

216

PART | II Models for assessing mitigation strategies

runway was added to FRA in late 2011, which relieved some pressure from the capacity situation. The development of average aircraft size at PEK is different from the other three airports. Overall, passengers per flight increased between 2002 and 2016 from 145 to 159, that is by ‘only’ 10%, but actually decreased between 2002 and 2008 from 145 to 126, hence, by 13%. Late in 2007, a third runway was opened, so that three independent runways have been in operation since then. From our calculations in Chapter 7, Modelling future airport capacity and capacity utilisation, we estimate the capacity increase of PEK to be 50%, compared to an estimated capacity increase at FRA of 19% (Table 7.4). While we see a more or less clear positive relationship between average aircraft size and capacity utilisation at the four airports, there are also factors which are specific to each airport. These factors are difficult to collect and analyse in their entirety on a global level for each airport, and for such analyses to be meaningful, they have to be conducted on the route level. Hence, we have to analyse non-stop connections between any two airports. Therefore we have chosen to employ an incremental DEA approach as described in this chapter as well as in Chapter 7, Modelling future airport capacity and capacity utilisation, to implicitly include as much information about the individual non-stop connections as possible. Nevertheless, forecasting the development of average aircraft size, that is the number of passengers per flight, for each non-stop connection worldwide remains an ambitious task; thus we call our approach a projection rather than a forecast. Table 9.4 displays the results of the aircraft size projection 2008
16 for the 40 non-stop connections of Table 9.1 that were served in 2008 by at least one flight per year. Most connections in Table 9.4 were served in 2008 at a higher flight frequency than the minimum condition of one flight per year. Most routes were served at least weekly. Columns 2 and 3 describe the flight route, that is the departure and arrival airports. Columns 4
6 display the route statistics for the year 2016, that is the annual passenger and flight volume and the number of passengers per flight. Columns 7
9 show the same statistics for the year 2008. Thus columns 4
9 represent actual, not forecast, values. Columns 10
12 show projected values for the year 2016 and the projection error, both in absolute and relative terms. On average, annual passenger volume increased by 164% between 2008 and 2016 on these 40 routes. This equals a compound annual growth rate (CAGR) of 12.9% per year, which is quite impressive. To accomplish this, the flight volume increased by 39% (CAGR of 4.2% per year) and the number of passengers per flight rose by 90% (CAGR of 8.4% per year). Thus, a large part of the demand growth was handled by increasing aircraft size. However, we have to consider that the majority of these routes are highgrowth routes with respect to demand development.

TABLE 9.4 Projection of average aircraft size (number of passengers per flight) on 40 sample routes for the time period 2008
16. 1

2

3

4

5

6

7

8

9

10

11

12

No.

IATA code DA

IATA code AA

APV 2016

AFV 2016

P/F 2016

APV 2008

AFV 2008

P/F 2008

pP/F 2016

PE

PE (%)

1

CUN

LGW

216,665

421

515

56,445

100

564

572

57

11.1

2

DXB

JFK

563,836

1098

514

225,179

734

307

513

21

2 0.2

3

LGW

CUN

213,399

422

506

63,654

100

637

571

66

13.0

4

YUL

ORY

46,650

93

502

35,990

110

327

432

2 70

2 13.9

5

MAN

AYT

170,530

340

502

29,028

104

279

521

20

3.9

6

DXB

MEL

395,287

791

500

82,878

366

226

508

8

1.6

7

LGW

AYT

233,447

470

497

52,028

195

267

503

6

1.2

8

MEL

DXB

391,036

789

496

85,067

366

232

506

11

2.1

9

JFK

DXB

533,025

1,098

485

207,995

803

259

464

2 21

2 4.4

10

DXB

LAX

262,886

545

482

6011

29

207

492

10

2.0

11

AYT

LGW

225,963

471

480

49,123

195

252

484

5

1.0

12

LGW

DXB

536,608

1129

475

222,567

1098

203

410

2 66

2 13.8

13

LAX

DXB

257,030

545

472

6392

29

220

483

12

2.5

14

DXB

CPH

168,498

387

435

12,749

62

206

461

26

5.9

15

CDG

AYT

12,607

29

435

253

15

17

427

27

2 1.7

(Continued )

TABLE 9.4 (Continued) 1

2

3

4

5

6

7

8

9

10

11

12

No.

IATA code DA

IATA code AA

APV 2016

AFV 2016

P/F 2016

APV 2008

AFV 2008

P/F 2008

pP/F 2016

PE

PE (%)

16

FRA

LAX

217,387

504

431

253,748

989

257

297

2 134

2 31.2

17

DXB

SFO

157,669

366

431

1463

7

209

438

7

1.7

18

MIA

FRA

156,822

366

428

113,430

364

312

376

2 53

2 12.3

19

AYT

MAN

144,427

340

425

34,723

113

307

465

40

9.5

20

CPH

DXB

168,053

396

424

11,567

62

187

446

21

5.0

21

DXB

LGW

475,858

1129

421

215,967

1098

197

365

2 56

2 13.4

22

LAX

FRA

210,132

504

417

249,475

988

253

287

2 129

2 31.0

23

DXB

SYD

456,588

1096

417

84,513

366

231

458

41

9.9

24

SFO

DXB

152,280

366

416

1508

7

215

424

8

1.9

25

FRA

MIA

151,864

366

415

108,185

364

297

363

2 52

2 12.6

26

HEL

LPA

106,719

258

414

30,122

182

166

378

2 36

2 8.7

27

DXB

MAN

450,567

1098

410

230,509

1377

167

334

2 76

2 18.6

28

DUS

DXB

299,262

732

409

130,109

732

178

361

2 48

2 11.7

29

DOH

MNL

350,528

862

407

141,780

539

263

421

15

3.6

30

FCO

DXB

408,050

1006

406

125,826

520

242

427

22

5.3

31

MAN

DXB

445,295

1098

406

227,774

732

311

447

41

10.2

32

YYZ

DXB

126,564

314

403

46,761

157

298

460

57

14.1

33

DXB

DUS

294,784

732

403

132,720

770

172

350

2 52

2 13.0

34

SYD

DXB

440,661

1097

402

83,878

366

229

445

43

10.8

35

CUN

SCL

26,402

66

400

6057

14

433

430

30

7.5

36

AMS

AYT

149,755

376

398

86,695

612

142

300

2 98

2 24.6

37

CDG

DXB

560,364

1407

398

302,647

1666

182

325

2 74

2 18.5

38

MUC

DXB

461,565

1166

396

168,380

1069

158

346

2 50

2 12.6

39

CUN

MAD

150,614

381

395

154,336

414

373

361

2 34

2 8.7

40

DXB

YYZ

124,110

314

395

50,453

157

321

468

73

18.5

272,845

624

437

103,200

449

230

428

2 11

2 2.6

Ø

AA, arrival airport name; AFV, annual flight volume; APV, annual passenger volume; DA, departure airport name; PE, projection error; P/F, passengers per flight; pP/F, projected passengers per flight.

220

PART | II Models for assessing mitigation strategies

The average projection error of the 40 routes of Table 9.4 is 211 passengers per flight (22.6%) with a standard deviation of 51 passengers per flight (12.0%). While the forecast is quite accurate for aggregates such as airports, it is not as accurate for individual routes. This is notably the case for routes that have a very dynamic demand development. To illustrate this point, we take a closer look at three airports, which are part of five or more routes in Table 9.4 (to produce more or less meaningful averages), either as departure or arrival airport: G G G

Antalya (AYT), which has an average projection error of 210.8% Dubai (DXB), which has an average projection error of 214.6% London Gatwick (LGW), which has an average projection error of only 20.9%

AYT is part of six routes, and the average projection error is mainly driven by the AMS
AYT route. On this route, average aircraft size increased from 142 to 376 passengers per flight and annual passenger demand volume increased from about 87,000 to nearly 150,000 passengers. This is an increase of 173% between 2008 and 2016. The projected aircraft size is 300 passengers per flight, that is an error of 276 passengers per flight (224.6%). DXB is part of 23 routes and the average projection error is mainly driven by the CDG
DXB and DXB
MAN routes. On these routes, average aircraft size increased from 182 to 398 and 167 to 410 passengers per flight, and annual passenger demand volume increased from about 303,000 to over 560,000 and around 231,000 to almost 451,000 passengers, that is by 185% and 195% between 2008 and 2016, respectively. The projected aircraft size is 325 and 334 passengers per flight, that is an error of 274 and 276 passengers per flight (218.5% and 218.6%). LGW is part of the six routes, and the average projection error is mainly driven by the LGW
DXB and DXB
LGW routes. On these routes, average aircraft size increased from 203 to 475 and 197 to 421 passengers per flight and annual passenger demand volume increased from almost 223,000 to nearly 537,000 and around 216,000 to almost 476,000 passengers, that is by 141% and 120% between 2008 and 2016. The projected aircraft size is 410 and 365 passengers per flight, that is an error of 266 and 256 passengers per flight (213.8% and 213.9%). Ultimately, we are interested in average aircraft size per airport to assess the capacity situation at an airport and potential mitigation measures. However, average aircraft size per airport is composed of the average aircraft size of many individual routes, so that individual projection errors are likely to be balanced to a large extent, as illustrated by the average values of Table 9.4. Furthermore, we have to take into account that the routes of Table 9.4 are mainly of exceptionally high demand growth, and that the projection error increases with more dynamic demand development, as the three

Modelling future development of the average number Chapter | 9

221

examples of Antalya, Dubai, and London Gatwick airports illustrate. However, these types of routes typically play only a minor role in a global air traffic demand assessment.

9.4

Conclusion

The purpose of this chapter is to present a simple to apply and robust approach to forecast (or rather project) the development of the key forecast variable passengers per flight (aircraft size) by airport pair to extend the forecast of airport capacities as described in Chapter 7, Modelling future airport capacity and capacity utilisation, to available passenger volume capacities. As in the case of airport capacities explained in Chapter 7, Modelling future airport capacity and capacity utilisation, the approach again is highly problem-specific and cannot be a substitute for a detailed flight route analysis in terms of aircraft fleet characteristics and their future development. However, the projected aircraft size development per airport pair can serve as a higher level parameter for top-down fleet modelling. The model approach chosen is very similar to the modelling of airport capacities in Chapter 7, Modelling future airport capacity and capacity utilisation. The basic analysis is a DEA, which is further refined by regression analyses to generalise the results, so that they can be used for forecasting and projection, respectively. The results are satisfactory on the aggregate airport level, that is average aircraft size by airport. This key characteristic is important for the capacity assessment in the next part of the book. However, forecast or projection accuracy varies considerably between different routes. Aircraft size tends to be underestimated, especially on routes which have a very dynamic passenger demand development, but these are typically a minority compared to the global passenger demand development. On the other hand, the projections are quite good for routes that have a near-average demand development. These types of routes form the majority. Nevertheless, the model presented cannot be a substitute for a more detailed approach if accuracy on single flight routes is essential. While the approach presented is based on route analyses, the main goal is to produce more or less accurate projections at the airport level.

References Berster, P., Gelhausen, M.C., Wilken, D., 2015. Is increasing aircraft size common practice at congested airports? J. Air Transp. Manage. 46, 40
48. Sabre AirVision Market Intelligence (MI), 2016. Data Based on Market Information Data Tapes (MIDT). Sabre, Southlake.

Part III

Forecasting future air traffic development up to 2040 and assessing mitigation strategies The current part of the book deals with the air traffic forecasts for the years 2030 and 2040. Here, we focus on the future development of air passenger and flight volume, and the number of passengers per aircraft (aircraft size) for each world region. Furthermore, we present airportspecific results for the top 20 airports in terms of unaccommodated demand in 2030 and 2040. After presenting the basic forecast results, we turn to the topic of assessing two mitigation strategies: increasing aircraft size and adding new runways to airports. Hence, we show how much capacity in terms of the number of passengers handled at an airport can be gained by these two measures up to the years 2030 and 2040. Based upon these findings, we develop some general mitigation strategies that will assist airports in deciding whether to focus on increasing aircraft size, enlarging runway capacity, or both. As in preceding chapters, we continue the airport case studies and present forecast results and capacity analyses for the airports San Diego (SAN), London Heathrow (LHR), Beijing Capital City (PEK), and now also Beijing Daxing (PKX) that opened in late 2019 and is relevant therefore for the forecasts and capacity analyses. Chapter 10, Traffic forecast and mitigation strategies, ends with some conclusion, and Chapter 11, Summary and conclusion, will close the book with an overall summary and conclusion.

223

Chapter 10

Traffic forecast and mitigation strategies In this chapter, we first present the traffic situation for the year 2016 and the forecasts for the years 2030 and 2040. Thereafter we assess different mitigation strategies for the capacity problem. The current chapter closes with the case study and finally the conclusion. The year 2016 serves as the base year in all our forecasts and analyses. With the exception of the top 20 airports in terms of unaccommodated passenger volume and the case study, the resolution level of the results presented is the seven world regions which we have already introduced earlier in this book.

10.1 Traffic situation in 2016 The traffic situation of the year 2016 has been extensively analysed in Part I of the book, the results of which serve in this section as the starting point for the 2030 and 2040 forecasts. Thus we briefly recapitulate passenger and flight volume, and aircraft size by world region.

10.1.1 Passenger volume In this chapter, passenger volume refers to the total passenger volume, including transfer passengers between two airports. Fig. 10.1 displays the annual passenger volume by world region in 2016. The number of passengers by world region corresponds thereby with the total number of enplaned passengers of that region. The sum of enplaned passengers of all seven world regions is equal to the total number of passengers carried worldwide. Almost four billion air passengers were transported in 2016 worldwide according to Sabre AirVision Market Intelligence (MI) (2016), which is a slightly higher value compared to the ICAO figure of almost 3.8 billion passengers [International Civil Aviation Organization (ICAO), 2017; see Chapter 1: Introduction]. However, for the forecast, we need data on passenger and flight volume by airport pair, so we have chosen Sabre MI data for reasons of consistency, and furthermore, forecast growth rates are not affected by this difference. Airport Capacity Constraints and Strategies for Mitigation. DOI: https://doi.org/10.1016/B978-0-12-812657-8.00010-5 © 2020 Elsevier Inc. All rights reserved.

225

Annual passenger volume

Million

226

PART | III Forecasting future air traffic development up to 2040 1600 1400 1200 1000 800 600 400

200 0

Africa Africa

Asia

Asia Europe

Europe Middle East

Middle East North America

North America South America

South America

Southwest Pacific

Southwest Pacific

FIGURE 10.1 Annual passenger volume in 2016 by world region; global sum is 4.0 billion passengers carried [Sabre AirVision Market Intelligence (MI), 2016].

More than 1.3 billion passengers enplaned in Asia, almost one billion in Europe and more than 900 million in North America in 2016. Africa, the Middle East, South America and the Southwest Pacific account for almost 710 million air passengers. About 85% of the passengers had trip origin and destination within world regions. Exceptions are, in particular, the Middle East and Africa, which have a share of intraregional traffic of only 43% and 62%, respectively, and thus are more connected to other world regions. Twenty-two per cent of the air passengers at African airports travelled between Africa and Europe and 13% between Africa and the Middle East. Here, a large part of the passenger volume flies to destinations such as Cape Town and tourism destinations in the northern part of Africa. The airports of the Middle East serve as transfer points between Europe and Asia. About 24% of air passengers travelled between the Middle East and Asia and 19% between the Middle East and Europe. Larger markets such as North America and Asia have a higher-than-average share of intraregional traffic (87% and 92%, respectively) due to more domestic traffic. However, Asia, especially, is a world region that comprises many countries including China, Japan and Indonesia. On the other hand, the North American region consists only of the United States and Canada.

10.1.2 Flight volume Fig. 10.2 shows the annual flight volume in 2016 by world region. There were 35.5 million flights in 2016 (see Fig. 3.9 in Section 3.4). The distribution of flight volume among the world regions is similar to the distribution of passenger

Million

Traffic forecast and mitigation strategies Chapter | 10

227

12

Annual flight volume

10

8

6

4

2

0

Africa Africa

Asia Asia

Europe

Europe Middle East

Middle East North America North America

South America

South America

Southwest Pacific

Southwest Pacific

FIGURE 10.2 Annual flight volume in 2016 by world region; global sum is 35.5 million flights [Sabre AirVision Market Intelligence (MI), 2016].

volume. The largest world regions in terms of flight volume are Asia (10.4 million), North America (10.0 million) and Europe (8.3 million). North America and Europe switched rankings, because average aircraft size, that is passengers per flight, is much lower in North America than in Europe, as shown in Section 10.1.3. The remaining four regions account for almost 6.7 million flights. More than 90% of the flight volume in 2016 took place within world regions. This is a slightly higher value compared to the passenger volume, as smaller aircraft are employed for shorter distances. Thus intraregional flights receive more weight in the distribution. This is especially true for the North American region, which is dominated by domestic flights. Here, 93% of the flights are intraregional, and North America accounts for 23% of the global passenger volume, and even for 28% of the global flight volume in 2016. Africa and the Middle East are again regions with the lowest share of intraregional flights (75% and 55%, respectively). However, these values are substantially higher than the corresponding shares of passenger volume, because passengers per flight tend to increase with distance travelled. This is particularly evident for flights between Africa and Europe. They account for 14% of all flights at African airports, and even for 22% of the passenger volume.

10.1.3 Aircraft size Table 10.1 presents the average number of passengers per flight (aircraft size) in 2016 by world region. This is the ratio of the values of Figs 10.1 and

TABLE 10.1 Passengers per flight (aircraft size) in 2016 by world region; global mean value is 111 [Sabre AirVision Market Intelligence (MI), 2016]. Africa

Asia

Europe

Middle East

North America

South America

Southwest Pacific

Africa

79

185

148

149

218

183

205

Asia

185

126

183

193

232

198

191

Europe

148

183

110

178

218

267

n.a.

Middle East

149

193

178

114

281

260

338

North America

218

232

218

281

85

121

245

South America

183

198

267

260

121

92

220

Southwest Pacific

205

191

n.a.

338

245

220

87

Ø

96

130

118

146

91

99

97

Traffic forecast and mitigation strategies Chapter | 10

229

10.2. The global mean value of passengers per flight is 111. Mean values by world region are shown in the last row of Table 10.1. Asia (130), Europe (118) and the Middle East (146) are above average, while Africa (96), North America (91), South America (99) and the Southwest Pacific (97) are below average. The lowest average number of passengers per flight can be found in North America due to the high share of domestic flights. Furthermore, the diagonal of Table 10.1 shows the average number of passengers per flight for intraregional flights. These values are, in every case, significantly lower than the values for interregional flights of the corresponding region. The three lowest values are Africa (79), North America (85) and South America (92). Off-diagonal values represent aircraft size of interregional flights. These values tend to increase with flight distance; however, aircraft size of flights from or to Asian airports and airports of the Middle East tend to be generally higher. South America and the Southwest Pacific are special cases due to their geographical position. While values of aircraft size for intraregional flights are among the lowest (92 and 87, respectively), values of passengers per flight for interregional flights are very high because of the large flight distances to most other regions. Finally, the top three values of passengers per flight for interregional flights are on routes of Middle East Southwest Pacific (338), North America Middle East (281) and Europe South America (267).

10.2 Forecast assumptions for 2030 and 2040 Before discussing the forecast results for the years 2030 and 2040, we need to describe the key assumptions that are essential for a sound interpretation of the results from a practitioner’s perspective. We distinguish between two kinds of assumptions as follows: G

G

Assumptions regarding input variables: this category comprises the development of future input variables of the models of Part II, such as gross domestic product (GDP) per capita, population and airfare. Assumptions resulting from the model design (Part II of the book): these assumptions are at first sight less obvious than those of the first category, but they have a substantial effect on model results.

Regarding input variables, we focus on the category ‘variable’ of Table 6.2, which means real GDP per capita, population and air fares. Forecasts of real GDP per capita and population have been retrieved from Information Handling Services (IHS) Markit (2017) for each country. Fig. 10.3 displays the distribution of growth rates of these variables between countries and the global mean values which have been calculated by weighting the original values by passenger volume of the year 2016 to account for their relevance for global air traffic development. Thus for real GDP per capita, we assume a compound annual growth rate (CAGR) of 2.49% for the

230

PART | III Forecasting future air traffic development up to 2040 5%

4%

CAGR population

3%

–4%

2% 2016–30 1%

–2%

0%

0%

–1%

2%

4%

6%

8%

10%

2030–40

–2%

CAGR real GDP per capita 2016–30

2030–40

Global mean values

FIGURE 10.3 Forecast growth rates (CAGR) for real GDP per capita and population for the time periods 2016 30 and 2030 40 [Information Handling Services (IHS) Markit, 2017]. CAGR, Compound annual growth rate.

period 2016 30 and 2.19% for the period 2030 40, which results in a CAGR of 2.36% for the period 2016 40. This means that global economic development slows down between 2030 and 2040. The economic development of more mature economies, such as North America and Europe, is forecast to be rather stable between 2016 and 2040, however, much slower compared to GDP per capita growth rates of emerging markets such as Asia and the Middle East. On the other hand, emerging markets, in particular Asia and the Middle East, are forecast to grow much faster in terms of GDP per capita, but they slow down between 2030 and 2040 compared to the 2016 30 period, for example, because of saturation effects. Regarding the development of the global population, we assume a CAGR of 0.55% for the period 2016 30 and 0.32% for the period 2030 40, which yields a CAGR of 0.46% for the period 2016 40. This means that global population growth also slows down between 2030 and 2040. While forecasts of GDP per capita and population are available, this is not the case for the future development of airfares. Based on the analyses described in Chapter 6, Modelling future air passenger demand, we assume that airfares decline globally by 1.5% p.a. on average in real terms because of further technological, institutional and organisational innovations, such as more cost-efficient aircraft, regulatory changes and improvements in air traffic management and control. Nevertheless, airfares are expected to rise in nominal terms, if inflation exceeds 1.5% per year on average. This is

Traffic forecast and mitigation strategies Chapter | 10

231

especially likely to be the case in countries with a higher level of inflation. Ultimately, this assumption does not mean that nominal airfares develop synchronously among countries and routes, as different levels of inflation lead to different levels of nominal airfare development. But, improvements in productivity, which are driven by the aforementioned innovations, are typically not limited to a particular country or flight route and are eventually spread around the globe in the long term, because air traffic is a global phenomenon. Furthermore, we do not consider measures, such as congestion pricing, because this presently is still a theoretical option, as discussed in Chapter 5, General strategies for mitigating airport capacity constraints; however, it may play a more important role in the future. Assumptions of the second category are hard-wired in the models, and the reader is referred to Part II of this book for more details. In this chapter, we focus on key assumptions, which have far-reaching effects on the forecast results and their analysis, especially from a practitioner’s perspective. These key assumptions comprise G G G G

technological innovations, airport capacity utilisation, aircraft load factor, and unaccommodated demand in 2016.

Regarding technological innovations, we assume gradual progress, but no radical change. This is reflected, for example, in the assumption about future airfare development and incorporated especially in the models of airport capacity and aircraft size. Data envelopment analysis (DEA) is not well suited for radical technological change, as it is essentially a sophisticated benchmarking approach based upon best practices. The second key assumption that underlies the forecasts is that of airport capacity utilisation. Here, we assume the highest possible utilisation that is supported by empirical evidence. This is the key assumption of the airport capacity model in Chapter 7, Modelling future airport capacity and capacity utilisation. Therefore, we assume that off-peak times during daytime will be increasingly filled up over time as traffic volume grows and available airport capacity runs short. While we do not model the flight load factor, the implicit assumption of the aircraft size model of Chapter 9, Modelling future development of the average number of passengers per flight, is that route-specific load factors achieve the highest possible level that is supported by empirical evidence, if passenger demand is sufficiently high. The higher the load factor is, the higher is the number of passengers per aircraft. Modelling load factors explicitly becomes necessary if we assign specific aircraft types to routes; however, this is not the case in this book. To sum up, we are cautious regarding rapid technological innovations and assume the best possible utilisation of airport and aircraft capacity. From our point of view, these assumptions are necessary to avoid producing forecast results that do not have a sound empirical foundation.

232

PART | III Forecasting future air traffic development up to 2040

Finally, unaccommodated passenger volume is defined in this book as the passenger demand that cannot be served because of limited airport capacity, including mitigation measures such as expanding airport capacity and employing larger aircraft with more seat capacity. However, we have chosen to disregard any unaccommodated passenger volume in 2016, because it is hard to identify with satisfactory precision by hindsight. Furthermore, as we will see in Section 10.3.1, the number of unaccommodated passengers is forecast to be still very low in 2030 on a global level, so the assumption of almost no unaccommodated passengers in 2016 seems to be not too restrictive. For 2030, this volume is relevant only for a small number of important airports (see Table 10.7 further down) and in 2016 most likely just for London Heathrow (LHR). Thus as it seems to be of only minor relevance for 2016 and to avoid arbitrary results, we maintain this assumption, but we have to keep in mind that strictly speaking the forecast unaccommodated passenger volume for 2030 and 2040 is only the difference to 2016, so that the numbers are just slightly on the low side in total terms, maybe with the exception of LHR.

10.3 Traffic forecasts for 2016 30 In this section, we present the results of the traffic forecast for 2030. As for the base year of 2016, we display passenger and flight volume, and aircraft size of the year 2030 and their growth rates (CAGR) for the time period from 2016 to 2030. Furthermore, we identify the unaccommodated demand by world region and the top 20 airports worldwide in terms of unaccommodated demand in 2030.

10.3.1 Passenger volume Fig. 10.4 shows the forecast annual passenger volume by world region for the year 2030. A global passenger volume of almost seven billion air passengers is expected for 2030, which is about 1.76 times the passenger volume of 2016, with a CAGR of 4.1% (see Table 10.2). More than 2.6 billion passengers are forecast for Asia, followed by Europe with almost 1.6 billion passengers and North America with nearly 1.5 billion passengers. Africa, the Middle East, South America and the Southwest Pacific are expected to account for about 1.3 billion air passengers in 2030. As in 2016, 85% of the passengers are forecast to have flight origins and destinations within world regions. Thus 5.9 billion passengers use intraregional air services. The relative intraregional distribution of passenger volume is forecast to remain more or less the same as in 2016. The Middle East and Africa are expected have a share of intraregional traffic of about 42% and 64%, respectively, which is only marginally different from the situation of 2016. Twenty per cent of air passengers of Africa are forecast to travel to

Million

Traffic forecast and mitigation strategies Chapter | 10

233

3000

Annual passenger volume

2500

2000

1500

1000

500

0

Africa Africa

Asia

Asia Europe

Europe Middle East

Middle East North America

North America South America

South America

Southwest Pacific

Southwest Pacific

FIGURE 10.4 Forecast annual passenger volume for the year 2030 by world region; global sum is 6.9 billion passengers carried.

Europe and 14% to the Middle East. This means that the Middle East is expected to gain slightly in importance for African airports, while European airports will become less important, however, only very slightly. In 2030 the airports of the Middle East will still serve as transfer points between Europe and Asia. A share of 27% of air passengers is forecast to travel between the Middle East and Asia and 18% between the Middle East and Europe. Thus the Asian region gains in importance for the Middle East, while the European region becomes slightly less important. The higher-than-average shares of intraregional travel of the large markets of North America and Asia (87% and 92%, respectively) will remain virtually constant in 2030 compared to 2016. Table 10.2 displays the annual passenger volume growth rates for the time period from 2016 to 2030. A CAGR of 4.1% is forecast for the global mean value; however, there is significant variation between world regions. Asia, the Middle East and Africa show the highest annual growth rates of 4.9%, 4.5% and 4.4%, respectively. However, the value of Africa, in particular, has to be put in perspective, because of the low demand volume level. Africa and the Southwest Pacific have by far the lowest passenger volume in 2016 and 2030, and about one-third percentage points above-average growth rates over a time period of 14 years will not change things significantly. The same applies more or less to the Middle East, which is the third smallest region in terms of passenger volume in 2016 and 2030. All the more important becomes the value of 4.9% of the Asian region, as it is the largest world

TABLE 10.2 Forecast annual passenger volume growth rates per year [compound annual growth rate (CAGR)] for the period 2016 30 by world region; global mean value is 4.1%. Africa (%)

Asia (%)

Europe (%)

Middle East (%)

North America (%)

South America (%)

Southwest Pacific (%)

Africa

4.6

5.4

3.9

4.6

4.0

4.1

4.1

Asia

5.4

5.0

3.9

5.0

3.9

4.2

4.1

Europe

3.9

3.9

3.3

4.1

3.3

3.7

n.a.

Middle East

4.6

5.0

4.1

4.4

4.2

4.6

4.5

North America

4.0

3.9

3.3

4.2

3.5

3.7

3.8

South America

4.1

4.2

3.7

4.6

3.7

4.1

4.2

Southwest Pacific

4.1

4.1

n.a.

4.5

3.8

4.2

3.7

Ø

4.4

4.9

3.4

4.5

3.5

4.0

3.8

Traffic forecast and mitigation strategies Chapter | 10

235

Unaccommodated annual passenger volume

Million

region in terms of passenger volume both in 2016 and 2030. Therefore the global mean value of 4.1% is essentially a result of the strong passenger volume growth of the Asian region, as the other two large regions, Europe and North America, are expected to grow much slower, because their markets are more developed. In fact, all regions except Africa, Asia and the Middle East show below-average growth rates in terms of forecast passenger volume in 2030. The highest growth rate of Table 10.2 with 5.4% is the growing air travel between Africa and Asia, followed by a value of 5.0% for air travel between Asia and the Middle East, as well as within the Asian region. However, while passenger volume between Africa and Asia accounts for only 0.1% and between Asia and the Middle East just for about 2.4% of the forecast global passenger volume in 2030, the Asian domestic volume has a share of more than 35%. In contrast, the three lowest growth rates can be found for the European domestic (3.3%), Europe North America (3.3%) and North American domestic (3.5%) travel flows. In particular, the growth of European and North American domestic passenger volumes has a significant impact on the global mean value, as they each account for about 19% of the forecast global passenger volume in 2030. Fig. 10.5 presents the unaccommodated annual passenger volume in 2030 by world region. As already explained, unaccommodated passenger volume is the forecast passenger demand that cannot be served as a result of a capacity shortage at airports, including mitigation measures such as expanding 25

20

15

10

5

0

Africa Africa

Asia Asia

Europe

Europe Middle East

Middle East North America North America

South America

South America

Southwest Pacific

Southwest Pacific

FIGURE 10.5 Forecast unaccommodated annual passenger volume for the year 2030 by world region; global sum is 49.4 million passengers.

236

PART | III Forecasting future air traffic development up to 2040

airport capacity and employing larger aircraft with more seat capacity. Thus the forecast ‘unconstrained’ passenger demand is defined by the sum of the values of Figs 10.4 and 10.5; however, this is only a theoretical value. Unaccommodated passenger volume is computed for each airport pair to consider route-specific developments regarding passengers per flight and airport capacity limits. The unaccommodated passenger volume totals up to 49 million passengers worldwide in 2030 and represents about 0.7% of the forecast global unconstrained passenger volume in 2030. The largest amounts of unaccommodated passenger volume in 2030 can be found in Asia (21.0 million), North America (19.2 million) and Europe (7.0 million), because airport capacity and aircraft size developments in these regions are insufficient for serving the unconstrained passenger demand. In these three regions, intraregional unaccommodated passenger volume has the highest share of total unaccommodated passenger volume per region (58% for Europe, 89% for North America and 92% for Asia). On the other hand, intraregional unaccommodated passenger volume does not exist in Africa, the Middle East, South America and the Southwest Pacific, as sufficient airport capacity is expected for 2030. Unaccommodated passenger volume in these regions is solely a result of insufficient airport capacity in other regions and sums up to about 2.2 million passengers in 2030. Here, we call unaccommodated demand of a region or airport, which is caused by a lack of airport capacity of the same region or airport, ‘direct capacity constraints’, while we name unaccommodated demand of a region or airport, which is caused by a lack of airport capacity of another region or airport, ‘indirect capacity constraints’. Thus the unaccommodated demand of Asia, Europe and North America is primarily caused by direct capacity constraints, while Africa, the Middle East, South America and the Southwest Pacific suffer from indirect capacity constraints. Table 10.3 displays the forecast share of unaccommodated passenger volume for 2030 by world region, which is defined as the ratio of the values of Fig. 10.5 to the sum of the values of Figs 10.4 and 10.5, that means the ratio of unaccommodated passenger volume to unconstrained passenger volume. North America, Asia and Europe show the highest values, while Africa, South America and the Southwest Pacific have almost negligible shares of unaccommodated demand, which can be found exclusively in interregional passenger volume. The case of the Middle East is quite unique. While there is no capacity shortage at Middle Eastern airports, the forecast share of unaccommodated demand is almost as high as that of Europe, because of airport capacity constraints in Asia and Europe.

10.3.2 Flight volume Fig. 10.6 shows the flight volume forecast for 2030 by world region. Nearly 46 million flights are forecast for the year 2030 globally, which means that

TABLE 10.3 Forecast share of unaccommodated passenger volume for the year 2030 by world region; global mean value is 0.9%. Africa (%)

Asia (%)

Europe (%)

Middle East (%)

North America (%)

South America (%)

Southwest Pacific (%)

Africa

0.0

2.4

0.5

0.0

1.1

0.0

0.0

Asia

2.4

0.8

1.2

0.8

0.7

0.0

0.4

Europe

0.5

1.2

0.3

0.8

2.1

0.3

n.a.

Middle East

0.0

0.8

0.8

0.0

0.5

0.0

0.0

North America

1.1

0.7

2.1

0.5

1.3

0.6

0.0

South America

0.0

0.0

0.3

0.0

0.6

0.0

0.0

Southwest Pacific

0.0

0.4

n.a.

0.0

0.0

0.0

0.0

Ø

0.1

0.8

0.4

0.4

1.3

0.1

0.0

Annual flight volume

Million

238

PART | III Forecasting future air traffic development up to 2040 16 14 12 10 8 6 4 2 0

Africa Africa

Asia Asia

Europe

Europe Middle East

Middle East North America North America

South America

South America

Southwest Pacific

Southwest Pacific

FIGURE 10.6 Forecast annual flight volume for the year 2030 by world region; global sum is 45.8 million flights.

the global flight volume grows by almost 29% compared to 2016; however, this is much lower than the growth of passenger volume of 76%. The largest regions in terms of flight volume are Asia (14.6 million), North America (12.3 million) and Europe (10.1 million). Africa, the Middle East, South America and the Southwest Pacific account for almost 8.7 million flights in 2030. As in 2016, 90% of the traffic is forecast to take place within world regions. Thus 41.3 million flights are intraregional. The relative intraregional distribution of flight volume is expected to remain more or less the same as in 2016. With the exception of Africa and the Middle East, the share of intraregional flights is between 87% (South America) and 95% (Asia). The Middle East and Africa are well below average, and they are expected to have a share of about 55% and 76%, respectively, which is virtually the same as in 2016. Table 10.4 presents the annual flight volume growth rates for the time period from 2016 to 2030. A CAGR of 1.8% is forecast for the global mean value; however, there is considerable variation between world regions. Asia, the Middle East and Africa show the highest annual growth rates of 2.4%, 2.2% and 1.9%, respectively. Europe and North America display forecast growth rates of 1.4% and 1.5%, respectively, which are well below average. South America and the Southwest Pacific each show values of 1.8%. The highest forecast values can be found for the routes of South America Middle East and Southwest Pacific Middle East (2.6%). On the other hand, the lowest value of 1.4% applies for North American and European

TABLE 10.4 Forecast annual flight volume growth rates per year [compound annual growth rate (CAGR)] for the period 2016 30 by world region; global mean value is 1.8%. Africa (%)

Asia (%)

Europe (%)

Middle East (%)

North America (%)

South America (%)

Southwest Pacific (%)

Africa

2.0

2.4

1.6

2.2

1.8

1.9

2.0

Asia

2.4

2.5

1.7

2.5

2.0

1.9

2.0

Europe

1.6

1.7

1.4

1.8

1.4

1.9

n.a.

Middle East

2.2

2.5

1.8

2.1

2.2

2.6

2.6

North America

1.8

2.0

1.4

2.2

1.4

1.6

2.1

South America

1.9

1.9

1.9

2.6

1.6

1.9

2.2

Southwest Pacific

2.0

2.0

n.a.

2.6

2.1

2.2

1.7

Ø

1.9

2.4

1.4

2.2

1.5

1.8

1.8

240

PART | III Forecasting future air traffic development up to 2040

domestic, and North American European routes. For 2030 we expect only minor effects of limited airport capacity in South America, the Southwest Pacific and the Middle East, while there will be a significant capacity shortage at airports in Europe, North America and even Asia due to the strong demand growth, eventually limiting further flight volume growth.

10.3.3 Aircraft size Table 10.5 presents the forecast values of the average number of passengers per flight (aircraft size) in 2030, which is the ratio of the values of Figs 10.4 and 10.6. On a global level, average aircraft size is forecast to increase from 111 in 2016 to 152 passengers per flight between 2016 and 2030. Aircraft size is expected to be clearly above average for Asia (181) and the Middle East (201), while the values for Africa (134), North America (120), South America (133) and the Southwest Pacific (128) are forecast to be well below average. The value of Europe (154) is close to the global mean value of 152 passengers per flight. As in 2016, intraregional flights carry on average a lower number of passengers per flight compared to interregional flights in 2030. The lowest number of passengers per flight is forecast for flights within North America (112), Africa (112) and the Southwest Pacific (114). On the other hand, the highest number of passengers per flight is expected for interregional flights, such as Middle East Southwest Pacific (437), Middle East North America (367), South America Middle East (342) and South America Europe (342). Table 10.6 displays the forecast annual growth rates of aircraft size for the time period 2016 30. The global mean value is 2.2%. Africa (2.4%), Asia (2.4%) and the Middle East (2.3%) show above-average values, while the average aircraft size development of Europe (2.0%), North America (2.0%), South America (2.1%) and the Southwest Pacific (2.0%) is below average. The three highest values are 2.9% for African Asian, 2.6% for African domestic and 2.5% for Asian Middle Eastern routes, while the lowest value is 1.7% for North American Southwest Pacific, 1.8% for Asian North American, Middle Eastern Southwest Pacific and European South American routes and 1.9% for European domestic, European North American and North American Middle Eastern routes.

10.3.4 Top 20 airports in terms of unaccommodated demand in 2030 Table 10.7 presents the top 20 airports regarding unaccommodated passenger demand volume in 2030. Among these 20 airports are ten from Asia, one from Europe, one from the Middle East and eight from North America. As expected, most airports are from world regions which suffer mostly from direct capacity constraints. Here, Atlanta Hartsfield–Jackson (ATL) is

TABLE 10.5 Forecast of the average number of passengers per flight (aircraft size) in 2030 by world region; global mean value is 152. Africa

Asia

Europe

Middle East

North America

South America

Southwest Pacific

Africa

112

275

201

206

294

250

273

Asia

275

177

246

274

298

269

253

Europe

201

246

144

240

284

342

n.a.

Middle East

206

274

240

156

367

342

437

North America

294

298

284

367

112

161

311

South America

250

269

342

342

161

125

292

Southwest Pacific

273

253

n.a.

437

311

292

114

Ø

134

181

154

201

120

133

128

TABLE 10.6 Forecast average aircraft size growth rates per year [compound annual growth rate (CAGR)] for the period 2016 30 by world region; global mean value is 2.2%. Africa (%)

Asia (%)

Europe (%)

Middle East (%)

North America (%)

South America (%)

Southwest Pacific (%)

Africa

2.6

2.9

2.2

2.4

2.2

2.2

2.1

Asia

2.9

2.4

2.1

2.5

1.8

2.2

2.0

Europe

2.2

2.1

1.9

2.2

1.9

1.8

n.a.

Middle East

2.4

2.5

2.2

2.2

1.9

2.0

1.8

North America

2.2

1.8

1.9

1.9

2.0

2.1

1.7

South America

2.2

2.2

1.8

2.0

2.1

2.2

2.0

Southwest Pacific

2.1

2.0

n.a.

1.8

1.7

2.0

2.0

Ø

2.4

2.4

2.0

2.3

2.0

2.1

2.0

TABLE 10.7 Top 20 airports in terms of unaccommodated passenger volume in 2030. No.

IATA code

Airport name

Aircraft movements (thousand)

Passengers (million)

Aircraft size

Unaccommodated passengers (million)

Share of unaccommodated passengers (%)

Capacity utilisation (%)

1

ATL

Atlanta Hartsfield Jackson

974

158

162

11.2

6.6

100.0

2

LHR

London Heathrow

523

117

224

8.6

6.8

100.0

3

BOM

Mumbai Chhatrapati Shivaji

447

98

220

6.7

6.4

100.0

4

CAN

Guangzhou Baiyun

592

121

205

6.7

5.2

100.0

5

ORD

Chicago O’Hare

945

117

123

6.4

5.2

100.0

6

CGK

Jakarta SoekarnoHatta

629

122

195

5.6

4.4

100.0

7

DEL

Delhi Indira Gandhi

599

127

212

3.6

2.7

100.0

8

LGA

New York LaGuardia

416

43

104

2.3

5.1

100.0

9

BLR

Bengaluru Kempegowda

274

51

188

1.0

1.8

100.0

10

DXB

Dubai

538

172

319

0.9

0.5

91.0

11

SIN

Singapore Changi

480

108

224

0.8

0.7

58.5 (Continued )

TABLE 10.7 (Continued) No.

IATA code

Airport name

Aircraft movements (thousand)

Passengers (million)

Aircraft size

Unaccommodated passengers (million)

Share of unaccommodated passengers (%)

Capacity utilisation (%)

12

LAX

Los Angeles

799

129

162

0.7

0.5

98.2

13

PEK & PKX

Beijing Capital City and Daxing

927

202

218

0.6

0.3

45.5

14

MIA

Miami

404

68

168

0.6

0.8

55.5

15

DFW

Dallas/Fort Worth

807

101

125

0.5

0.5

92.5

16

MCO

Orlando

368

65

176

0.5

0.8

51.6

17

HYD

Hyderabad Rajiv Gandhi

182

35

194

0.5

1.4

32.4

18

MAA

Chennai

218

41

188

0.5

1.2

49.6

19

SUB

Juanda

247

45

183

0.5

1.1

47.3

20

BOS

Boston Logan

422

57

135

0.5

0.9

49.0

Traffic forecast and mitigation strategies Chapter | 10

245

expected to be the airport with the highest unaccommodated passenger demand volume (11.2 million), followed by London Heathrow (LHR, 8. 6 million) and Mumbai Chhatrapati Shivaji (BOM, 6.7 million). Their unaccommodated passenger volume represents 6.6%, 6.8% and 6.4%, respectively, of their forecast unconstrained passenger volume in 2030. Airports such as ATL, LHR and BOM have serious capacity problems and thus suffer primarily from direct capacity constraints and have a capacity utilisation of 100%. On the other hand, the airports on the second half of the list of Table 10.7 have a capacity utilisation of less than 100% but suffer from indirect capacity constraints, meaning a capacity shortage at the destination airport. They typically have a forecast unaccommodated passenger volume of around one million per year or less, which is, in most cases, about 1% of their forecast unconstrained passenger volume in 2030 or less. Examples include the two hub airports of Beijing (PEK and PKX), which are treated as an airport system in terms of airport capacity in this book, Chennai (MAA) and Juanda (SUB). Unaccommodated demand volume of these airports lies in a range of between 500,000 and 600,000 passengers in 2030, which is about 0.3% 1.2% of their forecast unconstrained passenger volume. They suffer primarily from indirect capacity constraints, which means that they cannot handle the full demand potential due to a capacity shortage at other destinations. Dubai airport (DXB) is another remarkable case. While not having a capacity shortage itself (nor any other airport of the Middle East) in 2030, unaccommodated passenger volume totals almost one million (0.5% of forecast unconstrained passenger volume) solely because of capacity constraints at airports outside the Middle East. Most of the unaccommodated passenger volume is concentrated on a small number of airports in 2030. The top ten airports account for about 54% of the unaccommodated passenger volume worldwide. Here, the values of Table 10.7 need to be divided by a value of two to allow comparisons with Fig. 10.5, because passengers are counted twice in Table 10.7, as arriving and departing passengers per airport. Nevertheless, while unaccommodated passenger volume in 2030 is rather small compared to the forecast passenger volume on a global level, it is quite significant for particular world regions and airports. However, this is set to change for the time period from 2030 to 2040 and will be described in the next section.

10.4 Traffic forecasts for 2030 40 In this section, we continue with the results of the traffic forecast for 2040. This time, the baseline year is the forecast for 2030, to highlight differences in traffic development between the time periods 2016 30 and 2030 40. In particular, we expect airport capacity constraints to become more important during the second period compared to the first. As for the 2030 forecast, we show passenger and flight volume, and aircraft size of the year 2040 and

246

PART | III Forecasting future air traffic development up to 2040

their growth rates (CAGR) for the time period from 2030 to 2040. Again, we identify the unaccommodated demand by world region and the top 20 airports worldwide in terms of unaccommodated demand in 2040.

10.4.1 Passenger volume

Million

Fig. 10.7 presents the forecast annual passenger volume by world region for the year 2040. Nearly 9.4 billion air passengers are expected for 2040 globally, which is about 1.36 times the passenger volume forecast for 2030 and equals a CAGR of 3.1% (see Table 10.8). This is substantially less than for the period 2016 30 and is due to a slower economic development and population growth (see Fig. 10.3), as well as the increasing importance of airport capacity constraints. Almost 3.7 billion passengers are forecast for Asia, followed by Europe with around two billion passengers and North America with more than 1.9 billion passengers. Africa, the Middle East, South America and the Southwest Pacific are expected to account for nearly 1.8 billion air passengers in 2040. As before, 85% of the traffic is forecast to take place within world regions. Thus almost eight billion passengers will use intraregional air services. The relative intraregional distribution of passenger volume is forecast to remain more or less the same as in 2016 and 2030. The Middle East and Africa are expected to have a share of intraregional traffic of about 42%

4000

Annual passenger volume

3500 3000 2500 2000 1500 1000 500 0

Africa Africa

Asia

Asia Europe

Europe Middle East

Middle East North America

North America South America

South America

Southwest Pacific

Southwest Pacific

FIGURE 10.7 Forecast annual passenger volume for the year 2040 by world region; global sum is 9.4 billion.

TABLE 10.8 Forecast annual passenger volume growth rates per year [compound annual growth rate (CAGR)] for the period 2030 40 by world region; global mean value is 3.1%. Africa (%)

Asia (%)

Europe (%)

Middle East (%)

North America (%)

South America (%)

Southwest Pacific (%)

Africa

3.7

3.5

3.0

3.7

3.2

3.7

3.6

Asia

3.5

3.4

2.7

3.5

2.8

3.4

3.1

Europe

3.0

2.7

2.6

3.1

2.5

3.1

n.a.

Middle East

3.7

3.5

3.1

3.5

3.2

3.8

3.5

North America

3.2

2.8

2.5

3.2

2.7

3.1

2.9

South America

3.7

3.4

3.1

3.8

3.1

3.6

3.6

Southwest Pacific

3.6

3.1

n.a.

3.5

2.9

3.6

2.9

Ø

3.6

3.3

2.6

3.5

2.8

3.4

3.0

248

PART | III Forecasting future air traffic development up to 2040

and 65%, respectively, which is only a marginal difference compared to the situation of 2016 and 2030. Nineteen per cent of air passengers of Africa are forecast to travel to Europe and 14% to the Middle East. As in 2016 and 2030 the airports of the Middle East are major transfer points of passengers between Europe and Asia. A share of 27% of air passengers is forecast to travel between the Middle East and Asia and 17% between the Middle East and Europe, which are more or less the values of 2016 and 2030. The higher-than-average shares of intraregional travel of the large markets of North America and Asia (87% and 93%, respectively) will remain virtually constant in 2040 compared to 2030 and 2016. Thus as expected, there is no significant change in the traffic structure between world regions, which is no surprise given the share of interregional traffic of only 15%. Table 10.8 shows the average growth rates of annual passenger volume for the time period 2030 40. A CAGR of 3.1% is forecast for the global mean value, and there is significant variation between world regions. As for the period 2016 30, traffic of Africa, the Middle East and Asia grows with the highest annual growth rates of 3.6%, 3.5% and 3.3%, respectively; however, they are expected to decrease substantially compared to the previous period. Africa and the Middle East decline by 0.8% and 1.0% points, respectively, and Asia by even 1.6% points. In particular, Asia’s limited airport capacity plays a major role. Africa and the Middle East suffer from a capacity shortage in Asia, while there are no capacity constraints at their own airports. As in the previous period the global mean value of 3.1% is mainly a result of the forecast strong passenger volume growth of the Asian region, as the other two large regions, Europe and North America, are expected to grow much slower. The highest growth rate of Table 10.8 is 3.8% and can be found for air travel between South America and the Middle East, followed by a value of 3.7% for air travel between Africa and the Middle East and South America, respectively, as well as within the African region. The common feature of these relations is that airport capacity constraints play no role. In contrast, the three lowest growth rates can be found for European North American (2.5%), European domestic (2.6%) and European Asian as well as North American domestic (2.7%) routes. In particular, the Europe domestic figure has a considerable impact on the global mean value, as it still accounts for about 18% of the forecast global passenger volume in 2040. Nevertheless, these markets are all plagued by a capacity shortage at their main airports. Fig. 10.8 shows the unaccommodated annual passenger volume in 2040 by world region. The unaccommodated passenger volume sums up to almost 256 million passengers worldwide in 2040 and is 2.6% of the forecast global unconstrained passenger volume. This is a substantial increase compared to 2030, both in absolute as well as in relative terms. The largest levels of unaccommodated passenger volume in 2040 can be found in Asia (156.6 million), North America (63.2 million) and Europe (22.2 million), as airport capacity

Unaccommodated annual passenger volume

Million

Traffic forecast and mitigation strategies Chapter | 10

249

180 160 140 120

100 80 60 40 20 0

Africa Africa

Asia Asia

Europe

Europe Middle East

Middle East North America North America

South America

South America

Southwest Pacific

Southwest Pacific

FIGURE 10.8 Forecast unaccommodated annual passenger volume for the year 2040 by world region; global sum is 286.7 million passengers.

constraints become increasingly more important, and aircraft size developments cannot fully offset the capacity shortage. As in 2030 intraregional unaccommodated passenger volume has the highest share of total unaccommodated passenger volume in these regions (53% for Europe, 88% for North America and 93% for Asia). On the other hand, intraregional unaccommodated passenger volume does not exist in Africa, the Middle East and the Southwest Pacific and only marginally in South America, as more or less sufficient airport capacity is expected in 2040. Unaccommodated passenger volume in these regions is mainly a result of insufficient airport capacity in other regions and sums up to about 13.4 million passengers in 2040. However, we expect things to change beyond 2040, as can be seen exemplarily by the South American region. Here, airports such as Sao Paulo Congonhas (CGH), Mexico City (MEX) and Cancun (CUN) already suffer from direct capacity constraints; however, their share of unaccommodated demand is 1% or even less. Table 10.9 displays the forecast share of unaccommodated passenger volume for 2040 by world region. Asia (4.1%), North America (3.2%) and Middle East (1.8%) show the highest values, while Africa (0.4%), South America (0.5%) and the Southwest Pacific (0.2%) have low shares of unaccommodated demand, which can be found with the exception of South America exclusively in interregional passenger volume. The situation of the Middle East worsens compared to 2030. While there is still no capacity shortage at airports of the Middle East, the forecast share of

TABLE 10.9 Forecast share of unaccommodated passenger volume for the year 2040 by world region; global mean value is 3.0%. Africa (%)

Asia (%)

Europe (%)

Middle East (%)

North America (%)

South America (%)

Southwest Pacific (%)

Africa

0.0

8.1

1.0

0.0

2.5

0.0

0.0

Asia

8.1

4.1

4.3

5.1

2.7

0.0

1.3

Europe

1.0

4.3

0.7

1.7

4.9

0.8

n.a.

Middle East

0.0

5.1

1.7

0.0

1.8

0.0

0.0

North America

2.5

2.7

4.9

1.8

3.2

1.9

2.0

South America

0.0

0.0

0.8

0.0

1.9

0.2

0.0

Southwest Pacific

0.0

1.3

n.a.

0.0

2.0

0.0

0.0

Ø

0.4

4.1

1.1

1.8

3.2

0.5

0.2

Traffic forecast and mitigation strategies Chapter | 10

251

unaccommodated demand rises from 0.4% to 1.8% in 2030, in particular due to an increasing capacity shortage at Asian airports, but limited airport capacity in Europe and North America play a role as well.

10.4.2 Flight volume

Million

Fig. 10.9 presents the flight volume forecast for 2040 by world region. More than 52 million flights are forecast for the year 2040 globally, which means that the global flight volume grows by almost 15% compared to 2030; however, this is a considerably lower value compared to the period 2016 30 (129%). The largest regions in terms of flight volume are Asia (17.2 million), North America (13.9 million) and Europe (11.3 million). Africa, the Middle East, South America and the Southwest Pacific account for around 10.1 million flights in 2040. Ninety-one per cent of the traffic is forecast to have flight origins and destinations within world regions, which is virtually the same as in 2016 and 2030. As a result, 47.5 million flights depart and arrive within the same region. The relative intraregional distribution of flight volume is expected to remain more or less the same as in 2016 and 2030. With the exception of Africa and the Middle East, the share of intraregional flights is between 87% (South America) and 95% (Asia). The Middle East and Africa are well below average, and they are expected to have a share of about 55% and 76%, respectively, which is virtually the same as in 2016 and 2030. 20 18

Annual flight volume

16 14 12 10 8 6 4 2 0

Africa Africa

Asia Asia

Europe

Europe Middle East

Middle East North America North America

South America

South America

Southwest Pacific

Southwest Pacific

FIGURE 10.9 Forecast annual flight volume for the year 2040 by world region; global sum is 52.7 million flights.

252

PART | III Forecasting future air traffic development up to 2040

Table 10.10 illustrates the annual flight volume growth rates for the time period from 2030 to 2040. A CAGR of 1.4% is forecast globally; however, there is considerable variation between the different world regions. Asia, the Middle East and South America each have the highest annual growth rates of 1.6%. Europe and North America display forecast growth rates of 1.1% and 1.2%, respectively, which are well below average. Africa and the Southwest Pacific show values each of 1.3% and 1.5%, respectively. The highest forecast values can be found for South America Middle East and Southwest Pacific Middle East (2.3%). These markets do not have a shortage of airport capacity. On the other hand, the lowest value of 1.0% refers to Europe North America and Africa Europe, which is a result of the declining GDP growth rates that are expected until 2040 and increasing capacity problems at airports.

10.4.3 Aircraft size Table 10.11 displays the forecast values of the average number of passengers per flight (aircraft size) in 2040. On average, aircraft size is forecast to increase from 152 in 2030 to 179 passengers per flight in 2040 worldwide. It is expected to be clearly above average for Asia (214) and the Middle East (242), while the values for Africa (167), North America (139), South America (158) and the Southwest Pacific (148) are forecast to be well below average. The value of Europe (180) is very close to the global mean value. As in 2016 and 2030 intraregional flights carry on average a lower number of passengers per flight compared to interregional flights in 2040. The lowest number of passengers per flight is forecast for flights within North America (130), the Southwest Pacific (131) and Africa (142). On the other hand, the highest number of passengers per flight is forecast for interregional flights, such as Middle East Southwest Pacific (490), Middle East North America (422) and South America Middle East (396). Table 10.12 displays the forecast annual growth rates of aircraft size for the time period 2030 40. The global mean value is 1.7%, meaning that average growth rates slow down substantially compared to the previous period despite increasing capacity constraints, because there are upper limits to aircraft size and the rate at which it can be increased. This reveals the dilemma of the Asian region. Forecast demand growth for 2030 and 2040 is high, and the level of demand is already quite substantial in 2016 and 2030. However, in the long run, there are both limits to expanding airport capacity and to increasing aircraft size, thus leading to high levels of unaccommodated demand. While increasing aircraft size, typically, is easier than expanding airport capacity by adding new runways, there are still upper limits to aircraft size and the rate at which it can be increased, as a comparison of Tables 10.12 and 10.6 illustrates by the decreasing growth rates.

TABLE 10.10 Forecast annual flight volume growth rates per year [compound annual growth rate (CAGR)] for the period 2030 40 by world region; global mean value is 1.4%. Africa (%)

Asia (%)

Europe (%)

Middle East (%)

North America (%)

South America (%)

Southwest Pacific (%)

Africa

1.3

1.3

1.0

1.7

1.1

1.3

1.4

Asia

1.3

1.7

1.1

1.4

1.5

1.7

1.7

Europe

1.0

1.1

1.1

1.4

1.0

1.6

n.a.

Middle East

1.7

1.4

1.4

1.7

1.8

2.3

2.3

North America

1.1

1.5

1.0

1.8

1.2

1.2

1.4

South America

1.3

1.7

1.6

2.3

1.2

1.7

1.8

Southwest Pacific

1.4

1.7

n.a.

2.3

1.4

1.8

1.5

Ø

1.3

1.6

1.1

1.6

1.2

1.6

1.5

TABLE 10.11 Forecast of the average number of passengers per flight (aircraft size) in 2040 by world region; global mean value is 179. Africa

Asia

Europe

Middle East

North America

South America

Southwest Pacific

Africa

142

344

245

250

361

314

336

Asia

344

209

290

336

339

318

292

Europe

245

290

167

285

331

395

n.a.

Middle East

250

336

285

185

422

396

490

North America

361

339

331

422

130

193

358

South America

314

318

395

396

193

150

346

Southwest Pacific

336

292

n.a.

490

358

346

131

Ø

167

214

180

242

139

158

148

TABLE 10.12 Forecast average aircraft size growth rates per year [compound annual growth rate (CAGR)] for the period 2030 40 by world region; global mean value is 1.7%. Africa (%)

Asia (%)

Europe (%)

Middle East (%)

North America (%)

South America (%)

Southwest Pacific (%)

Africa

2.4

2.2

2.0

1.9

2.1

2.3

2.1

Asia

2.2

1.7

1.7

2.1

1.3

1.7

1.4

Europe

2.0

1.7

1.5

1.7

1.5

1.5

n.a.

Middle East

1.9

2.1

1.7

1.8

1.4

1.5

1.2

North America

2.1

1.3

1.5

1.4

1.5

1.8

1.4

South America

2.3

1.7

1.5

1.5

1.8

1.8

1.7

Southwest Pacific

2.1

1.4

n.a.

1.2

1.4

1.7

1.5

Ø

2.2

1.7

1.5

1.8

1.5

1.8

1.5

256

PART | III Forecasting future air traffic development up to 2040

Looking at the seven world regions and their interdependencies, Africa (2.2%), Asia (1.7%), the Middle East (1.8%) and South America (1.8%) show above-average values, while the average aircraft size development of Europe (1.5%), North America (1.5%) and the Southwest Pacific (1.5%) is below average. The three highest values are 2.4% for Africa domestic, 2.3% for Africa South America and 2.2% for Asia Africa, while the lowest values are 1.2% for Southwest Pacific Middle East, 1.3% for North America Asia and 1.4% for Middle East North America, Asia Southwest Pacific and North America Southwest Pacific.

10.4.4 Top 20 airports in terms of unaccommodated demand in 2040 Table 10.13 shows the top 20 airports regarding unaccommodated passenger demand volume in 2040. Among these 20 airports are 14 from Asia, one from Europe, one from the Middle East and four from North America. There are 14 airports suffering from direct capacity constraints (capacity utilisation of 100%), that is five more compared to 2030. Dubai (DXB) and Singapore Changi (SIN) airports are examples which still do not have a capacity shortage but suffer from capacity constraints at destination airports. Unaccommodated demand of these airports is expected to total between 5.7 and 6.3 million passengers in 2040, which is 2.6% and 3.8%, respectively, of their forecast unconstrained passenger volume for 2040. This is quite substantial, given the fact that these airports do not suffer from direct capacity constraints. Overall, unaccommodated passenger volume of airports with capacity reserves but suffering from indirect capacity constraints has increased substantially compared to 2030. Delhi Indira Gandhi (DEL) is expected to be the airport with the highest unaccommodated demand volume of 46.0 million passengers, followed by Mumbai Chhatrapati Shivaji (BOM, 42.1 million) and Jakarta Soekarno-Hatta (CGK, 41.8 million). Thus the top three airports are exclusively from Asia and their unaccommodated passenger volume represents 22.4%, 25.5% and 21.5% respectively, of their forecast unconstrained passenger volume in 2040. As in 2030, most of the unaccommodated passenger volume is still concentrated on a small number of airports in 2040. The top ten airports account for about 49% and the top 20 for about 59% of the unaccommodated passenger volume worldwide. As for 2030 the values of Table 10.13 need to be divided by a value of two to allow comparisons with Fig. 10.8, because passengers are counted twice in Table 10.13, that means arriving and departing passengers per airport. As indicated earlier in the book, unaccommodated passenger volume in 2040 is only a small part of the total forecast passenger volume; however, it is much more important for particular world regions and airports. The main reason for this development is that mitigation measures such as expanding airport capacity and increasing aircraft size increasingly

TABLE 10.13 Top 20 airports in terms of unaccommodated passenger volume in 2040. No.

IATA code

Airport name

Aircraft movements (thousand)

Passengers (million)

Aircraft size

Unaccommodated passengers (million)

Share of unaccommodated passengers (%)

Capacity utilisation (%)

1

DEL

Delhi Indira Gandhi

599

159

265

46.0

22.4

100.0

2

BOM

Mumbai Chhatrapati Shivaji

447

123

276

42.1

25.5

100.0

3

CGK

Jakarta Soekarno Hatta

629

152

242

41.8

21.5

100.0

4

ATL

Atlanta Hartsfield Jackson

974

195

200

33.6

14.7

100.0

5

LHR

London Heathrow

523

144

275

25.2

14.9

100.0

6

ORD

Chicago O’Hare

945

144

153

20.7

12.5

100.0

7

BLR

Bengaluru Kempegowda

274

65

237

17.6

21.3

100.0

8

KUL

Kuala Lumpur

593

140

236

9.4

6.3

100.0

9

MNL

Manila Ninoy Aquino

442

113

256

7.7

6.4

100.0

10

LAX

Los Angeles

814

167

205

7.1

4.1

100.0

11

DFW

Dallas/Fort Worth

881

130

148

6.1

4.4

100.0 (Continued )

TABLE 10.13 (Continued) No.

IATA code

Airport name

Aircraft movements (thousand)

Passengers (million)

Aircraft size

Unaccommodated passengers (million)

Share of unaccommodated passengers (%)

Capacity utilisation (%)

12

DXB

Dubai

669

241

360

6.3

2.6

94.8

13

SIN

Singapore Changi

599

146

244

5.7

3.8

73.0

14

HYD

Hyderabad Rajiv Gandhi

211

51

240

5.5

9.8

42.8

15

SGN

Tan Son Nhat

430

103

240

5.3

4.9

100.0

16

MAA

Chennai

261

60

229

5.2

8.0

66.2

17

CCU

Netaji Subhas Chandra Bose

207

52

252

5.0

8.8

55.1

18

BKK

Bangkok Suvarnabhumi

547

141

257

4.3

2.9

100.0

19

SZX

Shenzhen Bao’an

547

124

227

4.1

3.2

100.0

20

SUB

Juanda

330

66

199

3.8

5.5

62.6

Traffic forecast and mitigation strategies Chapter | 10

259

reach their limits regarding their maximum level and growth rate, respectively, despite the fact that factors driving passenger demand volume growth decline slightly between 2030 and 2040, which releases some pressure from airport capacities.

10.5 Assessing mitigation strategies In discussing general strategies for mitigating airport capacity constraints in Chapter 5, General strategies for mitigating airport capacity constraints, we have stated from the outset that each constrained airport has specific capacity problems and that solutions to capacity problems vary from airport to airport and from region to region, depending on the severity and type of the capacity bottleneck, the financial situation of the airport and the regulatory framework of the region or state, which the airport operator has to take into account in planning for the future. Our intention has been to analyse, on a global level, measures and options that have been successfully applied at airports and, therefore, may be recommended for further use. On the contrary, theoretical discussions, which do not contribute to reducing the capacity problem, should not be recommended accordingly. We have seen that a whole range of technological, investment and noninvestment options does theoretically exist, and furthermore, that a spectrum of supply and demand management measures may be applied, ranging from pure administrative measures such as regulations, hybrid measures, such as slot coordination with secondary slot trading, to market-based options, such as congestion pricing schemes and primary slot trading. In Table 5.1, we have shown a typology of these mitigation measures, whereby some of them have not yet been commonly applied, such as congestion pricing, since landing fees are regulated in many countries. We have concluded that rerouting of flights to secondary and less congested airports has been regarded by airlines as an undesirable measure and has not been widely used, in particular not at hubs, because of the interrelationship of incoming and outgoing flights. Of similar nature is the temporal shifting of flights to off-peak periods at the same airport. Nevertheless, this measure has been applied by slot coordinators at Level 3 airports, when airlines requested slots at peak times. These slots, however, were only available at off-peak times. The analysis has further shown that two mitigation measures new runways as an investment option and increasing seat capacity per flight as an operational measure have proven to be more effective than the other measures. The forecasts of passenger and flight volumes have therefore incorporated these measures, in addition to the prevailing measures of utilising runways and raising load factors as much as possible when and where needed. In the following, we discuss the forecast results with an emphasis on evaluating the absolute and relative effects of employing larger aircraft here measured in passengers per flight and enhancing runway utilisation and runway capacity enlargements.

260

PART | III Forecasting future air traffic development up to 2040

10.5.1 How do increasing aircraft size and adding new runways contribute to airport capacity improvements up to 2030 and 2040? Passenger volumes on links between airports and regions were forecast in a first phase in relation to mainly socio-economic factors. Future passenger volumes are thus reflecting the market demand without being constrained by a lack of infrastructure capacity. The methodological approach of forecasting passenger flows and flight volumes between airports and regions was conceived in a following phase in such a way that forecast horizon year capacities were estimated and future airport flight volumes in particular were mirrored against capacities. The forecast results have to be interpreted therefore as ‘capacity constrained’ estimates, which do not include passenger and flight volumes that cannot be accommodated due to bottlenecks caused by airports with insufficient capacity. We show, however, the unaccommodated demand as well, in order to provide information on the amount of future capacity constraint. Depending on the shortage of capacity reserves and/or the amount of unaccommodated flights, runway utilisation was intensified up to airport-specific limits, and load factor as well as seat capacity of flights were increased up to route type-specific limits. While the measures of increasing runway utilisation and load factors contribute rather modestly to mitigating capacity problems at constrained airports due to the already high degree of runway and seat capacity utilisation, higher seat capacities of aircraft and additional runways have proven to be effective measures for overcoming capacity shortages at these airports, which are in many instances major airports, in particular hubs with the largest traffic volumes. At non-constrained airports often secondary airports and those with mainly origin destination traffic increasing runway utilisation and raising load factors of flights in addition to employing aircraft with higher seat capacity are measures of first choice and contribute to overcoming future constraint situations to a higher degree than at already congested airports. In any case, all these measures have been applied in forecasting passenger and flight volumes of airports. In the following section, we present those forecast results for 2030 and 2040 which deal with the absolute and relative capacity gain due to increasing primarily passengers per flight, on the one hand, and improving runway utilisation and enlarging runway capacity, on the other. The measures of seat capacity per flight and load factor have been taken care of implicitly in the forecast by applying the product of these variables, that is, the number of passengers per flight.

10.5.1.1 Up to 2030 As presented and commented upon in the previous chapters, the forecast of passenger and flight volumes yields among others the following global forecast results for the year 2030: G G

flight volume: 45.8 million (2016: 35.5 million) flight volume growth 2030/16: 129%, 1.8% p.a.

Traffic forecast and mitigation strategies Chapter | 10 G G G G

261

passenger volume: 6.9 billion (2016: 4.0 billion) passenger volume growth 2030/16: 176%, 4.1% p.a. unaccommodated passenger volume: 49.4 million total (unconstrained) passenger volume:6.963 billion

Regarding the mitigation measures, the following forecast results were obtained for the time period 2016 to 2030: G

G

G G

capacity gain by more passengers per flight (average aircraft size in 2016: 111 passengers per flight, in 2030: 152 passengers per flight): 1765 million passengers capacity gain by increasing runway utilisation and higher runway capacity: 1220 million passengers capacity gain share by more passengers per flight: 59% capacity gain share by increasing runway utilisation and higher runway capacity: 41%

While the number of passengers transported in the global network increases by 76% from 2016 to 2030, the flight volume grows by only 29%. The growth difference results from the fact that the average seat capacity per flight multiplied by the load factor grows from 111 to 152 passengers per flight. In 2030 almost three billion passengers more than in 2016 will use the planes, and 59% of the passenger volume growth (1.8 billion passengers) can be accounted for by higher seat capacity and higher load factors per flight, while 41% (1.2 billion passengers) are attributable to higher capacity due to more runways and higher utilisation of runways. It is, thus, interesting to note that in our forecast, the measure of deploying larger aircraft will provide more additional capacity than better utilisation of runways and new runway capacity. The improvement of these measures is also one factor responsible for just a relatively small number of unaccommodated passengers (49 million or 0.7% of the total unconstrained passenger volume) in 2030. It should be added that the forecast methodology has not been conceived so as to produce biased results by preferring one type of measure against others. The realisation probabilities of single measures are primarily based on empirical functions rather than on scenario-type assumptions. The current situation of air transport demand and traffic supply varies from country to country and from world region to world region. It is therefore advisable to not only focus on global results but also to study regional forecast outcomes. Here, we concentrate again on the capacity gain due to increasing aircraft size, which means passengers per flight, and higher runway utilisation and capacity, which means more runways. In Figs 10.10 and 10.11, the regional distribution of capacity gain is shown; in Fig. 10.10 due to more passengers per flight and in Fig. 10.11 due to higher runway utilisation and capacity. Both figures give the capacity gain in terms of additional passengers served at airports as a consequence of realising the capacity

PART | III Forecasting future air traffic development up to 2040

Million

Capacity gain by larger aircraft in annual passenger volume

262

800 700 600 500 400 300 200 100 0

Africa Africa

Asia

Asia Europe

Europe Middle East

Middle East North America

North America South America

South America

Southwest Pacific

Southwest Pacific

Million

Capacity gain by improving runway utilisation and enlargements in annual passenger volume

FIGURE 10.10 Capacity gain by employing larger aircraft (that means more passengers per flight), measured in annual passenger volume between 2016 and 2030 by region; global sum is 1765 million passengers.

800 700 600 500 400 300 200 100 0

Africa Africa

Asia

Asia Europe

Europe Middle East

Middle East North America

North America South America

South America

Southwest Pacific

Southwest Pacific

FIGURE 10.11 Capacity gain by improving runway utilisation and runway enlargements measured in annual passenger volume between 2016 and 2030 by region; global sum is 1220 million passengers.

Traffic forecast and mitigation strategies Chapter | 10

263

shortage mitigating measures. The scales of the vertical axes are the same in both figures so that the effectiveness of the two types of measures can be compared directly. As passenger volumes vary greatly between the world regions so do the capacity gains. The passenger volumes of the three major regions of Asia, Europe and North America contribute with 5.7 billion (around 82%) to the global volume of 6.9 billion passengers in 2030. The capacity gains in these major world regions are corresponding high. Asia is the region with the highest passenger volume of 2.6 billion passengers with a passenger growth between 2016 and 2030 of 1.3 billion, of which 58% are handled by larger aircraft with higher load factors and 42% by better runway utilisation and capacity enlargements at Asian airports. Passenger growth in Europe and North America is, with 586 million and 563 million passengers respectively, less than half as large as in Asia. The passenger volumes are 1.6 and 1.5 billion passengers in 2030. In Europe and North America, 62% of the passenger growth is realised through more passengers per flight and 38% through higher runway utilisation and capacity enlargements. The capacity gain caused by employing larger aircraft is, thus, in the three major regions as well as in the other regions, more effective than the measure of higher runway utilisation and more runways. While airlines are constantly enlarging their fleets with aircraft with higher seat capacity and succeed in raising load factors, major airports, in particular, have problems building new runways or even new airports. The measure of transporting more passengers per flight has proven to be in the past and will be in future the single most effective non-investment measure of mitigating airport capacity constraints. According to the forecast the average number of passengers per flight will grow from 111 in 2016 by 41 to 152 passengers per flight in 2030.

10.5.1.2 Between 2030 and 2040 We first present the main global forecast results in order to better relate the capacity gain to total passenger volumes of the year 2040. The forecast of passenger and flight volumes yields among others the following global forecast results for the year 2040: G G

G G

G G

flight volume: 52.7 million (2016: 35.5 million) flight volume growth 2040/30: 115%, 1.4% p.a.; 2040/16: 148%, 1.7% p.a. passenger volume: 9.4 billion (2016: 4.0 billion) passenger volume growth 2040/30: 136%, 3.1% p.a.; 2040/16: 1138%, 3.7% p.a. unaccommodated passenger volume: 255.5 million total (unconstrained) passenger volume: 9.630 billion

264

PART | III Forecasting future air traffic development up to 2040

Regarding the mitigation measures, the following forecast results were obtained for the time period 2030 to 2040: G

G

G G

capacity gain by more passengers per flight (average aircraft size in 2030: 152 passengers per flight, in 2040: 179 passengers per flight): 1375 million passengers (2030: 1765 million) capacity gain by increasing runway utilisation and higher runway capacity: 1079 million passengers (2030: 1220 million) capacity gain share by more passengers per flight: 56% capacity gain share by increasing runway utilisation and higher runway capacity: 44%

Global flight volume will increase further by 15% in the second forecast period from 2030 to 2040; passenger volume will grow even more by 36% to about 9.4 billion passengers. Compared with the base year 2016, the overall demand is forecast to more than double and grow by 138%. Compared with 2030, another 2.5 billion passengers will use the air transport system in 2040 and will fly in even larger aircraft; the average number of passengers per flight will continue to rise to almost 180 passengers. About 56% of the passenger growth can be attributed to the effect of larger aircraft with higher load factors and 44% to the measure of higher runway utilisation and more runway capacity. As in the previous forecast period, the capacity gain stemming from more passengers per flight will be higher than the runway utilisation and enlargements effect, however, with a relative gain of the latter effect. The number of unaccommodated passengers will grow disproportionately to more than 255 million passengers, accounting for around 2.6% of the global passenger volume. Although this share seems to be rather small due to the effectiveness of the mitigation measures applied, capacity shortages especially at major airports will become more severe. As has been shown in discussing the forecast results in detail, major airports will have much higher shares of unaccommodated passengers than secondary airports because the investment options are typically difficult to realise at high-volume airports, as has already been the case in the past. The capacity gain due to increasing aircraft size, here passengers per flight, and better runway utilisation and capacity extensions (that means more runways) by world region is shown in Figs 10.12 and 10.13. The passenger volume of the three major world regions, Asia, Europe and North America, will grow to more than 7.6 billion passengers; the volume share of these regions will slightly lose in relative importance and decrease to 81%. While the passenger volume of Asia continues to grow disproportionately high to 3.7 billion passengers, the volumes of Europe and North America will grow at a disproportionately slower pace to around two billion each. In all three regions, growth rates will be lower than in the first forecast period to 2030.

Million

Capacity gain by larger aircraft in annual passenger volume

Traffic forecast and mitigation strategies Chapter | 10

265

800 700 600 500 400 300 200 100 0

Africa Africa

Asia

Asia Europe

Europe Middle East

Middle East North America

North America South America

South America

Southwest Pacific

Southwest Pacific

Million

Capacity gain by improving runway utilisation and enlargements in annual passenger volume

FIGURE 10.12 Capacity gain by employing larger aircraft (that means more passengers per flight) measured in annual passenger volume between 2030 and 2040 by region; global sum is 1375 million passengers.

800 700 600 500 400 300 200 100 0

Africa Africa

Asia

Asia Europe

Europe Middle East

Middle East North America

North America South America

South America

Southwest Pacific

Southwest Pacific

FIGURE 10.13 Capacity gain by improving runway utilisation and enlargements measured in annual passenger volume between 2030 and 2040 by region; global sum is 1079 million passengers.

266

PART | III Forecasting future air traffic development up to 2040

Average aircraft size and the degree of capacity shortage vary in all seven regions, however, the relative mitigation effect of increasing aircraft size, measured in passengers per flight, will be stronger than the measure of utilising runways more intensively and building more runways in all regions, except in the Southwest Pacific, where the aircraft size effect at 46% is lower than the runway effect. While the aircraft size measure contributes with 56% to the total capacity constraint mitigation worldwide, this measure at 64% is highest in Africa, followed by 61% in Europe, 58% in North America, 55% in the Middle East and 54% in Asia. The reasons for the variation of these mitigation effects are most likely to be found in local and country-specific conditions of transport policy and airport circumstances, such as varying airport size and degree of airport congestion. In the following section, we take a more detailed look at the relationship between flight volume and aircraft size and the development of this relationship at airports from 2016 to 2040 in all world regions.

10.5.2 General mitigation measures in world regions The question is as follows: can we draw some general conclusions from the analysis in the global network and in world region networks and give recommendations on capacity restraint mitigation strategies? Clearly, airportspecific recommendations regarding the enlargement of capacity can be deduced only from an analysis of future needs of capacity of that specific airport. On the other hand, the driving factor of employing larger aircraft lies within the strategic planning of airlines, and this measure affects the capacity needs of airports with hardly any possibility for them to interfere and influence this process. The capacity planning of airports concentrates more on the investment option of enlarging infrastructure capacity. Since both system partners airlines and airports affect future capacity needs of airports, there may be a case for a combined airport-specific, as well as a general capacity planning, strategy. Given this hypothesis, we want to derive some general recommendations on capacity mitigation strategies. Before looking at mitigation strategies, we want to highlight the different effects of the two optional mitigation types, the aircraft size and the runway measure, on the relative capacity gain. Fig. 10.14 shows the relationships between capacity gain and aircraft size growth (including higher load factors) and enlargement of runway capacity (including higher runway utilisation by more aircraft movements). As can be seen, there is a linear relationship between capacity gain and rising aircraft size. The more passengers are carried per flight the higher is the capacity gain, with the result that the throughput of a runway increases in terms of passengers carried per hour without the need to increase runway capacity in terms of flights per hour. There is a great probability that the linear relationship does not hold forever. At some point, in the future, aircraft

Capacity gain in %

Traffic forecast and mitigation strategies Chapter | 10

267

Long-term maximum viable aircraft size

Short-term maximum viable aircraft size

Increasing number of runways / increasing aircraft size Capacity gain by adding runways Capacity gain by adding runways, weighted by realisation probability Capacity gain by increasing aircraft size

FIGURE 10.14 Comparing capacity gains achieved by increasing the number of runways at an airport and by increasing aircraft size.

capacity will most likely grow more slowly until there comes a limit of aircraft size, based on technical and/or economic reasons. In our forecast, however, we do not yet see such saturation tendencies. The pace of employing larger aircraft will slow down somewhat in the second forecast period from 2030 to 2040; however, due to the fact that some capacity bottlenecks will continue to exist and others may develop, the average size of aircraft will increase from 111 passengers per flight in 2016 to 152 in 2030 and continue to grow to 179 passengers per flight in 2040. The effect of enlarging runway capacity is quite different (see upper function). Adding a new runway offers a capacity gain of around 40 flight movements per hour, if the airport is a single-runway airport and both runways can be operated independently. The relative capacity gain amounts, in such a case, to around 100%. Since the vast majority of airports worldwide have only one runway, such an investment would yield a significant growth in throughput in terms of flights per hour. The more runways an airport has the smaller becomes the relative gain in capacity by adding another runway, as has been explained in Chapter 7, Modelling future airport capacity and capacity utilisation (see Table 7.4). For example, an airport already operating with four runways, as we find often in North America, will gain only about 20% of the capacity of a single runway by adding a new runway. The option of enlarging runway capacity loses importance with the number of runways in operation and traffic volume, because of the growing complexity of flight operations, until the marginal benefit diminishes almost completely. The runway

268

PART | III Forecasting future air traffic development up to 2040

Thousand

option is, thus, primarily a first choice of airports with only one or a few runways. We have seen in the past and can also see in the forecast that, in particular, major airports with high traffic volumes and a hub function are often confronted with the need to raise the capacity and, at the same time, have great problems pursuing such projects. Taking these circumstances into consideration, we have to admit that in a probabilistic sense the relative gain in adding capacity by new runways becomes less realistic. The effect of the lower realisation probability has been at least in an abstract form accounted for in the lower function of Fig. 10.14. For the airport population as a whole, the relative gain in capacity will diminish fast when these enlargement measures are concentrated on major airports, which already have a high runway capacity. Their chances of realising new runway capacity are in general relatively small as compared with secondary and smaller airports; for the former the option of being served with larger aircraft will offer relatively more benefits. Having discussed the different effects of the two types of mitigation measures in general, we will now identify the position of airports worldwide and in world regions with respect to average aircraft size and flight volume at these airports in 2016 and as forecast for 2030 and 2040, and derive some general mitigation strategies for airports in each world region, depending on these two characteristics. Fig. 10.15 shows the distribution of all airports

1200

‘Increase of aircraft size’ IV

Aircraft movements

1000

‘Gradual increase of aircraft size’ III

2016: 13.1% (11.0%) 2030: 1.9% (1.2%) 2040: 1.2% (0.7%)

2016: 29.2% (36.6%) 2030: 49.9% (56.0%) 2040: 55.0% (60.3%)

800

600

400

200

0

‘New runway or increase of aircraft size’ I 0

50 2016: 31.2% (21.6%) 2030: 14.6% (7.2%) 2040: 9.6% (4.0%)

100

‘New runway’ II

150 200 250 Passengers per flight (aircraft size) 2016

2030

2040

300

350

400

2016: 26.5% (30.8%) 2030: 33.6% (35.6%) 2040: 34.2% (35.0%)

FIGURE 10.15 Global classification of airports according to mitigation strategies and their share of global aircraft movements for the years 2016, 2030 and 2040 (in brackets: airports’ share of global passenger volume).

Traffic forecast and mitigation strategies Chapter | 10

269

regarding their traffic volume (vertical axis) and average aircraft size, which means passengers per flight (horizontal axis). Each point in the diagram represents aircraft movements and average aircraft size data of a specific airport for the years 2016, 2030 and 2040. The distribution of airports shows that, by far, most of the airports are situated in the low range of flight volume, which means from very small volumes to around 200,000 movements per year over the whole range of average aircraft sizes to around 400 passengers per flight. A small number of airports will, in future years, reach volumes of up to one million movements and even more with aircraft sizes between around 120 and 250 passengers per flight. The diagram is divided into four quadrants, separated by a horizontal line at a value of 200,000 aircraft movements and a vertical line at 111 passengers per flight. The volume dividing line has been selected as a practical annual service volume of a single runway. The capacity of a runway may be higher, say around 240,000 movements, but the limit of 200,000 corresponds with a service volume that can be handled with a high level of service. The vertical line represents the average aircraft size of 111 passengers in 2016. Airports in these quadrants are characterised by certain ranges of flight volume and aircraft size. Airports with small flight volumes and aircraft can be found in quadrant I. Airports with small traffic volumes, but with higher numbers of passengers per flight, are in quadrant II, whereas airports with greater flight volumes as well as aircraft size are located in quadrant III. Finally, airports with high volumes and rather small aircraft are in quadrant IV. The general mitigation strategies vary with the position of airports regarding their volume and aircraft size. While airports in quadrant I handled, in 2016, 22% of the global passenger demand and 31% of the flight volume, their relative importance will go down with the growth of traffic in future. In 2030 the airports remaining in this category will handle only 7% of the passenger demand and 15% of the traffic; this trend will continue to the year 2040. Airports in quadrant I would take full advantage of both mitigation types, that is increasing aircraft size as well as runway capacity, provided that they have the choice of realising these options. As mentioned above, the investment option can be proactively pursued by airport operators, whereas the aircraft size option depends on the fleet and network strategy of airlines serving the airport. The percentage of airports of this type with capacity problems is compared with larger airports rather small. Airports in quadrant II handle larger aircraft, however, rather small traffic volumes. These airports will gain in importance; their traffic share will rise from 27% in 2016 to 34% in 2030, while their passenger volume share will rise from 31% to 36%. In the following period, from 2030 to 2040, their traffic share will not further increase. About one-third of the total traffic will thus be handled by airports in this quadrant. These airports would primarily benefit from enlarging runway capacity. A new runway would double the capacity if the new runway could be realised so that flight operations on

270

PART | III Forecasting future air traffic development up to 2040

each runway can be managed independently from each other. Increasing aircraft size would be a secondary option, especially for those airports with average aircraft sizes in the range of 120 200 passengers per flight. Airports with higher traffic volumes and large average aircraft size belong to a higher percentage to those with capacity problems. Airports in quadrant III form this category and mitigation strategies are often needed here more than at airports in other quadrants. In 2016 airports of this category handled a traffic share of 29%, and in 2030 their share will rise to 50%. Their importance will continue to grow until 2040, when they will handle a traffic share of 55%. The passenger volume share goes up correspondingly from 37% to 60% in 2040. This group of airports will be, thus, the most important group. Most of the airports handle aircraft ranging in their size between 120 and 230 passengers per flight; hence, a gradual increase of aircraft size might be the most beneficial strategy to cope with capacity shortages. For airports with traffic volumes of below around 400,000 movements per year, an investment in a new runway may be an option, too, especially if their average aircraft size already exceeds around 230 passengers per flight. For airports with very high traffic volumes, new runways provide a relatively small additional value. A solution may rather be to open a new airport in the same agglomeration, as is the case in Beijing. In such a case the full capacity gain may be realised by the additional runway system. The airports in quadrant IV handle high traffic volumes with aircraft of rather small size of below 111 passengers per flight. There are only relatively few airports of this type, and as can be seen in Fig. 10.15, their traffic share was very small in 2016 and is diminishing in future. Their first mitigation strategy would be to handle more and more aircraft of larger size, thereby increasing the passenger throughput per runway without raising the flight volume correspondingly. Since the occurrence of airports of this type will be negligible in future, mitigation strategies would not really contribute to solving the global capacity problem. As we have seen, airports with high traffic volumes, as well as aircraft with many passengers on board (in quadrant III), represent both the most important group as measured by their traffic share of the global traffic, as well as the most problematic ones, since they have the highest share of airports with highly utilised infrastructure with partly severe capacity constraints. Solutions to mitigate the capacity problem vary, certainly from airport to airport. As a general strategy, we would propose, first, a further increase of aircraft size and, second, new runway capacity for airports with a small number of runways and, for airports with a complex runway system, new capacity in the form of a new airport nearby. We have to state, however, that for a series of airports of this group the proposed strategies may be out of the question due to local and political constraints. The second most important airport group consists of airports with traffic volumes below 200,000 annual aircraft movements and aircraft with a size

Traffic forecast and mitigation strategies Chapter | 10

271

Thousand

exceeding 111 passengers per flight (in quadrant II). The general mitigation strategy would be an investment into additional runway capacity and, only second, for airports with relatively small aircraft of up to 200 passengers per flight, more flights with more passenger capacity. Here again, new runways may be not a feasible solution in many instances for airport-specific reasons, and airports have to rely on non-investment solutions such as a growth of average aircraft size. After examination of the global development, we now analyse airport strategies in world regions since airport sizes, developmental conditions and traffic growth patterns vary from region to region. In addition, the mitigation strategies proposed may be better tailored to the needs of airports in each region. Africa is a continent with a rather low air traffic volume of nearly 1.1 million flights on almost 390 airports in 2016. This is forecast to grow above average to more than 1.4 million flights in 2030 and further to 1.6 million flights in 2040. Fig. 10.16 shows the distribution of African airports regarding their aircraft movements and passengers per flight numbers in 2016, 2030 and 2040. Quite in contrast to the airport size structure in the global network, 60% of the traffic of all airports in Africa has been handled in 2016 by small airports with low traffic volumes and numbers of passengers per flight below the global average of 111 passengers (in quadrant I), with a passenger share

1200 ‘Gradually increase aircraft size’ III

‘Increase aircraft size’ IV 1000 2016: 9.8% (11.2%)

2016: 0.0% (0.0%) 2030: 16.8% (18.6%) 2040: 17.8% (18.7%)

Aircraft movements

2030: 0.0% (0.0%) 2040: 0.0% (0.0%)

800

600

400

JNB

200

0

‘New runway or increase of aircraft size’ NBO I 0

50

2016: 59.8% (48.7%) 2030: 23.0% (11.9%) 2040: 16.1% (7.2%)

100

‘New runway’ II

150 200 250 Passengers per flight (aircraft size)

2016

2030

2040

300

350

400

2016: 30.4% (40.1%) 2030: 60.2% (69.5%) 2040: 66.1% (74.1%)

FIGURE 10.16 Classification of African airports according to mitigation strategies and their share of African aircraft movements for the years 2016, 2030 and 2040 (in brackets: airports’ share of passenger volume at African airports).

272

PART | III Forecasting future air traffic development up to 2040

of almost 50%. Due to the traffic growth foreseen, airports will partly move into quadrant II and handle more aircraft movements with more passengers per flight. In 2016 30% of the traffic in Africa was handled by airports in this group. This share will double in 2030 and further grow to 66% in 2040, when these airports will handle almost three quarters of the passenger volume of Africa. Airports with high traffic volumes and high numbers of passengers per flight in quadrant III will gain in importance as well but will not reach a traffic share of all airports of more than 18%. The traffic importance of African airports in quadrant IV with high traffic volumes, but with small aircraft is very small and negligible in future. For airports with capacity constraints in Africa, which are in quadrant I, mitigation strategies could be new runways and/or increasing aircraft size, since most of these airports are rather small, most likely with one runway, and being served with aircraft of low seat capacity. Two example airports are depicted in Fig. 10.16, Johannesburg (JNB) and Nairobi (NBO); JNB is the largest airport in Africa with slightly more than 200,000 passengers in 2016 and equipped with two parallel runways, whereas NBO is a mediumsized airport with around 100,000 passengers and one runway. Both airports can grow on existing infrastructure without facing greater capacity problems in future. Passenger growth can be handled by aircraft growing in size correspondingly. The average number of passengers per flight in Africa was 96 in 2016; this figure is expected to grow to 134 in 2030, which is still well below the global average of 152 passengers. In 2016 Asia handled a traffic volume of 10.4 million flights on more than 920 airports, carrying 1.3 billion passengers. Each flight had on average 130 passengers on board, almost 20 passengers more than the global average. Asia has not only the highest traffic share of all regions but is also the region with the highest passenger volume growth of 4.9% p.a. between 2016 and 2030 so that the number of passengers will grow to more than 2.6 billion on 14.6 million flights. Demand will continue to grow until 2040 to almost 3.7 billion passengers. Outstanding examples of traffic size and growth are the two hub airports of Beijing, Capital City (PEK) and Daxing (PKX), which are shown among others in Fig. 10.17. Most of the airports in Asia are larger than those in Africa, and especially small airports in quadrant I play only a marginal role in handling traffic in future. The traffic share of airports with small traffic volumes and aircraft with high numbers of passengers per flight (in quadrant II) will decrease from 42% in 2016 to 39% in 2030 and further on to 35% in 2040, whereas the traffic share of airports with both high traffic volumes and flights with many passengers on board (in quadrant III) will grow to 56% in 2030 and to 62% in 2040. Airports in Asia are typically confronted with high traffic growth rates, as can be particularly seen in the example of the largest Asian airport Beijing Capital City (PEK) in 2016. The traffic will almost double from more than 600,000 aircraft movements in 2016 to almost 1.2 million in

Thousand

Traffic forecast and mitigation strategies Chapter | 10 1200 ‘Increase aircraft size’ IV 1000

Aircraft movements

273

‘Gradually increase aircraft size’ III 2016: 41.2% (47.1%) 2030: 56.4% (60.8%) 2040: 62.2% (65.5%)

2016: 0.0% (0.0%) 2030: 0.0% (0.0%) 2040: 0.0% (0.0%)

800 PEK & PKX

600

400

200

0

‘New runway or ITM increase of aircraft size, I 0

50 2016: 16.4% (11.0%) 2030: 4.4% (2.0%) 2040: 2.5% (0.9%)

100

CJU

‘New runway’ II

150 200 250 Passengers per flight (aircraft size) 2016

2030

2040

300

350

400

2016: 42.4% (41.9%) 2030: 39.2% (37.2%) 2040: 35.3% (33.6%)

FIGURE 10.17 Classification of Asian airports according to mitigation strategies and their share of Asian aircraft movements for the years 2016, 2030 and 2040 (in brackets: airports’ share of passenger volume at Asian airports).

2040, and most of the growth is accomplished by a new airport with sufficient runway capacity. Average aircraft size will also grow by 87, from 159 passengers to 246 passengers per flight, in the same period. PEK is an outstanding example of a large airport where a first choice strategy to mitigate capacity shortages would be to increase aircraft size further but, if possible, create new runway capacity by building a new airport (PKX). Given the strong growth of traffic in Asia, we would recommend for airports with limited runway capacity in quadrant II a new runway or two, and for airports with more capacity available (in quadrant III) a further increase of aircraft size. In both cases, however, we would not exclude the alternative strategy, which is larger aircraft for airports in quadrant II and new runways for airports in quadrant III as a secondary option. For airports with multiple runways, we do not recommend as a first choice new runway capacity on the same airport, but instead on a secondary airport in the same agglomeration, as the example of Beijing shows. Two other examples are shown in Fig. 10.17, the development of more typical airports of Osaka (ITM) and Jeju (CJU). Both airports have traffic volumes below the practical capacity of a single runway of 200,000 movements and participate in the traffic growth in Asia. Both airports can still handle the growing traffic by increasing average aircraft size without a need to realise an investment option. If capacity constraints become a problem, we would recommend a runway capacity extension as a first option.

274

PART | III Forecasting future air traffic development up to 2040

Thousand

Europe is an important traffic region, too, with 685 airports and a traffic volume of 8.3 million flights in 2016. Some 980 million passengers were carried on these flights, which had an average aircraft size of 118 passengers. European airports are similarly distributed as Asian airports regarding traffic volume and average aircraft size. However, Europe does not have an airport as large as Beijing Capital City (PEK). The distribution and development of airports in Europe is shown in Fig. 10.18. Traffic at European airports was roughly equally distributed in 2016 over the three quadrants I III with about 30% each in 2016, although the passenger demand share grew from 21% at smaller airports to 34% at airports with volumes of less than 200,000 aircraft movements per year in quadrant II and to 40% at the large airports with more traffic and larger aircraft in quadrant III. Traffic share at small airports will diminish in future, while airports in quadrants II and III will take over and handle more and more traffic. Large airports with more flights of larger aircraft size will handle even more than airports in quadrant II, which still have rather small traffic volumes. Airports struggling with capacity constraints are mainly in quadrant III, such as two of the example airports: London Heathrow (LHR) and London Gatwick (LGW). Both airports are in need of substantial capacity growth, however, have not yet succeeded in realising new runway projects due to public resistance. Their traffic and aircraft size developments are shown in Fig. 10.18. According to the forecast, LHR will grow almost exclusively by aircraft size, while LGW will get a capacity extension and grow by traffic 1200

1000

Aircraft movements

‘Gradually increase aircraft size’ III

‘Increase aircraft size’ IV

2016: 35.4% (40.3%) 2030: 47.8% (52.2%) 2040: 51.3% (55.2%)

2016: 5.5% (4.9%) 2030: 1.0% (0.6%) 2040: 1.0% (0.6%)

800

600 FRA

400

LHR

LGW

200

0

‘New runway’ II

GVA

‘New runway or STR increase of aircraft size’ I 0

50

2016: 29.6% (21.2%) 2030: 11.1% (5.6%) 2040: 6.6% (2.6%)

100

150 200 250 Passengers per flight (aircraft size) 2016

2030

2040

300

350

400

2016: 29.5% (33.6%) 2030: 40.1% (41.6%) 2040: 41.2% (41.6%)

FIGURE 10.18 Classification of European airports according to mitigation strategies and their share of European aircraft movements for the years 2016, 2030 and 2040 (in brackets: airports’ share of passenger volume at European airports).

Traffic forecast and mitigation strategies Chapter | 10

275

volume as well as by aircraft size. With four runways, Frankfurt airport still has capacity reserves and can grow by traffic volume and aircraft size without additional runway capacity. Two other example airports of smaller size are depicted in Fig. 10.18, namely Geneva (GVA) and Stuttgart (STR), each with one runway. Both airports will grow without new runways in the forecast period by handling more aircraft with more passengers on board. Although the general recommendations derived on the global level apply to Europe as well, airports with capacity constraints in Europe have, in many instances, severe problems with enlarging capacity. We would recommend new runways for airports with capacity bottlenecks in quadrant II, but we admit, however, that the realisation chances are not great, if at all. This means that airports in Europe have primarily to rely on the growth of aircraft size in order to cope with further traffic growth. For capacity constrained airports with high traffic volumes and aircraft size (in quadrant III), we would recommend, first, a further increase of aircraft size. However, we would also recommend capacity growth, knowing the problems associated with investment proposals for European airports. It seems that airports in Europe of this type, and other types as well, are in a particularly difficult position to realise plans to add runway capacity because the population living around airports and local politics oppose such proposals, often successfully, on grounds of nuisance, noise and gaseous emissions. The Middle East region is, with 111 airports and a flight volume of 1.2 million, one of the smaller air traffic regions. As can be seen in Fig. 10.19, Dubai airport (DXB) stands out as a prominent airport, while most others are smaller and can be found mainly in quadrants II and III. While the relative traffic importance of smaller airports in quadrant II will remain the same with around 45% of all aircraft movements in the Middle East but with a declining importance of passenger demand, larger airports in quadrant III will take over and increase their share of traffic volume from 36% to 53% in 2040. As a general strategy for capacity constrained airports here, we would recommend raising the capacity by adding new runways, since average aircraft size is already above average in many instances. According to the forecast, DXB will grow from around 400,000 aircraft movements in 2016 to almost 670,000 movements in 2040. This growth will be realised through both new runway capacity as well as larger aircraft size. Average aircraft size will thereby increase further by about 140 passengers per flight, from around 220 to 360, a size reached by just a few large airports worldwide. Riyadh airport (RUH) has less than 200,000 movements and is only half as large as DXB. Here, traffic is expected to grow to almost 300,000 movements in 2040. To accomplish the traffic growth the airport will use both existing runways more intensively with aircraft which will grow in size from about 140 to around 220 passengers per flight. In 2016 North America had more than 1000 airports and a flight volume of ten million flights. It was almost as important a traffic region as Asia.

Thousand

276

PART | III Forecasting future air traffic development up to 2040 1200

‘Increase aircraft size’ IV

1000

Aircraft movements

‘Gradually increase aircraft size’ III

2016: 36.2% (46.6%) 2030: 51.4% (61.9%) 2040: 53.1% (62.2%)

2016: 0.0% (0.0%) 2030: 0.0% (0.0%) 2040: 0.0% (0.0%)

800

600

400

DXB

200

0

‘New runway or increase of aircraft size’ I 0

50 2016: 18.6% (11.9%) 2030: 1.3% (0.5%) 2040: 0.5% (0.2%)

100

RUH

‘New runway’ II

150 200 250 Passengers per flight (aircraft size) 2016

2030

2040

300

350

400

2016: 45.2% (41.5%) 2030: 47.3% (37.6%) 2040: 46.4% (37.6%)

FIGURE 10.19 Classification of airports of the Middle East according to mitigation strategies and their share of aircraft movements at airports of the Middle East for the years 2016, 2030 and 2040 (in brackets: airports’ share of passenger volume at airports of the Middle East).

The passenger volume was greater in Asia with 1.3 billion passengers, more than in North America which had 0.9 billion passengers. The reason for the difference is the average aircraft size; on Asian flights, average aircraft size is highest with 130 passengers per flight, while in North America, only 91 passengers are on board, the lowest value of all regions. North America has mainly domestic US traffic with frequent services with rather small aircraft, whereas the Asian network is served with more international and medium to long range flights with higher seat capacity. The distribution and development of airports regarding their traffic volume and aircraft size in North America is shown in Fig. 10.20. Quite in contrast to the size distribution of airports in other regions, average aircraft size of North American airports varies mainly between 70 and 200 passengers per flight as can be seen in Fig. 10.20. North America also has a higher share of high-volume airports, with Atlanta Hartsfield Jackson (ATL) and Chicago O’Hare (ORD) reaching a threshold of almost one million aircraft movements in 2030 and 2040. While, for the time being, about three-quarters of the traffic in North America is handled by below-average sized aircraft at small- and large-volume airports, traffic volumes and average aircraft size will grow in future so that in 2040 almost two-thirds of the traffic is concentrated on larger airports in quadrant III. For the capacity constrained airports among them, especially for those with complex runway systems, we would recommend a further increase of aircraft size as the more

Thousand

Traffic forecast and mitigation strategies Chapter | 10 1200 ‘Gradually increase aircraft size’ III

‘Increase aircraft size’ IV

1000

Aircraft movements

277

2016: 38.6% (39.0%) 2030: 6.1% (5.2%) 2040: 3.6% (2.7%)

800

2016: 20.3% (27.3%) 2030: 58.4% (67.3%) 2040: 63.7% (72.3%)

ATL

DFW

600

400

200

0

"New runway or increase of aircraft size" I 0

50

2016: 37.6% (29.0%) 2030: 24.7% (15.5%) 2040: 17.7% (9.9%)

100

SAN

‘New runway’ II 150 200 250 Passengers per flight (aircraft size) 2016

2030

2040

300

350

400

2016: 3.5% (4.7%) 2030: 10.8% (12.0%) 2040: 15.0% (15.1%)

FIGURE 10.20 Classification of North American airports according to mitigation strategies and their share of North American aircraft movements for the years 2016, 2030 and 2040 (in brackets: airports’ share of passenger volume at North American airports).

efficient strategy, thus improving runway utilisation rather than raising infrastructure capacity. For smaller airports with capacity problems, infrastructure enlargements may be the better choice, if such options are feasible. The development of aircraft size and traffic volume is shown in Fig. 10.20 for three example airports: Atlanta Hartsfield Jackson (ATL), Dallas/Fort Worth (DFW) and San Diego (SAN). While ATL and DFW are high-capacity airports with five and seven runways, respectively, SAN belongs to the 30 core airports of the United States, but with one runway only. All three airports have high capacity utilisation rates, whereby ATL is, with about 875,000 movements, operating at capacity level. Traffic volumes will increase in future, most of all in DFW and least in SAN. ATL will enlarge capacity by a new runway in order to cope with demand growth, although one would propose as a general rule for airports with complex runway systems a higher runway utilisation through larger aircraft size. ATL has five parallel runways, and adding another parallel runway would still bring a relatively great capacity increase. DFW has no capacity shortage to handle the future flight volume, while SAN, on the contrary, has no possibility of raising capacity but has to rely on larger aircraft size. South America has 530 airports and a traffic volume of 3.3 million flights in 2016. It is one of the smaller world regions. Traffic is expected to grow to 4.2 million flights in 2030 and to almost five million in 2040. Average aircraft size, with 99 passengers per flight in 2016, is below the global average of 111 and will remain below average in future with values of 133

Thousand

278

PART | III Forecasting future air traffic development up to 2040 1200

1000

Aircraft movements

‘Gradually increase aircraft size’ III

‘Increase aircraft size’ IV

2016: 8.2% (10.8%) 2030: 20.8% (24.1%) 2040: 30.9% (36.0%)

2016: 6.4% (6.3%) 2030: 0.0% (0.0%) 2040: 0.0% (0.0%)

800

600 MEX

400

200

0

‘New runway or increase of aircraft size’ SDU I 0

50

2016: 51.8% (39.0%) 2030: 25.2% (13.0%) 2040: 13.9% (4.9%)

100

‘New runway’ II

150 200 250 Passengers per flight (aircraft size)

2016

2030

2040

300

350

400

2016: 33.6% (43.9%) 2030: 54.0% (62.9%) 2040: 55.2% (59.1%)

FIGURE 10.21 Classification of South American airports according to mitigation strategies and their share of South American aircraft movements for the years 2016, 2030 and 2040 (in brackets: airports’ share of passenger volume at South American airports).

passengers in 2030 and 158 in 2040. The development of aircraft size and flight volume of each airport from 2016 to 2030 and 2040 is shown in Fig. 10.21. Most airports in South America have traffic volumes below 200,000 aircraft movements, volumes which can be handled by a single runway. Due to traffic growth the traffic share of small airports with average aircraft size below the average of 111 passengers per flight (in quadrant I) will go down from 52% to 14% in 2040, while the traffic share of airports with larger aircraft size (in quadrant II) will increase from 34% to 55% in 2040, and the traffic share of airports with both larger traffic volumes and aircraft size (in quadrant III) will rise from 8% to 31%. This means, on the other hand, that almost 70% of the South American traffic will be handled in future by rather small-volume airports. For those of them with capacity problems, a general mitigation measure of first choice would be to add another runway and, secondly, to increase runway utilisation by larger average aircraft size. The outstanding airport in South America is Mexico City (MEX) with more than 426,000 aircraft movements in 2016, which is expected to grow to around 681,000 movements in 2040 (see Fig. 10.21). The airport has been operating at a high capacity utilisation rate for years. To accomplish the future growth, MEX will need a substantial capacity enlargement, which is forecast to happen probably and better so by a secondary airport. Therefore the passenger growth will be realised by additional flights rather

Traffic forecast and mitigation strategies Chapter | 10

279

than by an increase of average aircraft size. It will grow from 97 in 2016 to around 143 passengers per flight in 2040, a value still below the South American average of 158 passengers. The long-term growth potential lies therefore in a stronger rise of average aircraft size. The metropolitan area Rio de Janeiro airport has two airports, Rio de Janeiro-Galeao (GIG) as an international airport and Rio de Janeiro Santos Dumont (SDU) for domestic traffic, primarily to Sao Paulo. SDU has a single runway and a traffic volume of around 100,000 aircraft movements (see Fig. 10.21). Traffic is expected to grow to 167,000 movements in 2040, which means that a new runway is not needed. The airport will handle the growing passenger demand mainly by increasing average aircraft size from around 100 passengers per flight to nearly 150 passengers. We see, in both airport examples, that the measure of enlarging infrastructure capacity will not be the solution to cope with growing passenger demand, but rather the option of increasing runway utilisation by larger aircraft size. The Southwest Pacific region with Australia as the main contributor is also one of the smaller traffic regions, with 401 airports and a traffic volume of 1.1 million flights in 2016. According to the forecast, traffic will grow to 1.5 million flights in 2030 and to 1.7 million in 2040. Average aircraft size in 2016 was 97 passengers per flight and thus 14 below the world average of 111, caused mainly by the small aircraft size on about one million domestic flights. This main measure of mitigating capacity shortages grows to about 128 passengers per flight in 2030 and to 148 in 2040 and remains, thereby, below the global average. The development of average aircraft size and traffic volume of each airport in the Southwest Pacific is shown in Fig. 10.22. As in other regions, most of the traffic in the Southwest Pacific is handled by small-volume airports of category I and II; in 2016 almost two-thirds of all flights operated from these airports. Airports with both small volumes and average aircraft size, in particular, will lose traffic in future, and airports with more passengers per flight and with larger volumes will gain correspondingly. Airports of categories II and III handled about half of the traffic in 2016; this share will rise to two-thirds in 2030 and to more than 70% in 2040. Since average aircraft size will not exceed 150 passengers per flight until 2040, and many airports handle traffic volumes of below 400,000 aircraft movements, both mitigation options may be potential candidates for airports with capacity problems, depending on the local situation regarding infrastructure enlargement possibilities. Airports with traffic volumes of below 200,000 aircraft movements would rather prefer a new runway in the case of a need for further runway capacity, while larger airports might look for more flights with more passengers on board, to improve runway utilisation further. Two example airports have been selected, the developments of which are shown in Fig. 10.22. Sydney (SYD) is the airport with the highest traffic volume (about 334,000 aircraft movements in 2016) in the Southwest Pacific

Thousand

280

PART | III Forecasting future air traffic development up to 2040 1200 ‘Gradually increase aircraft size’ III

‘Increase aircraft size’ IV 1000 2016: 0.0% (0.0%)

2016: 34.1% (45.3%) 2030: 42.2% (54.1%) 2040: 43.8% (54.1%)

Aircraft movements

2030: 0.0% (0.0%) 2040: 0.0% (0.0%)

800

600

400

200

0

SYD

‘New runway or increase of aircraft size’ADL I 0

50 2016: 48.0% (31.1%) 2030: 33.1% (17.3%) 2040: 28.7% (14.9%)

100

‘New runway’ II

150 200 250 Passengers per flight (aircraft size)

2016

2030

2040

300

350

400

2016: 17.9% (23.6%) 2030: 24.7% (28.6%) 2040: 27.5% (31.0%)

FIGURE 10.22 Classification of airports of the Southwest Pacific according to mitigation strategies and their share of aircraft movements at airports of the Southwest Pacific for the years 2016, 2030 and 2040 (in brackets: airports’ share of passenger volume at airports of the Southwest Pacific).

region. Adelaide (ADL) represents rather an airport of typical size with almost 82,000 movements and is equipped with two crossing runways. It has no capacity problems. Since average aircraft size with less than 100 passengers per flight is rather small, future passenger volume growth can be handled first of all by larger aircraft size and in addition by more flights. SYD has three runways and a capacity of around 550,000 aircraft movements, which is sufficient to handle the growing flight volume in the forecast period until 2040. Therefore no additional runway capacity has been forecast. With an additional 200,000 aircraft movements and an increase of average aircraft size from 128 to 183 passengers per flight, the airport will be in a position to serve the growing passenger demand until 2040. At that time, however, the airport will operate at capacity level, and an enlargement of capacity seems to be needed then. In discussing the airport situation in world regions, we have seen that the share of congested airports as well as the distribution and future development of traffic volumes and average aircraft size of airports vary substantially between regions. This means that mitigation strategies of airports coping with capacity shortfalls differ correspondingly in regions. Africa has a great number of small-volume airports today and in future, with a great range of average aircraft size, which in most cases do not have major capacity problems. With growing passenger demand, traffic volumes at airports will grow as well; however, airports will not reach volumes of more than

Traffic forecast and mitigation strategies Chapter | 10

281

200,000 aircraft movements in most instances. Options to react to capacity constraints would be, first, new runways and, second, more passengers per flight. Asia is, in contrast, the most important traffic region worldwide and has a high share of airports with large traffic volumes and average aircraft size. More than 40% of the traffic of Asia is handled by these airports today, which will increase to more than 60% in 2040. Twenty-two of the top 50 airports worldwide with high capacity utilisation are located in Asia (see Table 7.1). In addition, Asian airports have to cope with high traffic growth rates. For airports with smaller traffic volumes, but with large average aircraft size, which need additional capacity, we would first recommend new runways, whereas for the larger airports with larger aircraft size, a further increase of average aircraft size may be more appropriate. Given the strong growth of passenger demand, we know, however, that the latter may need additional runway capacity as well. Traffic of European airports is fairly evenly distributed over smaller airports with low average aircraft size, airports with small traffic volumes and larger aircraft size, and airports with higher volumes with more passengers on board. Only the latter two groups of airports will gain in importance in future. Although traffic growth will be smaller in Europe than in Asia and North America, Europe has been and will remain a region where airports with capacity constraints face great problems of enhancing runway capacity by new infrastructure. Eleven of the 50 top airports worldwide with high capacity utilisation rates can be found in Europe. Improving runway utilisation by increasing average aircraft size will probably be the main mitigation strategy. Air traffic volumes of North American airports vary over a wide range, but most of them handle rather small aircraft; more than 70% of the traffic is handled by airports with below-average aircraft size. This will change in future. In 2030 almost 60% of the traffic will be handled by large airports with aircraft with high seat capacity. A great number of airports with high traffic volumes have more or less capacity constraints, and thirteen of the top 50 airports in terms of high capacity utilisation worldwide in 2016 are in North America. For these airports, we would strongly recommend, to cope with the growing demand, increasing average aircraft size. Adding new runways at airports with complex runway systems is most likely not the most efficient strategy. The three remaining regions of the Middle East, South America and the Southwest Pacific are traffic regions with fewer airports and with smaller flight volumes than the three large regions of Asia, North America and Europe. Most of the traffic is handled by rather small-volume airports with a wide range of average aircraft size, in the Middle East and the Southwest Pacific more than 60% and in South America even more than 80%. The largest airports of these regions are Dubai (DXB), Mexico City (MEX) and

282

PART | III Forecasting future air traffic development up to 2040

Sydney (SYD), which will be able to grow by means of a new airport and, as in the case of SYD, by improving runway utilisation by more aircraft and higher average aircraft size. For the other airports with capacity problems, we would recommend new runway infrastructure as well; however, a better utilisation of runways by larger aircraft should always be a strategy to follow if appropriate.

10.6 Case study: traffic forecast for San Diego (SAN), London Heathrow (LHR), Beijing Capital City (PEK) and Beijing Daxing (PKX) This is the final part of the case study of the three example airports San Diego, London Heathrow and Beijing Capital City, which now includes a fourth airport, the second fully fledged hub airport Beijing Daxing (PKX). It opened in September 2019 in the south of Beijing and started with four runways and is expected to be equipped with up to eight runways for civil aviation. Nevertheless, the two hub airports taken as a whole and treated as a whole in this case study will have ample capacity reserves for future growth of passenger and flight volume in the Beijing region. As in the preceding sections, we present results for 2030 and 2040 regarding passenger volume, number of aircraft movements, aircraft size and capacity gain. The purpose is to demonstrate the forecast capability of the models of Part II of the book on the airport level for the example airports. We have included an additional optimistic scenario with a third runway in LHR as proposed by the Airports Commission (2015) to solve the capacity crunch. In this case the maximum number of aircraft movements that LHR can handle per year could increase by about 260,000 to around 740,000 aircraft movements, as already discussed in Chapter 5, General strategies for mitigating airport capacity constraints (Airports Commission, 2015). This value is close to our own estimate of the airport capacity model of Chapter 7, Modelling future airport capacity and capacity utilisation, which forecasts a 50% increase of airport capacity given the runway configuration, which means to almost 785,000 aircraft movements per year. But, for the analyses to follow, we keep the value of 740,000 aircraft movements per year of the Airports Commission (2015), which has assumed that the new runway will be fully available no later than 2030 so that there is no demand lost because of a capacity shortage at LHR (see Chapter 8: Modelling future airport capacity enlargements and limits and especially Fig. 8.3 for more details on demand accumulation). Nevertheless, we still expect the status quo to remain at LHR, which means the current runway system to be the more likely case for the years 2030 and 2040. However, we assume that the movement cap will be dropped so that the airport can handle about 523,000 aircraft movements annually, which is a capacity increase of almost 10%. Eventually, we have to take note of the fact that capacity realisation forecasts, especially at LHR,

Traffic forecast and mitigation strategies Chapter | 10

283

are inherently difficult. For SAN we do not expect an enlargement of runway capacity, as it is sufficient to handle the forecast traffic volumes of 2030 and 2040.

10.6.1 Passenger volume

Million

Fig. 10.23 shows the annual passenger volume of 2016 and the forecast years 2030 and 2040 for the airports LHR, PEK and PKX, and SAN. The annual growth rates (CAGR) for the time periods 2016 30 and 2030 40 are displayed on top of the columns for 2030 and 2040, respectively. About 76 million passengers were handled at LHR in 2016, while around 117 and 144 million passengers per year are forecast for 2030 and 2040, respectively, in the case of two runways. This corresponds to a CAGR of 3.1% for the time period 2016 30 and 2.1% for 2030 40. Overall, passenger volume rises by almost 90% between 2016 and 2040, which is quite amazing given the current capacity situation at LHR. However, the prevailing decreasing growth rate of passenger volume essentially arises from an ever-expanding capacity shortage at LHR, while the demand drivers are more or less stable (see Section 10.2). In the case of an additional runway, passenger volume at LHR could even rise to more than 126 million and 168 million passengers in 2030 and 2040, respectively. This corresponds to a 350

Annual passenger volume

300

3.7%

250 5.4%

200 2.9% 2.1%

150 3.1%

3.7%

100 50 0

3.5%

LHR two runways

LHR three runways 2016

PEK and PKX 2030

2.8%

SAN

2040

FIGURE 10.23 Actual and forecast annual passenger volume for London Heathrow (LHR) with two and three runways, Beijing Capital City (PEK) and Beijing Daxing (PKX), and San Diego (SAN) for the years 2016, 2030 and 2040 (values above the columns for 2030 and 2040 represent annual growth rates (CAGR) for the time periods 2016 30 and 2030 40). CAGR, Compound annual growth rate.

284

PART | III Forecasting future air traffic development up to 2040

CAGR of 3.7% for the time period 2016 30 and 2.9% for 2030 40. In total, passenger volume could even increase by 121% between 2016 and 2040, if a third runway were available at LHR. More than 96 million passengers arrived at or departed from PEK in 2016 and for 2030 and 2040, 202 million and 290 million passengers, respectively, are expected to be handled at the two hub airports of Beijing PEK and PKX. This corresponds to a CAGR of 5.4% for the time period 2016 30 and 3.7% for 2030 40 and, thus, passenger volume will triple between 2016 and 2040. The growth rate slows down substantially over time, because economic development in China is expected to slow down as well in the long term (see Section 10.2). However, growth rates of passenger volume as well as economic development are still significantly higher than in more mature economies such as Europe or North America. Direct capacity constraints play no role because of the new airport PKX; however, there is some unaccommodated demand volume as a result of indirect capacity constraints. Twenty-one million passengers were handled at the single-runway airport SAN in 2016, and passenger volume is forecast to increase to almost 34 million and more than 44 million passengers in 2030 and 2040, respectively. This corresponds to a CAGR of 3.5% for the time period 2016 30 and 2.8% for 2030 40, and passenger volume therefore grows by about 112% between 2016 and 2040. There is no significant capacity shortage expected until 2040 because of a trend to larger aircraft that means more passengers per flight. However, further economic development mirrors that of a mature economy, which limits further growth potential of passenger volume. Fig. 10.24 displays the forecast unaccommodated annual passenger volume for 2030 and 2040 by airport. The share of unaccommodated passenger volume, which is related to the unconstrained passenger volume, is displayed on top of the columns for 2030 and 2040. In the case that the current runway system will not be extended, about 8.6 million passengers, corresponding to 6.8% of the forecast unconstrained passenger volume of London Heathrow (LHR), cannot be served in 2030 due to a capacity shortage. For 2040 this value increases to nearly 25.2 million passengers, equalling almost 15% of the unconstrained passenger demand volume of LHR. However, if a third runway were to be realised, unaccommodated passenger volume could be reduced to about 340,000 in 2030 and 2.4 million passengers in 2040, which is 0.3% and 1.4% of the unconstrained passenger volume forecast. Any unaccommodated demand is due to capacity constraints at other airports, as there would be sufficient capacity at LHR up to 2040. However, LHR will be almost at its capacity limit of 740,000 aircraft movements per year in 2040. Unaccommodated passenger demand at PEK and PKX and SAN is rather small and caused by indirect capacity constraints. Due to the new airport, Beijing Daxing (PKX), there is a massive capacity increase for the Beijing

Million

Unaccommodated annual passenger volume

Traffic forecast and mitigation strategies Chapter | 10

285

30 14.9%

25

20

15

10

6.8%

5 1.4%

0

0.3%

0.3%

LHR two runways

LHR three runways 2030

0.6%

PEK and PKX

0.6%

1.6%

SAN

2040

FIGURE 10.24 Forecast unaccommodated annual passenger volume for London Heathrow (LHR) with two and three runways, Beijing Capital City (PEK) and Beijing Daxing (PKX), and San Diego (SAN) for the years 2030 and 2040 (values above the columns for 2030 and 2040 represent the share of unaccommodated demand relative to the unconstrained demand, that means forecast passenger volume plus unaccommodated demand).

region and, thus, just between 633,000 (0.3%) and 1.7 million (0.6%) passengers of the unconstrained passenger volume cannot be served, but mainly because of capacity constraints at other Asian airports. SAN has ample capacity reserves in 2016 but will reach its capacity limit in 2040. Around 189,000 passengers (0.6%) in 2030 and 736,000 passengers (1.6%) in 2040 cannot be handled because of a capacity shortage at destination airports. While the level of unaccommodated demand is relatively low at these three airports, capacity constraints are becoming increasingly important, as the trend of the share of unaccommodated demand illustrates: it doubles for the Beijing airports and almost triples for SAN.

10.6.2 Number of aircraft movements Fig. 10.25 displays the forecast number of aircraft movements for 2030 and 2040 for the airports LHR, PEK and PKX, and SAN. The annual growth rates (CAGR) for the time periods 2016 30 and 2030 40 are shown on top of the columns for 2030 and 2040, respectively. In the most likely case, which is the do-nothing scenario for LHR, traffic volume increases only by 0.7% p.a. until 2030. Thereafter the number of aircraft movements remains constant and reaches a level of about 523,000. However, in the optimistic

Annual aircraft movements

Thousand

286

PART | III Forecasting future air traffic development up to 2040 1400

2.4%

1200

1000

3.1%

800

600

1.7% 1.8% 0.7% 0.0%

400 1.7%

200

0

LHR two runways

LHR three runways 2016

PEK and PKX 2030

1.5%

SAN

2040

FIGURE 10.25 Actual and forecast number of annual aircraft movements for London Heathrow (LHR) with two and three runways, Beijing Capital City (PEK) and Beijing Daxing (PKX), and San Diego (SAN) for the years 2016, 2030 and 2040 [values above the columns for 2030 and 2040 represent annual growth rates (CAGR) for the time periods 2016 30 and 2030 40]. CAGR, Compound annual growth rate.

case, which means a third runway, the number of aircraft movements is forecast to rise to more than 610,000 in 2030 and almost 725,000 in 2040, which is close to the capacity limit. CAGR of flights is 1.8% for the time period 2016 30 and 1.7% between 2030 and 2040 and, thus, substantially higher due to ample capacity reserves. On the other hand, PEK and PKX, and SAN are examples of airports that do not suffer from direct capacity constraints as there is sufficient capacity available. As a result, the number of aircraft movements grows much faster than at LHR in the case of two runways, especially at PEK and PKX, because of the strong growth of passenger demand volume. For 2030 almost 927,000 aircraft movements are expected at the two Beijing hub airports, and this value is forecast to increase to almost 1.2 million in 2040. This corresponds to a CAGR of 3.1% for 2016 30 and 2.4% for 2030 40. Between 2016 and 2040, the number of aircraft movements will rise by 94%. Because of a more mature market, it is expected to grow much slower at SAN. For 2030 about 218,000 aircraft movements are forecast, which corresponds to a CAGR of 1.7%. The number of aircraft movements is expected to reach a level of almost 254,000, which corresponds to a CAGR of 1.5% between 2030 and 2040. Thus we expect SAN to reach its capacity limit around the year 2040, which is about 260,000 aircraft movements.

Traffic forecast and mitigation strategies Chapter | 10

287

10.6.3 Aircraft size Fig. 10.26 presents the average aircraft size for 2016 and the forecast years 2030 and 2040 for the airports LHR, PEK and PKX, and SAN. The annual growth rates (CAGR) for the time periods 2016 30 and 2030 40 are displayed on top of the columns for 2030 and 2040, respectively. Average aircraft size is expected to increase substantially at all three airports. However, the strongest increase, both in absolute as well as in relative terms, can be found at LHR in the case of no further runway extension. The number of passengers per flight is on average 160 in 2016 and is forecast to increase to 224 in 2030 and 275 in 2040. This corresponds to an annual increase of 2.4% for the period 2016 30 and 2.1% for 2030 40. Thus despite an increasing capacity shortage, growth of passengers per flight slows down slightly, as there are limits to increasing average aircraft size, in particular at hub airports. In the case of a third runway, aircraft size still increases considerably at LHR. In 2030 average aircraft size is forecast to be 207 and 232 in 2040. This equals a CAGR of 1.9% for 2016 30 and 1.1% for 2030 40. Because of the eased capacity situation, the number of passengers per flight is, in 2030, 17 and, in 2040, 43 less than in the case of two runways. Due to capacity constraints at destination airports, average aircraft size increases at PEK and PKX, and SAN considerably as well, however, much 300

Passengers per flight (aircraft size)

2.1% 1.2%

250

1.1%

2.4%

2.3%

1.9%

200 1.3% 1.8%

150

100

50

0

LHR two runways

LHR three runways 2016

PEK and PKX 2030

SAN

2040

FIGURE 10.26 Actual and forecast average aircraft size, that means passengers per flight, for London Heathrow (LHR) with two and three runways, Beijing Capital City (PEK) and Beijing Daxing (PKX), and San Diego (SAN) for the years 2016, 2030 and 2040 [values above the columns for 2030 and 2040 represent annual growth rates (CAGR) for the time periods 2016 30 and 2030 40]. CAGR, Compound annual growth rate.

288

PART | III Forecasting future air traffic development up to 2040

less than at LHR. Furthermore, the strong growth of passenger volume that is expected for 2030 and 2040 at PEK and PKX supports increasing aircraft size. At PEK and PKX, average aircraft size increases from 159 in 2016 to 218 (12.3% p.a. between 2016 and 2030) in 2030 and finally to 246 (11.2% p.a. between 2030 and 2040) in 2040. Average aircraft size increases at SAN from 121 passengers per flight in 2016 to 155 in 2030 and 175 in 2040. This corresponds to a CAGR of 1.8% for 2016 30 and 1.3% for 2030 40.

10.6.4 Capacity analyses

Million

Capacity gain by larger aircraft in annual passenger volume

Figs 10.27 and 10.28 display the capacity gain by larger aircraft and by runway measures, respectively, for the airports LHR, PEK and PKX and SAN for the periods 2016 30 and 2030 40. As explained earlier, the capacity gain is measured in annual passenger volume to account for aircraft size development. The values on the top of the columns describe the measure-specific relative increase of passenger volume compared to the base year of the forecast, which means 2016 for the first period and 2030 for the second period. In the case of a do-nothing scenario, passenger volume at LHR will increase by 42% due to larger aircraft and by 12% through better runway utilisation between 2016 and 2030. Overall, passenger volume will increase by about 54% during that period. Thus around 78% of the necessary capacity 60 56.1%

50

40 16.0%

42.2%

30

22.7%

34.8% 17.5%

20

10

0

35.6%

LHR two runways

LHR three runways 2016–30

PEK and PKX

15.8%

SAN

2030–40

FIGURE 10.27 Capacity gain by employing larger aircraft, that means more passengers per flight, measured in annual passenger volume for London Heathrow (LHR) with two and three runways, Beijing Capital City (PEK) and Beijing Daxing (PKX), and San Diego (SAN) for the time periods 2016 to 2030 and 2030 to 2040 (values above the columns for 2016 30 and 2030 40 represent the relative capacity gain during that period).

Capacity gain by improving runway utilisation and enlargements in annual passenger volume

Million

Traffic forecast and mitigation strategies Chapter | 10

289

60 27.3% 53.3%

50

40

30 30.5% 15.4%

20

10

11.8% 25.6% 15.8%

0

0.0%

LHR two runways

LHR three runways 2016–30

PEK and PKX

SAN

2030–40

FIGURE 10.28 Capacity gain by improving runway utilisation and runway enlargements measured in annual passenger volume for London Heathrow (LHR) with two and three runways, Beijing Capital City (PEK) and Beijing Daxing (PKX), and San Diego (SAN) for the time periods 2016 30 and 2030 40 (values above the columns for 2016 30 and 2030 40 represent the relative capacity gain during that period).

will be provided by increasing aircraft size and only 22% of the necessary capacity comes from better runway utilisation so that more traffic can be handled by the airport. For the period 2030 40, only further increasing aircraft size will contribute to capacity gains at LHR to accommodate the rise of passenger volume of 23% compared to 2030. As a result, the increase of total annual passenger volume between 2016 and 2040 at LHR of almost 68 million passengers is made up of 59 million passengers (87%) capacity gain by larger aircraft and 9 million passengers (13%) capacity gain by better runway utilisation. This is no surprise given the tight capacity situation at LHR and things would be very different in the case of a third runway. In this case, passenger volume would rise by 35% due to larger aircraft and by 31% through better runway utilisation and enlargements between 2016 and 2030, which is more balanced because of ample capacity reserves. For 2030 40 passenger volume is expected to increase by 18% due to larger aircraft and by 15% through better runway utilisation, which is roughly a fifty-fifty ratio. Due to the large capacity increase, total passenger volume is forecast to increase by about 91.3 million (1121%) between 2016 and 2040, which is made up of 48.6 million (53%) capacity gain by larger aircraft and 42.7 million (47%) capacity gain by better runway utilisation and enlargements. The airports PEK and PKX are a different case, because the new hub airport PKX means a massive capacity expansion in the Beijing region. Thus

290

PART | III Forecasting future air traffic development up to 2040

passenger volume will increase by 56% due to larger aircraft and by 53% through better runway utilisation and enlargements between 2016 and 2030. This is a more balanced ratio compared to LHR with two runways, and whereas there is sufficient airport capacity in the Beijing region, there are major capacity bottlenecks at other major Asian airports (see Tables 10.7 and 10.13). In total, passenger volume will rise by almost 110% between 2016 and 2030, and each measure provides about half of the necessary capacity. For the 2030 40 period, better runway utilisation becomes more important relative to increasing aircraft size, because aircraft size is already very large and unlike LHR (two runways), there are still ample capacity reserves. Nevertheless, aircraft size continues to increase between 2030 and 2040 substantially. The total increase of annual passenger volume between 2016 and 2040 at PEK and PKX of more than 193 million passengers is made up of more than 86 million (45%) capacity gain by larger aircraft and nearly 107 million (55%) capacity gain by better runway utilisation and enlargements. This is a more balanced ratio compared to the most likely LHR case. Finally, passenger volume at SAN will increase by 36% due to larger aircraft and by 26% through better runway utilisation between 2016 and 2030. Therefore passenger volume will increase by 61% during that period, and about 58% of the necessary capacity will be provided by increasing aircraft size. On the other hand, 42% of the necessary capacity comes from better runway utilisation, as there are considerable capacity reserves at SAN. For the years 2030 40, as at PEK and PKX, better runway utilisation and increasing aircraft size each contribute about 50% to the necessary capacity gain, because aircraft size is already relatively large for a North American airport, and there are still sufficient capacity reserves up to around the year 2040. However, aircraft size is still expected to rise between 2030 and 2040, and the increase totals more than 50 passengers per flight between 2016 and 2040. The overall increase of annual passenger volume between 2016 and 2040 at SAN of 23.5 million passengers is made up of almost 12.8 million (54%) capacity gain by larger aircraft and nearly 10.7 million (46%) capacity gain by better runway utilisation. Thus while SAN and PEK and PKX differ clearly in terms of passenger and flight volume as well as average aircraft size, they are quite similar regarding the composition of mitigation measures.

10.7 Conclusion In this chapter, we presented the main results of the global, regional and airport-specific passenger and flight volume forecast. Future growth of traffic volumes and limiting factors of growth were discussed in the context of measures applied to mitigating existing or forecast airport capacity shortages. As a consequence, some general mitigation strategies have been derived for

Traffic forecast and mitigation strategies Chapter | 10

291

airports of different types regarding their traffic volume and average aircraft size, which means passengers per flight. As was to be expected, our forecast yields a continuation of the longterm growth of air passenger demand and flight volume until 2040, however, with growth varying strongly between world regions. The pace of growth depends on development assumptions of model input variables such as, in particular, socio-economic factors and model design variables such as technological innovations, airport capacity enlargements and growth of aircraft size and utilisation. A key input is the economic development; here, it has been assumed that the real GDP per capita will globally rise by 2.36% p.a. on average between 2016 and 2040. The model’s inherent assumptions of capacity enlargements and aircraft size directly influence forecast results and are seen as the main characteristics of mitigation strategies. Air traffic is forecast to grow substantially until 2040 despite the growing capacity shortage, especially at major airports. Passenger volume is expected to rise from about four billion in 2016 to 9.4 billion passengers in 2040. Depending on the demand growth, the number of flights is forecast to increase from 35.5 million in 2016 to 52.7 million in 2040. Finally, the number of passengers per flight is forecast to rise from an average of 111 in 2016 to 179 passengers in 2040. We have subdivided the forecast into the subperiods 2016 30 and 2030 40 to highlight the effects of capacity constraints at airports on air traffic development. While there is only a marginal unaccommodated passenger volume in 2030 of 49 million on the global level (but not for particular airports), and a lack of airport capacity can still be offset by further increasing aircraft size, there is a substantial shortage in 2040. Almost 256 million passengers cannot be served due to capacity constraints, equalling 2.6% of the global unconstrained demand volume, and more importantly, when relating the value of 2040 to the 2030 value, this share is expected to rise fast beyond 2040. The airports with the largest unaccommodated passenger volume are located in Asia, North America and Europe including Delhi Indira Gandhi (DEL), Atlanta Hartsfield Jackson (ATL) and London Heathrow (LHR). For these airports, the share of unaccommodated passenger volume reaches levels of 15% 25% of their unconstrained passenger volume. Furthermore, we have analysed how much increasing aircraft size as well as better runway utilisation and enlargements contribute to the capacity needed to handle the forecast passenger volume in 2030 and 2040. On the global level, larger aircraft account for 3.1 billion additional passengers and better runway utilisation and enlargements account for 2.3 billion additional passengers. Thus about 57% of the passenger volume growth between 2016 and 2040 is enabled by more passengers per flight and 43% by runwayrelated measures. For heavily capacity constrained airports, increasing aircraft size is even more important, while airports with ample capacity reserves show a more balanced distribution.

292

PART | III Forecasting future air traffic development up to 2040

Increasing aircraft size and runway capacity by adding new runways and raising runway utilisation turned out to be the most powerful measures of all options available to airport planners and operators. Theoretically, a whole range of technological, investment and operational options for mitigating a capacity shortage at airports does exist, which have been discussed in Chapter 5, General strategies for mitigating airport capacity constraints. Some of these measures have been successfully applied, especially increasing aircraft size and, whenever possible, adding runway capacity, while others have turned out to be rather undesirable measures, such as rerouting flights from congested airports, in particular hubs, to secondary airports, and others are still discussed. However, these have not been proven as effective mitigation measures, such as market-based options, in particular peak hour pricing of flights and slot auctions in slot coordination. Based on these findings, we have derived general mitigation strategies. We have subdivided airports according to their average aircraft size and traffic volume into four classes. Depending on their values, we recommend focusing on increasing aircraft size, improving runway utilisation and enlargements or both. We have also conducted such analyses on the global level for the seven world regions and particular airports in those regions. Between 2016 and 2040, more and more airports fall into the category with a large traffic volume as well as large average aircraft size, making further capacity gains difficult. These airports account for 29% of the traffic volume in 2016 and 55% in 2040. On the other hand, airports with a relatively low traffic volume and aircraft size have a share of 31% in 2016, thus being the largest class, but only 10% in 2040, dropping back to third place. As a result, mitigation measures that we have found to be unfavourable until 2040 in this book (see Chapter 5: General strategies for mitigating airport capacity constraints) probably need to be reconsidered beyond 2040 despite their drawbacks, especially shifting traffic to neighbouring airports. The forecasts for 2030 and 2040 demonstrate that increasing aircraft size and airport capacity will have limits in future. This is especially true for Asian airports because of the forecast strong demand growth of that region and the already substantial level of passenger and traffic volume. The case study shows the distinct effect of a substantial increase of airport capacity on future air traffic development by means of the new airport in Beijing, which opened late in 2019, and the third runway at London Heathrow, which may still have an uncertain future because of public resistance, although the British government has endorsed the project. Mitigation measures of airports coping with capacity shortfalls differ between and within world regions, since the share of constrained airports as well as the distribution and future development of traffic volumes and average aircraft size vary substantially. Africa has a relatively high number of small-volume airports with a wide range of average aircraft size, which in most cases do not have major capacity problems. Options to react to future

Traffic forecast and mitigation strategies Chapter | 10

293

capacity constraints would be, first, new runways and for airports with rather small aircraft to increase the seat capacity of flights. Asia has, in contrast, a high share of airports with high traffic volumes as well as aircraft with high seat capacity. In addition, almost half of the 50 top-ranking airports with the highest capacity utilisation in 2016 are located in Asia. Given the strong passenger demand growth forecast for Asia, a general mitigation strategy does not exist. For airports with more runways in operation, and correspondingly large traffic volumes, we would recommend additional airports and, as an intermediate measure, a further increase of average aircraft size. We know, however, that both types of measures may no longer be feasible in the long-term future due to local constraints and non-availability of suitable aircraft types. Future traffic in Europe will be distributed over airports with a wide range of traffic volumes with rather high seat capacity. Most major airports do not yet have complex runway systems; therefore for those with further capacity needs, we would recommend adding new runways. Public opposition to capacity enlargements, however, has been strong in Europe. Hence, improving runway utilisation by further increasing average aircraft size may be the only feasible mitigation strategy for such airports for a longer period of time. More than other world regions, North America has many airports with high runway capacity and large traffic volumes. In 2040 almost threequarters of all flights will be handled by high-volume airports, of which, in 2016, around 13 airports already had high utilisation rates. On the other hand, traffic growth in North America will be relatively low and average aircraft size is the smallest of all world regions. A clear recommendation for the capacity constrained airports among them is therefore to increase average aircraft size. Most of the traffic in the three remaining world regions, the Middle East, South America and the Southwest Pacific, is handled by airports with smaller traffic volumes than in the major regions of Asia, Europe and North America and with an average aircraft size that varies greatly among airports. Two of the three biggest airports, Dubai (DXB) and Mexico City (MEX) will need and probably get new runway capacity. For other airports with a lack of capacity, we would recommend additional runways as well; however, improving runway utilisation remains a good choice, too.

References Airports Commission, 2015. Airports Commission: Final Report, July 2015. International Civil Aviation Organization (ICAO), 2017. ICAO Traffic Statistics. Montreal. Information Handling Services (IHS) Markit, 2017. Global Economy: Forecast of GDP per Capita and Population Growth Rates (CAGR) 2016 to 2030 and 2030 to 2040. IHS Markit, London. Sabre AirVision Market Intelligence (MI), 2016. Data Based on Market Information Data Tapes (MIDT). Sabre, Southlake.

Chapter 11

Summary and conclusion The global airport network with regular scheduled services consists of more than 4000 airports. The great majority of them are single runway airports with relatively small traffic volumes and ample capacity. At first sight, these airports may be regarded as a vast capacity reservoir for handling future traffic. The analysis of traffic distribution in the global network has disclosed another story. Air traffic has been, until today, very much concentrated on a small number of important airports, in particular hub airports. Half of the global traffic is handled by just 120 airports (3% of all airports), with average traffic volumes of around 300,000 aircraft movements in 2016, and the other half by the remaining 3934 airports, which handle just about 9000 aircraft movements on average per airport and year. The global airport constraint analysis has revealed that these airports are not only the most important nodes of the network, but they are also to a great extent confronted with severe capacity problems, such as London Heathrow, New York LaGuardia and Beijing Capital City. If these airports already struggle with capacity shortfalls today, what are their chances of keeping up with the traffic forecast by industry and international institutions to continue to grow in the long term? The more general question is whether or not airport capacity constraints will become such a problem in the global air transport network as to form a substantial barrier to future growth of demand for air services. The prime objectives of this book have been G

G

to present a detailed empirical and model-based analysis of the impact of limited airport capacity on the future development of air traffic and to derive general strategies for mitigating capacity constraints at airports concerned in world regions.

This book is structured in three parts. The first part deals with a discussion of concepts of airport capacity and capacity utilisation and an analysis of the global capacity constraint situation of airports. The second part concentrates on methodological questions of mapping out general mitigation strategies, of forecasting passenger demand, airport capacity and utilisation, realisation probabilities of airport capacity enlargements and aircraft size development. The third part presents our forecasts of passenger and traffic

Airport Capacity Constraints and Strategies for Mitigation. DOI: https://doi.org/10.1016/B978-0-12-812657-8.00011-7 © 2020 Elsevier Inc. All rights reserved.

295

296

PART | III Forecasting future air traffic development up to 2040

volumes at airports in world regions until 2030 and 2040 and the assessment of strategies to mitigate capacity constraints at airports in world regions. We have been aware of the fact that each constrained airport has specific capacity problems and that solutions to mitigate these problems vary from airport to airport and from region to region, depending among other things, on the severity and type of capacity bottleneck, the financial situation of the airport and the regulatory framework of the region and state, which the airport operator has to take into account in planning for the future. Nevertheless, our intention has been to analyse, on a global level, airport capacity constraints and mitigation measures and options, which either have been successfully applied at airports and may be recommended for further use, or to discuss, on theoretical grounds, those which have yet not proven their practical relevance and of which at least some do not contribute to alleviating the capacity problem and should not be recommended accordingly. The constraint analysis has clearly shown that the employment of aircraft with higher seat capacity has proven to be a powerful instrument for improving the capacity utilisation of existing facilities. The driving factor of operating larger aircraft, however, lies within the strategic planning of airlines with hardly any possibility for airports to influence this process. The capacity planning of airports concentrates more on the investment option of enlarging infrastructure capacity. Since both system partners airports and airlines affect future capacity as well as capacity needs of airports, there is a case for a combined airport-specific as well as a general capacity planning strategy. Given this hypothesis, we derived some general recommendations on capacity constraint mitigation strategies. To familiarise the reader with the problem of airport capacity, we have first clarified the terms of capacity in Chapter 2, Concepts of capacity and methods of estimation, and capacity utilisation in Chapter 3, Capacity utilisation at airports worldwide. Since at major airports with capacity shortfalls the runway system is typically the weakest component in terms of airport capacity and the most difficult one to enlarge on grounds of the involvement of the public, we concentrate here on the runway capacity. In addition, we are interested in a practical or sustained capacity rather than a theoretical capacity which does not account for an acceptable level of service. We have seen that different capacity concepts exist as well as approaches of estimating runway capacities. Applying more recent methods such as simulation models requires the input of detailed data of the airport under consideration. Our objective has been to analyse the global capacity situation of airports, which means that the empirical approach to estimating capacity has to allow to be applied to all airports of the analysis in an equal and consistent way. We have, therefore, developed a new approach which includes empirical and benchmark elements. The methodological nucleus is traffic ranking functions of congested airports, based on detailed flight schedule data by airport. The model calculates hourly proxy values of capacity and capacity utilisation

Summary and conclusion Chapter | 11

297

from these functions and relates them to annual service volumes. This approach enables us to analyse the capacity situation of airports which are already congested or are likely to be in the future. Traffic ranking functions are a tool to visualise the capacity situation at airports and to determine specific volumes at certain hours. Aircraft movements per hour are ranked in descending order for all hours of the year, starting with the busiest hour. The functional form describes the degree of capacity utilisation. The less inclined the slope of the curve over long periods of time, the closer the airport is to reaching capacity in these periods, if and only if the top-ranking hourly volumes reach levels near capacity. If traffic ranking curves are belly shaped, as, for instance, those of London Heathrow (see Fig. 11.1), then the 5% peak hour volumes are indicative of the practical capacity of the airport, thus allowing identification of airports with a high capacity utilisation. Capacity utilisation of airports is regarded as a prime indicator of capacity reserves, or more importantly for high-volume airports, of a lack of capacity reserves. For assessing airport capacity reserves, the “capacity utilisation index” (CUI) has been defined as the ratio of the average hour and the 5% peak hour traffic volume. The hour has been chosen as a suitable time unit for measuring the capacity since the maximum throughput of an infrastructure such as a runway can be achieved by the demand in real-life conditions only in short time intervals. This is not the case for a whole year because of many low or zero demand periods at night and during weekends. The absolute slope of the traffic ranking function serves as a measure for capacity utilisation; the CUI is an indicator of that slope. For London Heathrow, for instance, the airport has been operating at capacity for many years; the CUI of the year 2016 has been calculated at 0.82. In contrast, 100

London Heathrow 5% peak hour volume: 87 Average hour volume: 72 CUI: 0.82

Aircraft movements per hour

90 80 70 60 50 40

London Stansted: 5% peak hour volume: 40 Average hour volume: 24 CUI: 0.59

30 20 10 0

1 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75 78 81 84 87 Hours of the year

FIGURE 11.1 Air traffic ranking functions with capacity utilisation index (CUI) of London Heathrow (LHR) and London Stansted (STN) 2016 [Official Airline Guide (OAG), 2016].

298

PART | III Forecasting future air traffic development up to 2040

London Stansted, an airport with capacity reserves, has a CUI of only 0.59 (see Fig. 11.1). Traffic volumes reach capacity only in a few peak hours per year. Capacity utilisation indices have been developed on the basis of traffic ranking functions for each airport in the global network. The CUI values are, thus, comparable between airports of different size and with different runway systems. By means of the CUI, we have categorised airports according to their importance within the global air traffic system and to the level of capacity reserve or of capacity constraint, respectively. So, we have identified those airports which serve a large share of the global traffic volume and, at the same time, suffer from capacity bottlenecks. To identify constrained airports and categorise them according to the level of constraint, we first had to give an answer to the problem of how to define and measure capacity constraint. Airport congestion encompasses the whole transition area between constrained flow conditions in some peak hours and dense traffic conditions with high delays for each aircraft over longer periods of time. Due to the lack of delay data comparable between airports, we have pursued an approach which describes the intensity of capacity constraint in relation to the relative amount of time in which aircraft movements at an airport are handled under near-capacity conditions. We first eliminated those airports which have no capacity problems or will not face them in the near future by identifying airports with peak hour flight volumes being smaller than about 50% of their practical capacity. The vast majority of airports, almost 3900 out of the total of more than 4000 airports in 2016, belong to the group of unconstrained airports. The remaining 172 airports are likely to operate at or near capacity. For these airports we have determined, in a further step, the relative time span during a year in which they operate in near-capacity conditions by means of traffic ranking functions with traffic volumes and hours operating at each traffic volume in relative terms. Traffic volumes are expressed, thus, as shares of practical capacity and the number of hours as shares of the total operating time per year. Four capacity utilisation classes have been derived. Airports are in the lowest class A if their 5% peak hour volumes exceed 70% of the 5% peak hour capacity in more than one hour and in less than 5% of all operating hours of the year. Airports belonging to the highest capacity utilisation class D are those with 5% peak hour volumes exceeding 70% of the 5% peak hour capacity in more than 65% of all operating hours. The class A is the largest among the four capacity utilisation classes with 137 airports. Class A airports experience some congestion in peak hours, however, only for short periods in up to about 300 400 hours per year. As such, we have not yet classified them as congested airports, although we would not contradict operators of airports with traffic volumes approaching capacity in 5% of operating hours, if they think that their airport has a capacity problem. In our global analysis, we have classified the class B, C and D airports as constrained. Thus 35 airports have been more or less constrained in 2016,

Summary and conclusion Chapter | 11

299

with London Heathrow, New York LaGuardia and Beijing Capital City being the most constrained airports of class D. Other predominant examples of constrained airports are Atlanta Hartsfield-Jackson, Chicago O’Hare, Los Angeles and Newark in North America; Mexico City in South America; Singapore, Bangkok, Delhi, Jakarta and Hong Kong in Asia; Munich, London Gatwick and Istanbul (IST) in Europe and Dubai in the Middle East. The 35 congested airports are the main airports of the global network; they handled a traffic volume of 13.3 million aircraft movements, corresponding to a share of nearly 19% of the global traffic in 2016. The global constraint analysis has clearly confirmed that traffic is concentrated on a relatively small number of airports. Furthermore, these are the most important nodes of the network and they are to a great extent confronted with severe capacity problems. The second part of this book deals with the methodological background of the forecast model, which is applied in the third part. The forecast model consists of four submodels for forecasting: G

G G G

air passenger demand, that is origin destination (OD) passenger flows and total passenger flows including transfer passengers between countries and airports, respectively airport capacity and capacity utilisation realisation probabilities of airport capacity enlargements and limits average number of passengers per flight (so-called aircraft size)

The air passenger demand model is based on the fundamental theory of the gravity law. While variables such as distance, population and gross domestic product (GDP) are rather common in gravity models, we expected further insights by including an airfare variable. For better discrimination between different types of OD flows, for example, tourist destinations, we have included variables, such as tourism receipts and expenditures and population density. Furthermore, we have employed a Poisson pseudo maximum likelihood estimator to produce better and more reliable forecast results, thereby enhancing the out-of-sample results, that is forecast efficacy. A series of relevant demand elasticities have been derived from the model such as, for instance, airfare elasticity. On the one hand, if airfares rise by 1%, OD passenger volume declines by 1.11%. On the other hand, if GDP per capita increases in the origin country by 1%, then OD passenger demand rises by 0.45%, and if GDP per capacity rises in the destination region by 1%, the OD demand grows by 0.23%. For assessing the effects of limited airport capacity on global air traffic, we have developed a problem-specific approach suited for application at the global level. However, it cannot be a substitute for a detailed capacity analysis in airport-specific studies. We employed a data envelopment analysis (DEA) and conducted a regression analysis based upon the DEA results. DEA is a non-parametric empirical method of operations research to estimate

300

PART | III Forecasting future air traffic development up to 2040

production frontiers employing linear programming techniques. This approach enabled us to compute current and future annual service volumes of airports worldwide. While the analysis of Chapter 3, Capacity utilisation at airports worldwide and Chapter 4, Constrained and underutilised airports concentrated on the current traffic situation, the model has allowed us to forecast 5% peak hour volume and average hourly volume in a situation of the highest possible capacity utilisation. With the model of forecasting realisation probabilities of airport capacity enlargements and limits, we have introduced an approach which incorporates limited airport capacity in air transport forecasts. The key idea of the model is that delayed runway expansions are a result of opposition due to negative effects of such plans on the airport’s neighbourhood. Factors of the model are noise annoyance, level of welfare, economic opportunities, intermodal substitution possibilities and level of participation. This enables us to estimate the probability of realisation of a new runway which can be transformed into an expected value of delay. The fourth submodel is a simple to apply and robust approach to forecast the development of the key forecast variable passengers per flight by airport pair to extend the forecast of airport capacities to available passenger volume capacities. As in the case of airport capacities, the approach is highly problem-specific and cannot be a substitute for a detailed flight route analysis in terms of aircraft fleet characteristics and their future development. The method is very similar to the modelling of airport capacities. The basic analysis is a DEA, which is further refined by regression analyses to generalise the results, so that they can be used for forecasting and projection, respectively. With all the input data needed for running the four-part model, passenger and flight volumes of different geographical granularity have been forecast for the years 2030 and 2040. At the most detailed level, passenger flows and flight volumes on routes between airports have been estimated, which have been aggregated to world region specific flows. Airport-specific results are presented for the top 20 airports in terms of unaccommodated demand volume. Given the models of estimating future runway capacities and the probability of realising new runways, we are able to forecast both unconstrained as well as constrained passenger flows. Hereby, mainly two model variables have played a major role in estimating constrained passenger flows and resulting flight volumes: average aircraft size and capacity enlargements through new runways and, if the airport is not yet constrained, improving runway utilisation by more aircraft movements. Increasing aircraft size and adding new runways to a constrained airport are regarded as the main options for mitigating capacity shortfalls. Hence, we show how much capacity in terms of passengers handled can be gained by these measures up to the years 2030 and 2040.

Summary and conclusion Chapter | 11

301

Before applying the two key variables in the forecast, we have discussed the broad spectrum of investment and non-investment mitigation measures, including those which have been commonly practised so far as well as those which are discussed on theoretical grounds, but have not yet found their way into real-life operations. Investment measures encompass direct investment options, such as new runways or terminals, and indirect investments, such as new rail terminals near airports, and non-investment measures involve supply and demand management options, ranging from administrative or regulatory measures to hybrid to pure market-based options. A typology of options to mitigate the negative effects of airport capacity constraints has been developed of which the following measures are practised or may be candidates for future application: G G G G G

adding new runways as investment options rerouting flights to less-utilised airports nearby diverting flights to off-peak hours raising seat capacity of flights raising load factors of flights

We have observed that the global network of 4054 airports in 2016 has been enlarged by about 400 runways since 2008; the majority of them, however, in the network of 3954 secondary airports. The top 100 airports in terms of traffic volume have seen almost no runway capacity extension, although these airports handle about 45% of the total traffic, in many instances in severe capacity constrained circumstances. The fact that these airports were not in a position with a few exceptions to add capacity is indicative of the difficult political situation regarding airport extensions in many states. We have discussed the measure of rerouting flights to secondary airports nearby and concluded that this option is not widely used, in particular not at hub airports, because of the interrelationship of incoming and outgoing flights. Of similar nature is the temporal diversion of flights to off-peak periods at partly congested airports. Only in exceptional cases do airlines have an interest in accepting shifting flights from their originally scheduled time. In contrast to the flight rerouting and diverting options, raising seat capacity per flight has been the most effective measure among the noninvestment options to keep pace with growing bottlenecks at airports. The average number of seats per flight has grown globally from 105 seats in 2000 to 138 seats per flight in 2016, hence in 16 years by 33 seats or by 31% in relative terms. This option has been chosen at the 100 major airports as well as at all other airports. Average seat capacity at the 100 top-ranking airports has risen by 24 seats to 156 seats from 2008 to 2016 and at all other airports by 18 seats to 123 seats per flight. At the 100 top-ranking airports, of which about 40 airports have been more or less capacity constrained, flights offered in 2016 had, thus, 33 seats more than at all other airports.

302

PART | III Forecasting future air traffic development up to 2040

In terms of passenger throughput efficiency, this means that at the major airports 127 passengers were transported per flight, whereas at the great number of all other airports 98 passengers, and thus 29 passengers less, were on average on board a flight. Besides the fact that a lack of available slots hinders airlines from increasing frequencies at congested airports, it is an economic interest which leads airlines to operate larger aircraft with lower unit costs, both at constrained and unconstrained airports. All airlines have an economic interest in raising the load factors of flights. In fact, they have already achieved rather high levels of passenger occupancy rates with average values of 80% and more. The analysis of mitigation measures has shown that new runways as an investment option and raising seat capacity per flight as an operational option have proven to be more effective than the other measures. Other measures, in particular technological, regulatory, hybrid and market-based options have been discussed as well. Regulatory measures such as the widely applied International Air Transport Association (IATA) slot coordination have been introduced by public authorities to optimise traffic flows in capacity constrained conditions or to allow or prioritise certain types of traffic such as scheduled flights, and exclude others, for instance, general aviation at main airports. The IATA slot coordination deals with capacity scarcity without applying market-based measures, although slots at constrained airports have a high market value and are traded, wherever allowed, at high prices among airlines, after they have been allocated in the first place (secondary slot trading). Pricing options, in contrast, aim to optimise and maximise traffic flows by allowing airlines to trade with slots and airport operators to charge different prices for aircraft operations depending on the degree of constraint. Primary slot trading (without any preceding slot coordination regulation) and congestion pricing are, however, still theoretical options and not yet applied in air transportation. These measures have, therefore, not been included in the traffic forecast. Nevertheless, their potential effects of mitigating capacity scarcity deserve further research in future projects. As was to be expected, the demand and flight forecast yields a continuation of the long-term growth of air passenger demand and flight volume until 2040, however, with growth varying strongly between world regions and airports. The pace of growth depends on development assumptions of model input variables, in particular, socio-economic factors and model design variables such as technological innovations, airport capacity enlargements and growth of aircraft size and utilisation. A key input is the economic development; here it has been assumed that the real GDP per capita will globally rise by 2.36% p.a. on average between 2016 and 2040. The model’s inherent assumptions of capacity enlargements and aircraft size directly influence the forecast results and are seen as the main characteristics of mitigation strategies.

Summary and conclusion Chapter | 11

303

Air traffic is forecast to grow substantially until 2040 despite the growing capacity shortage, especially at major airports. Passenger volume is expected to grow from about four billion in 2016 to 9.4 billion passengers in 2040 (see Table 11.1). Depending on the demand growth, the number of flights is forecast to increase from 35.5 million in 2016 to 52.7 million in 2040. Finally, the number of passengers per flight is forecast to rise from an average of 111 in 2016 to 179 passengers in 2040. We have subdivided the forecast into the subperiods 2016 30 and 2030 40 to highlight the effects of capacity constraints at airports on air traffic development. While there is only a marginal unaccommodated passenger volume in 2030 of 49 million on the global level (but not for particular airports), and a lack of airport capacity can still be offset by further increasing aircraft size, there will be a substantial shortage in 2040. Almost 256 million passengers cannot be served due to capacity constraints, equalling 2.6% of the global unconstrained demand volume, and more importantly, when relating the value of 2040 to the 2030 value, this share is expected to rise fast beyond 2040. The airports with the largest unaccommodated passenger volume are located in Asia, North America and Europe, such as Delhi Indira Gandhi (DEL), Atlanta Hartsfield-Jackson (ATL) and London Heathrow (LHR). For these airports the share of unaccommodated passenger volume reaches levels of 15% 25% of the total unconstrained passenger volume. Direct constrained airports are those which suffer from constraints caused by lack of own capacity. Unaccommodated demand at Asian, European and North American airports is primarily caused by direct capacity constraints, while airports in other regions suffer mainly from indirect constraints, which means they are limited by capacity constraints at destination airports. The most constrained airports are typically identical with those with the highest capacity utilisation. In fact, the nine most constrained airports in 2030 and the ten most constrained airports in 2040 have a capacity utilisation of 100%. Tables 11.2 11.3 display the ten most constrained airports worldwide in 2030 and 2040. According to the forecast, Asian airports will primarily suffer from capacity shortfalls in future, mainly caused by the strong demand growth in that world region. Most of these airports belong to the largest airports of that region or the global network. However, two of the biggest airports worldwide, Atlanta Hartsfield-Jackson and Chicago O’Hare, will remain constrained as well, such as London Heathrow in Europe. Traffic growth is expected to vary greatly between world regions. While the global annual passenger volume will grow by 4.1% p.a. on average from 2016 to 2030 and further on to 2040 with 3.1% p.a., passenger volume of Asia, the region with the highest passenger volume of 1.3 billion in 2016, will grow by 4.9% p.a. in the first period and 3.3% p.a. from 2030 to 2040. North America and Europe are also regions with high passenger volumes of more than 900 million and almost one billion, respectively, in 2016.

TABLE 11.1 Main global air traffic forecast results for the years 2030 and 2040.

Passenger volume (billion)

2016

2030

CAGR (2016 30)

2040

CAGR (2030 40)

CAGR (2016 40)

4.0

6.9

4.1%

9.4

3.1%

3.7%

Unaccommodated passenger volume (million)

49.4

255.5

Total (unconstrained) passenger volume (billion)

7.0

9.7

Flight volume (million)

35.5

45.8

1.8%

52.7

1.4%

1.6%

Average number of passengers per flight (‘aircraft size’)

111

152

2.3%

179

1.7%

2.0%

Capacity gain by: (million passengers) More passengers per flight

1765

1375

Increasing runway utilisation and higher runway capacity

1220

1079

TABLE 11.2 Ten most constrained airports worldwide in 2030. Airport name

IATA code

Aircraft movements (thousand)

Passengers (million)

Unaccommodated passengers (million)

Share of unaccommodated passengers (%)

Atlanta Hartsfield-Jackson

ATL

974

158

11.2

6.6

London Heathrow

LHR

523

117

8.6

6.8

Mumbai Chhatrapati Shivaji

BOM

447

98

6.7

6.4

Guangzhou Baiyun

CAN

592

121

6.7

5.2

Chicago O’Hare

ORD

945

117

6.4

5.2

Jakarta Soekarno-Hatta

CGK

629

122

5.6

4.4

Delhi Indira Gandhi

DEL

599

127

3.6

2.7

New York LaGuardia

LGA

416

43

2.3

5.1

Bengaluru Kempegowda

BLR

274

51

1.0

1.8

Dubai

DXB

538

172

0.9

0.5

TABLE 11.3 Ten most constrained airports worldwide in 2040. Airport name

IATA code

Aircraft movements (thousand)

Passengers (million)

Unaccommodated passengers (million)

Share of unaccommodated passengers (%)

Delhi Indira Gandhi

DEL

599

159

46.0

22.4

Mumbai Chhatrapati Shivaji

BOM

447

123

42.1

25.5

Jakarta SoekarnoHatta

CGK

629

152

41.8

21.5

Atlanta HartsfieldJackson

ATL

974

195

33.6

14.7

London Heathrow

LHR

523

144

25.2

14.9

Chicago O’Hare

ORD

945

144

20.7

12.5

Bengaluru Kempegowda

BLR

274

65

17.6

21.3

Kuala Lumpur

KUL

593

140

9.4

6.3

Ninoy Aquino

MNL

442

113

7.7

6.4

Los Angeles

LAX

814

167

7.1

4.1

Summary and conclusion Chapter | 11

307

However, they will grow less strongly with 3.5% p.a. and 3.4% p.a. in the first period and 2.8% p.a. and 2.6% p.a. in the second period. This means that more than 2.6 billion passengers are forecast for Asia in 2030 and 3.7 billion passengers in 2040. For North America nearly 1.5 billion passengers in 2030 and 1.9 billion in 2040 are forecast, while for Europe passenger volumes of almost 1.6 billion and around two billion passengers are expected in 2030 and 2040, respectively. The other world regions of Africa, the Middle East, South America and the Southwest Pacific have much smaller passenger volumes. They will gain, however, in importance by growing with above-average growth rates until 2040. How much will increasing aircraft size and better runway utilisation and enlargements of runway capacity contribute to the capacity needed to handle the forecast passenger volume? On the global level, larger aircraft account for 3.1 billion additional passengers and better runway utilisation and enlargements account for 2.3 billion additional passengers. Thus about 57% of the passenger volume growth between 2016 and 2040 is enabled by more passengers per flight and 43% by runway-related measures. For heavily capacity constrained airports, increasing aircraft size is even more important, while airports with ample capacity reserves show a more balanced distribution. Increasing aircraft size and runway capacity by adding new runways and raising runway utilisation turned out to be the most powerful measures of all options available to airport planners and operators. Based on these findings, we have derived general mitigation strategies. We have subdivided airports according to their average aircraft size and traffic volume into four classes. Depending on their values, we recommend focusing on increasing aircraft size, improving runway utilisation and enlargements or both. We have conducted such analyses on the global level as well as for the seven world regions and particular airports in those regions. Between 2016 and 2040, more and more airports fall into the category with a large traffic volume as well as high average aircraft size, making further capacity gains difficult. These airports account for 29% of the traffic volume in 2016, and 55% in 2040. On the other hand, airports with a relatively low traffic volume and aircraft size have a share of 31% in 2016, thus being the largest class, but only 10% in 2040, dropping back to third place. As a result, mitigation measures which we have found to be unfavourable until 2040 in this book (see Chapter 5: General strategies for mitigating airport capacity constraints), probably need to be reconsidered beyond 2040 despite their drawbacks, especially shifting traffic to neighbouring airports. The forecasts for 2030 and 2040 demonstrate that increasing aircraft size and airport capacity will have limits in future. This is especially true for Asian airports because of the forecast strong demand growth of that region and the already substantial level of passenger and traffic volume. The case study shows the distinct effect of a substantial increase of airport capacity on future air traffic development by means of the new airport in Beijing, which opened late in 2019, and the third

308

PART | III Forecasting future air traffic development up to 2040

runway at London Heathrow, which may still have an uncertain future because of public resistance, although the British government has endorsed the project. Mitigation measures of airports coping with capacity shortfalls differ between and within world regions, since the share of constrained airports as well as the distribution and future development of traffic volumes and average aircraft size vary substantially. Africa has a relatively high number of small volume airports with a wide range of average aircraft size, which in most cases do not have major capacity problems. Options to react to future capacity constraints would be, first, new runways and, for airports with rather small aircraft, to increase seat capacity of flights. Asia has, in contrast, a high share of airports with high traffic volumes as well as aircraft with high seat capacity. In addition, almost half of the 50 top-ranking airports with the highest capacity utilisation in 2016 are located in Asia. Given the strong passenger demand growth forecast for Asia, a general mitigation strategy does not exist. For airports with more runways in operation and correspondingly large traffic volumes, we would recommend additional airports and as an intermediate measure a further increase of average aircraft size. Future traffic in Europe will be distributed over airports with a wide range of traffic volumes with rather high seat capacity. Most major airports do not yet have complex runway systems, so for those with further capacity needs we would recommend adding new runways. Public opposition to capacity enlargements, however, has been strong in Europe so that improving runway utilisation by further increasing average aircraft size may be the only feasible mitigation strategy for such airports for a longer period of time. More than other world regions, North America has many airports with high runway capacity and large traffic volumes. In 2040 almost threequarters of all flights will be handled by high-volume airports, of which, in 2016, around 13 airports already had high utilisation rates. On the other hand, traffic growth in North America will be relatively low and average aircraft size is the smallest of all world regions. A clear recommendation for capacity constrained airports with complex runway systems among them is, therefore, an increase of average aircraft size. Most of the traffic in the three remaining world regions, the Middle East, South America and the Southwest Pacific is handled by airports with smaller traffic volumes than in the major traffic regions of Asia, Europe and North America and with an average aircraft size that varies greatly among airports. Two of the three largest airports, Dubai (DXB) and Mexico City (MEX), will need and probably get new runway capacity. For other airports with lack of capacity, we would recommend additional runways; however, improving runway utilisation also remains a good choice. We are aware of the fact that these general mitigation strategies may not be of great help in some specific airport planning projects since local

Summary and conclusion Chapter | 11

309

circumstances may include constraints which are unique for the airport under consideration. We hope, however, that the options proposed show some valid directions of how to proceed in mitigating capacity problems, in particular, because the traffic and capacity forecast of all airports worldwide has proven the usefulness of the main mitigation contributors, that is increasing aircraft size and investing in new runways. We are convinced that for an airport planner a reflection of measures which have been successfully applied at a greater or even global level can be of great assistance in advocating in favour of such measures at that airport. In addition, as airports, airlines and air traffic management as system partners pursue proven mitigation strategies, the chances of realising operational and investment proposals will also augment specific airport planning projects.

Reference Official Airline Guide (OAG), 2016. Market Analysis. Reed Travel Group, Dunstable.

Appendix List of abbreviations 18/365 AA AACM ACM ADD ADL ADSIM AEP AFV AMS ANSP AP AP1 AP2 APV ASV ATC ATL ATM ATMB ATS AUH AYT BAA BCN BKK BLR BNE BOG BOM BOS BROAD BWI

18 hours 365 days a year arrival airport name annual aircraft volume number of aircraft movements Addis Ababa Airport Adelaide Airport Airport Delay Simulation Model Buenos Aires Jorge Newbery Airport annual flight volume Amsterdam Schiphol Airport air navigation service provider arrival priority configuration binary variables that take values of one if the number of aircraft movements is below 100,000 per year binary variables that take values of one if the number of aircraft movements is between 100,000 and 200,000 per year annual passenger volume annual service volume air traffic control Atlanta Hartsfield Jackson Airport air traffic management Air Traffic Management Bureau air traffic service Abu Dhabi Airport Antalya Airport British Airport Authority Barcelona Airport Bangkok Suvarnabhumi Airport Bengaluru Kempegowda Airport Brisbane Airport Bogota Airport Mumbai Chhatrapati Shivaji Airport Boston Logan Airport number of broadband subscribers per 100 people Baltimore Washington Airport

311

312

Appendix

CAAC CAGR CAI CAN CCC CCU CDG CGH CGK CJU CKG CLT CMO Cols. 8 11 CPH CPT CRS CSX CTU CU CUC CUI CUN DA DAFIF DCA DEA DEA2 DEA3 DEL DEN DFW Dist DLR DMK DMU DOH DP DPS DTW DUB DUS DXB ESRA EU EUR EWR

Civil Aviation Administration of China compound annual growth rate Cairo Airport Guangzhou Baiyun Airport Capacity coverage chart Netaji Subhas Chandra Bose Airport Paris Charles de Gaulle Airport Sao Paulo Congonhas Airport Jakarta Soekarno Hatta Airport Jeju Airport Chongqing Jiangbei Airport Charlotte Douglas Airport Boeing Current Market Outlook difference between the values of columns 8 and 11 Copenhagen Airport Cape Town Airport constant returns to scale Changsha Huanghua Airport Chengdu Shuangliu Airport Capacity utilisation in % Capacity utilisation class Capacity utilisation index Cancun Airport departure airport name Digital Aeronautical Flight Information Files Ronald Reagan Washington Airport data envelopment analysis data envelopment analysis with two inputs data envelopment analysis with three inputs Delhi Indira Gandhi Airport Denver Airport Dallas/Fort Worth Airport distance German Aerospace Center Bangkok Don Mueang Airport decision-making units Doha Airport departure priority configuration Denpasar Ngurah Rai Airport Detroit Metropolitan Wayne County Airport Dublin Airport Du¨sseldorf Airport Dubai Airport Eurocontrol Statistical Reference Area European Union Europe Newark Liberty Airport

Appendix FAA FCO FLL FRA FUK GDP GF GLA GMF GRGDP GRU GVA HAM HKG HKIA HND HNL HS Hub HYD IAD IAH IATA ICAO ICN IFR IHS Markit IST ITM JED JFK JNB km KMG KUL LAS LAX LCC LF LGA LGW LHR LIM LIMC LIN LIS

313

Federal Aviation Administration Rome Fiumicino Airport Fort Lauderdale-Hollywood Airport Frankfurt Airport Fukuoka Airport gross domestic product Greater Five Glasgow Airport Airbus Global Market Forecast growth rate of GDP per capita in purchase power parity of constant 2005 international US dollars Sao Paulo Guarulhos Airport Geneva Airport Hamburg Airport Hong Kong Airport Hong Kong International Airport Tokyo Haneda Airport Honolulu Airport hub-and-spoke hub routes Hyderabad Rajiv Gandhi Airport Washington Dulles Airport Houston George Bush Airport International Air Transport Association International Civil Aviation Organization Seoul-Incheon Airport instrument flight rules Information Handling Services Markit Istanbul Ataturk Airport Osaka Airport Jeddah Airport New York John F. Kennedy Airport Johannesburg Airport kilometres Kunming Changshui Airport Kuala Lumpur Airport Las Vegas McCarran Airport Los Angeles Airport low-cost carrier low-frequency routes New York LaGuardia Airport London Gatwick Airport London Heathrow Airport Lima Jorge Chavez Airport low instrument Milan-Linate Airport Lisbon Airport

314

Appendix

LL LN (AACM) MAA MAACM/RWY/OH MAC/RWY/OH MAN MCO MDW MEL MEM MEX MIA MNL MSP MTOM MUC NBO NCE OAG OD ODP ODPS OECD OH OLS ORD OSL p.a. P/F P2P P2P PART PC PC/F PCU PE PEK PER PHCAP PHL PHTV PHX PKX PMI POP10KM

log-likelihood function logarithm of annual aircraft volume Chennai Airport maximum average aircraft movements per runway per operating hour maximum average capacity per runway per operating hour Manchester Airport Orlando Airport Chicago Midway Airport Melbourne Airport Memphis Airport Mexico City Benito Juarez Airport Miami Airport Manila Ninoy Aquino Airport Minneapolis Saint Paul Airport maximum take-off mass Munich Airport Nairobi Airport Nice Airport Official Airline Guide origin destination OD passenger volume OD passenger share Organisation for Economic Co-operation and Development operating hours ordinary least squares method Chicago O’Hare Airport Oslo Airport per annum passengers per flight point-to-point point-to-point routes binary variable that takes a value of one if the type of government conforms to democratic principles and a value of zero if not estimated annual passenger capacity potential estimated passenger capacity potential per flight passenger capacity potential utilisation in % projection error Beijing Capital City Airport Perth Airport practical hourly capacity Philadelphia Airport peak hour volume traffic volume Phoenix Sky Harbor Airport Beijing Daxing Airport Palma de Mallorca Airport number of people living within 10 km of an airport

Appendix pP/F PPML PVG RA RAILKM2 RDSIM RP RPK RSS RUH RWY RWYs Sabre MI SAN SAW SAW SDU SEA SFO SGN SHA SID SIN SLC STAR STN STR SUB SVO SYD SZX TAAM TAO TLV TOUR TP TPA TXL UK UN UPG URC US USD VFR VIE

315

projected passengers per flight poisson pseudo maximum likelihood Shanghai Pudong Airport regression analysis number of railway kilometres per square kilometre of the country Runway Delay Simulation Model realisation probability revenue passenger kilometres residual sum of squares Riyadh Airport runway runways Sabre Market Intelligence San Diego Airport Sabiha Go¨kcen Airport Istanbul Sabiha Gokcen Airport Rio de Janeiro-Santos Dumont Airport Seattle Tacoma Airport San Francisco Airport Tan Son Nhat Airport Shanghai Hongqiao Airport standard instrument departures Singapore Changi Airport Salt Lake City Airport standard arrival routes London Stansted Airport Stuttgart Airport Juanda Airport Moscow Sheremetyevo Airport Sydney Kingsford Smith Airport Shenzhen Bao’an Airport Total Airspace and Airport Modeller Qingdao Liuting Airport Tel Aviv Airport receipts from international tourism as a percentage from total exports total passenger volume Tampa Airport Berlin Tegel Airport United Kingdom United Nations Sultan Hasanuddin Airport ¨ ru¨mqi Diwopu Airport U United States US dollar visual flight rules Vienna Airport

316 WAW WDI XMN YVR

Appendix Warsaw Airport World Development Indicators Xiamen Gaoqi Airport Vancouver Airport

Definition of world regions Country

World region

Afghanistan Albania Algeria American Samoa Angola Anguilla Antigua and Barbuda Argentina Armenia Aruba Australia Austria Azerbaijan Bahamas Bahrain Bangladesh Barbados Belarus Belgium Belize Benin Bermuda Bhutan Bolivia Bosnia and Herzegovina Botswana Brazil Brunei Darussalam Bulgaria Burkina Faso Burundi Cambodia Cameroon Canada Cape Verde Cayman Islands Central African Republic Chad Chile China

Asia Europe Africa Southwest Pacific Africa South America South America South America Europe South America Southwest Pacific Europe Europe South America Middle East Asia South America Europe Europe South America Africa South America Asia South America Europe Africa South America Asia Europe Africa Africa Asia Africa North America Africa South America Africa Africa South America Asia (Continued )

Appendix

(Continued) Country

World region

Colombia Comoros Congo Democratic Republic of the Congo Costa Rica Cote D’Ivoire Croatia Cuba Curacao Cyprus Czech Republic Denmark Djibouti Dominica Dominican Republic Ecuador Egypt El Salvador Equatorial Guinea Eritrea Estonia Ethiopia Fiji Finland France French Guiana Gabon Gambia Georgia Germany Ghana Gibraltar Greece Grenada Guam Guatemala Guinea Guinea-Bissau Guyana Haiti Honduras Hong Kong Hungary Iceland India Indonesia Iran

South America Africa Africa Africa South America Africa Europe South America South America Europe Europe Europe Africa South America South America South America Africa South America Africa Africa Europe Africa Southwest Pacific Europe Europe South America Africa Africa Europe Europe Africa Europe Europe South America Southwest Pacific South America Africa Africa South America South America South America Asia Europe Europe Asia Asia Middle East (Continued )

317

318

Appendix

(Continued) Country

World region

Iraq Ireland Israel Italy Jamaica Japan Jordan Kazakhstan Kenya Kiribati Kuwait Kyrgyz Republic Lao People’s Democratic Republic Latvia Lebanon Lesotho Liberia Libya Lithuania Luxembourg Macau Macedonia Madagascar Malawi Malaysia Maldives Mali Malta Martinique Mauritania Mauritius Mexico Moldova Mongolia Montenegro Morocco Mozambique Myanmar (Burma) Namibia Nepal The Netherlands New Zealand Nicaragua Niger Nigeria North Korea Norway

Middle East Europe Middle East Europe South America Asia Middle East Asia Africa Southwest Pacific Middle East Asia Asia Europe Middle East Africa Africa Africa Europe Europe Asia Europe Africa Africa Asia Asia Africa Europe South America Africa Africa South America Europe Asia Europe Africa Africa Asia Africa Asia Europe Southwest Pacific South America Africa Africa Asia Europe (Continued )

Appendix

(Continued) Country

World region

Oman Pakistan Panama Papua New Guinea Paraguay Peru Philippines Poland Portugal Puerto Rico Qatar Reunion Romania Russia Rwanda Samoa Sao Tome and Principe Saudi Arabia Senegal Serbia Seychelles Sierra Leone Singapore Sint Maarten Slovakia Slovenia Solomon Islands Somalia South Africa South Korea South Sudan Spain Sri Lanka St. Kitts and Nevis St. Lucia St. Vincent and the Grenadines Sudan Suriname Swaziland Sweden Switzerland Syria Taiwan Tajikistan Tanzania Thailand Togo

Middle East Asia South America Southwest Pacific South America South America Asia Europe Europe South America Middle East Africa Europe Asia/Europe Africa Southwest Pacific Africa Middle East Africa Europe Africa Africa Asia South America Europe Europe Southwest Pacific Africa Africa Asia Africa Europe Asia South America South America South America Africa South America Africa Europe Europe Middle East Asia Asia Africa Asia Africa (Continued )

319

320

Appendix

(Continued) Country

World region

Tonga Trinidad and Tobago Tunisia Turkey Turkmenistan Tuvalu Uganda Ukraine United Arab Emirates United Kingdom United States of America Uruguay US Virgin Islands Uzbekistan Vanuatu Venezuela Vietnam Zambia Zimbabwe

Southwest Pacific South America Africa Europe Asia Southwest Pacific Africa Europe Middle East Europe North America South America South America Asia Southwest Pacific South America Asia Africa Africa

Index Note: Page numbers followed by “f” and “t” refer to figures and tables, respectively.

A Abu Dhabi (AUH) airport, 95 97 Addis Ababa (ADD) airport, 95 97 Adelaide (ADL) airport, 95 97, 279 280 African airports, classification of according to mitigation strategies, 271f Aircraft size, 204 205, 205f, 214 221, 214f, 217t Airfares, 137 139, 155, 157 Air-fare variable, 133 134 Air navigation service providers (ANSP), 15 16, 108 109 Air passenger demand model, 155f, 299 Airport and Airspace Simulation Model, 27 Airport capacity enlargements, 301 302 Airport capacity improvements increasing aircraft size and adding new runways for, 260 266 between 2030 and 2040, 263 266 up to 2030, 260 263 Airport capacity utilisation, 231 Airport classes, 166 Airport congestion, 87, 102, 205, 298 Airport Dallas/Fort Worth (DFW), 70 71 Airport Delay Simulation Model, 27 Airport Machine, 27 Airspace surrounding airports, 12 13 Air traffic, 3 4, 22, 66, 303 Air traffic control (ATC), 10 11, 14, 109 110, 173 174 Air traffic distribution, regional characteristics of, 74 87 Air traffic management (ATM), 14 16 Air traffic ranking curves, 34 43 Amsterdam Schiphol (AMS) airport, 166 168, 179 181 Analytical approaches, 20 21 Annual airport capacity, 162 163 Annual capacities, 8 12, 32 Annual flight volume

growth rates for the time period from 2030 to 2040, 253t in 2016 by world region, 226 227, 227f Annual passenger volume, 225 by world region in 2016, 226f Annual service volume (ASV), 8 9, 88 89, 162 163, 166, 168 173, 175 177, 187 Antalya (AYT) airport, 207, 220 Asian airports, 226, 292 293 classification of, 273f Asian domestic volume, 235 Atlanta Hartsfield Jackson (ATL) airport, 54 55, 66, 94, 162 163, 167 168, 173 174, 240 245, 276 277, 291, 298 299, 303 Average hour volume, 50 51

B BAA, 42 43 Bangkok (DMV) airport, 97 Barcelona (BCN) airport, 97 Beijing Capital City (PEK) airport, 32, 94 95, 126 130, 163 164, 168, 179 180, 272 274, 283, 283f, 285f, 286f, 289 290, 295, 298 299 average load factor per flight at, 130f average seat capacity per flight at, 129f capacity utilisation development at, 55 61, 59f development of capacity utilisation by capacity utilisation class at, 97 101 forecast number of aircraft movements for 2030 and 2040 for, 285 286, 286f hourly traffic pattern at, 128f unaccommodated passenger demand at, 284 285 Beijing Daxing (PKX) airport, 272 273, 282 283, 283f, 285f, 286f, 289 290

321

322

Index

Beijing Daxing (PKX) airport (Continued) forecast number of aircraft movements for 2030 and 2040 for, 285 286 unaccommodated passenger demand at, 284 285 Benchmarking, 18, 21, 28 29 Bengaluru Kempegowda (BLR) airport, 173 174, 240 245 Blumstein formulations, 17 Bogota (BOG) airport, 97 Boston (BOS) airport, 95 97 British Airport Authority, 8 BROAD, 194 196 Buenos Aires (AEP) airport, 95 97

C Cairo (CAI) airport, 95 97 Cancun (CUN) airport, 95 97, 207 211, 248 249 Capacity, 7 8 approaches to estimate runway capacity, 16 27 analytical approaches, 20 21 empirical approaches, 21 27 queuing models, 18 19 simulation models, 27 factors influencing, 12 16 meaning of, 8 12 Capacity constrained airports and underutilised airports, 92 97 Capacity coverage charts (CCCs), 10 11 Capacity envelope, 17, 17f, 168 Capacity gain by employing larger aircraft, in 2030, 262f by employing larger aircraft, in 2040, 265f by improving runway utilisation and enlargements, in 2040, 265f by improving runway utilisation and runway enlargements, in 2030, 262f Capacity gap, 197 198, 201t Capacity models, 18 Capacity utilisation, 31, 297 air traffic ranking curves, 34 43 capacity utilisation index, 48 55 measure of, 31 34 peak hour volume and annual service volume, 43 48 San Diego (SAN), London Heathrow (LHR) and Beijing (PEK), 55 61 Capacity utilisation classes, 90 92

Capacity utilisation index (CUI), 31, 48 55, 62, 162 163, 165 166, 178 187, 179f, 180t, 297 298 Cape Town (CPT) airport, 97 Cargo flights, 120 Ceteris paribus, 154 155 Chengdu (CTU) airport, 97 Chicago O’Hare (ORD) airport, 66, 94, 164, 166 168, 176 177, 199 200, 276 277, 298 299, 303 Cobb Douglas production function, 135 Compound annual growth rate (CAGR), 74, 189, 197, 229 230, 238 Computer-based mathematical models, 17 Congested airports, 119 120 Congested to uncongested airports, flights diversion from, 117 Congestion pricing, 230 231, 302 Consensus approach, 22, 28 29 Constrained airports worldwide in 2030, 305t in 2040, 306t Constrained and under-utilised airports, 65 Beijing (PEK) transport, 97 101 capacity constrained airports and underutilised airports, 92 97 description of levels of airport capacity constraint, 87 92 development of global air traffic distribution over time, 70 74 London Heathrow (LHR) transport, 97 101 point-to-point versus hub networks, 68 70 regional characteristics of air traffic distribution, 74 87 San Diego (SAN) transport, 97 101 Constraint analysis, 296 Coordination committee, 25, 29 Copenhagen (CPH) airport, 95 97 Country ties, 137 139 Cumulative distribution of air traffic, 49, 65, 71 72, 72f, 84 85

D Dallas/Fort Worth (DFW) airport, 166 168, 175 176, 277 Data envelopment analysis (DEA), 88 89, 163 167, 169t, 173 175, 206 207, 211, 214, 299 300 Decision-making units (DMUs), 163 Declared capacity, 9, 19, 25, 29 Delhi Indira Gandhi (DEL) airport, 256

Index Denpasar (DPS) airport, 97 Denver (DEN) airport, 22, 166 168, 179 182 Description of levels of airport capacity constraint, 87 92 Detroit Metropolitan Wayne County (DTW), 166 167 Digital Aeronautical Flight Information Files (DAFIF), 111 Direct capacity constraints, 236, 284 Doha (DOH) airport, 97 Dubai (DXB) airport, 94, 168, 207 211, 220, 240 245, 256, 275, 281 282, 308 Dublin (DUB) airport, 94 Du¨sseldorf (DUS) airport, 97, 182 capacity utilisation index, 55

E Empirical approaches, 21 27 EU/IATA slot coordination, 24 25 Eurocontrol, 189 European airports, 274 classification of, 274f traffic of, 274, 281 European Council Regulation No 95/93, 22 24 Event-based simulations, 27 Expanding airport capacity, 232, 235 236, 252, 256 259

F Facility Reported Rate, 21 Federal Aviation Administration (FAA), 8 9, 13 5% peak hour capacity, 90, 102 103 5% peak hour volume, 31, 38 39, 51 52, 89, 162 163, 173, 178 179, 181 188, 181t, 183t, 186t Flight rerouting and diverting options, 301 302 Forecast annual flight volume for year 2030 by world region, 238f for year 2040 by world region, 251f Forecast annual flight volume growth rates per year for the period of 2016 30 by world region, 239t for the period of 2030 40 by world region, 253t Forecast annual passenger volume

323

for the year 2030 by world region, 232, 233f for the year 2040 by world region, 246, 246f Forecast annual passenger volume growth rates, 233 235, 248 for the period 2016 30 by world region, 234t for the period 2030 40 by world region, 247t Forecast assumptions, for 2030 and 2040, 229 232 Forecast average aircraft size growth rates per year for the period 2016 30 by world region, 242t for the period 2030 40 by world region, 255t Forecast efficacy, 134 135, 144 145 Forecast model, 299 Forecast of average number of passengers per flight in 2030 by world region, 241t in 2040 by world region, 254t Forecast unaccommodated annual passenger volume for the year 2030 by world region, 235 236, 235f for the year 2040 by world region, 248 249, 249f Forecast ‘unconstrained’ passenger demand, 235 236 Four capacity utilization classes, 90 91 ‘Fraction’ approach, 164 Frankfurt (FRA) airport, 35 38, 41f, 43, 48 49, 97, 118 119, 119f, 190, 193, 215 216, 274 275 Frankfurt/Paris Charles de Gaulle, 199 200 Fukuoka (FUK) airport, 97 Future air passenger demand, modelling, 133 background, 133 135 model application: comparing different forecasts, 155 158 model estimation and testing, 141 155 model theory, 135 141 Future airport capacity and capacity utilisation, 161 background, 162 163 comparison of model results with actual traffic data, 178 187 model theory and parameter estimation, 163 178

324

Index

Future airport capacity enlargements and limits, modelling, 189 background, 190 model theory and parameter estimation, 190 197 potential impact of limited airport capacity on future growth of flight volume, 197 203

G Geneva airport, 89 92, 274 275 relative traffic ranking function of, 91f German airports, 33 35, 91 92 Gini coefficient, 74, 74t Glasgow (GLA) airport, 95 97 Global airport network, 72, 295 Global air traffic, 3, 4f, 107, 161 162, 189 Global air traffic distribution, development of, 70 74 Global air traffic forecast results for the years 2030 and 2040, 304t Global constraint analysis, 103 Global economic development, 229 230 Global flight volume, 227, 264 Global mean value, 227 229 Global passenger volume, 227, 232 Grandfather right, 24 25 Gravity models, 133 136, 134t, 144, 149 150 Guangzhou (CAN) airport, 94

H Handbook method, 8 9 Hanover airport, 51 52 capacity utilisation index, 51 52, 52f Ho Chi Minh (SGN) airport, 97 Hong Kong (HKG) airport, 94, 134, 168, 240 245 Honolulu (HNL) airport, 95 97 Hourly capacity, 8 12, 23t Hub airports, 32, 47, 79 80 Hub-and-spoke (HS) system, 68 69 Hub routes (Hub), 153 154 Hyderabad (HYD) airport, 95 97

I Indirect capacity constraints, 236 Infrastructural characteristics of runways, 15 Infrastructure, airport-related, 108 109

Instrument flight rules (IFR), 8 9, 14, 20 21, 39 40, 42 43, 189 Instrument meteorological conditions, 14 Intercity trains, 108 109 International Air Transport Association (IATA) slot coordination, 109 110, 116 117, 302 International Civil Aviation Organisation (ICAO), 13 Intraregional flights, 227 average number of passengers per flight for, 229 Intraregional traffic, 226 Investment in new runways, 110 115 Istanbul (IST) airport, 298 299 Istanbul Sabiha Go¨kcen (SAW), 94

J Jakarta Soekarno Hatta (CGK) airport, 94, 167, 173 174, 240 245, 256, 291, 303 Jeju (CJU) airport, 273 Jensen’s inequality, 136 Johannesburg (JNB) airport, 272 Juanda (SUB) airport, 173 174

K Kuala Lumpur (KUL) airport, 97

L Las Vegas (LAS) airport, 95 97 Less-congested airports, diversion to, 116 117 Level 3 airports, 116 117 Liberalised markets, airlines in, 119 120 Lima Jorge Chavez (LIM) airport, 173 174 Limited airport capacity, 4, 197 203 Load factor per flight, 119 126 Load factors, modelling, 231 Logit model, 191 192, 194 Log-likelihood function, 144 London Gatwick (LGW) airport, 94, 117 118, 118f, 167 168, 173 174, 207 211, 220, 274 275, 298 299 London Heathrow (LHR) airport, 7 8, 32 33, 39, 94 95, 126 130, 162 164, 167, 199 200, 215 216, 232, 240 245, 274 275, 283, 291, 295, 297 299, 303

Index actual and forecast number of annual aircraft movements for, 286f actual and forecast annual passenger volume for, 283f air traffic ranking curve of, 39f average load factor per flight at, 130f average seat capacity per flight at, 129f capacity utilisation development at, 55 61, 57f capacity utilisation index, 54 55, 297f development of capacity utilisation by capacity utilisation class at, 97 101 forecast number of aircraft movements for 2030 and 2040 for, 285 286 forecast unaccommodated annual passenger volume for, 285f hourly traffic pattern at, 127f third runway at, 292 London Heathrow airport, 111, 179 181 London Stansted (STN), 35 38, 42 43 capacity utilisation index, 51, 51f, 297f Longitudinal separation minima, 15 Long-term infrastructure planning, 10 11 Lorenz curve, 49 Los Angeles (LAX) airport, 94, 168, 298 299 Low-cost carriers (LCCs), 3 4, 69 70, 126 Low-frequency (LF) routes, 153 154

M Manchester (MAN) airport, 97 Manila (MNL) airport, 97 Market Intelligence (MI) data, 133 134 Markov chains, 190 191 Maximum average hourly volume, 89, 187 McFadden’s pseudo-R2, 144, 212 213 Mexico City (MEX) airport, 94, 248 249, 278 279, 281 282, 298 299, 308 Miami (MIA) airport, 97, 240 245 Middle East and Munich (MUC) airport, 94 Middle East region, 275 classification of airports of, 276f Milan Linate (LIN) airport, 95 97 Minneapolis (MSP) airport, 97 Mitigation measures, analysis of, 302 Mitigation strategies, assessing, 259 282 airport capacity improvements up to 2030 and 2040, 260 266 general mitigation measures in world regions, 266 282 Moscow (SVO) airport, 97

325

Mumbai Chhatrapati Shivaji (BOM) airport, 256 Munich airport, 17, 298 299

N Nairobi (NBO) airport, 95 97, 272 Newark airport, 298 299 New runways, 259, 261, 263, 267 270, 300 302, 307 308 investment in, 110 115 New York (JFK) airport, 97, 207 211, 215 New York LaGuardia (LGA) airport, 94 95 Nice (NCE) airport, 97 9/11 crisis, 3 4 Nolens volens measure, 117 Nominal airfare development, 230 231 North American airports, 226 air traffic volumes of, 281 classification of, 277f North American market, 66 68 Null hypothesis, 144

O Off-diagonal values, 229 Official Airline Guide (OAG), 34 35, 42 Off-peak period, 32, 116 119, 131 Operational measures, 109 Ordinary least squares (OLS) method, 134 136, 147 149 Origin destination (OD) demand, 68 69, 133, 140 Origin destination passenger volume (ODP), 150 ODP elasticity of TP, 150 ODP share (ODPS), 150 152 Osaka (ITM) airport, 273 OurAirports, 111

P Palma de Mallorca (PMI) in Spain, 35 37, 97 PART, 194 196 Passengers per flight, 205 background, 205 206 model theory and parameter estimation, 206 214 top 50 routes in terms of, in 2016, 208t 2008 16 projection of aircraft size on 40 sample routes, 215 221 Peak hour volume and annual service volume, 43 48

326

Index

Peking Capital City (PEK) airport, 181, 215 216 Peking Daxing (PKX), 200 Perimeter rule, 109 Planning approval procedure, 7 8 Point-to-point (P2P) system, 68 69, 153 154 versus hub networks, 68 70 Poisson distribution, 136 137 Poisson pseudo maximum likelihood (PPML) estimator, 133 134, 136, 141 149, 158 Poisson-type distributed arrivals and departures, 18 Practical hourly capacity (PHCAP), 10 11, 18 19, 27 28 Precipitation, 14 15 Primary slot trading, 302 Pseudo-R2, 144, 194

Q Queuing models, 18 19

R

RAILKM2, 194 196 Realisation probability (RP), 191 192, 194, 196 197 Regression analysis, 163, 206 Regression-based tests, 144 Regulatory measures, 109 110 Relative traffic ranking curve, 91 92, 98 100, 98f, 99f Rerouting traffic to under-utilised airports and/or using more off-peak hours, 116 119 Residual sum of squares (RSS), 145, 145f Rio de Janeiro-Galeao (GIG) airport, 279 Riyadh (RUH) airport, 95 97, 275 Rome (FCO) airport, 97 Runway capacity, methods of estimation of, 9, 16 27 Runway Delay Simulation Model, 27 Runway expansion, 190, 193, 196 197, 199 200, 199f, 200f Runway extensions at airports, 111, 112t RunwaySimulator, 21 Runway system, 162 164, 166 167, 175, 181 182

S Sabre Market Intelligence (MI) data, 133 134

San Diego (SAN) airport, 20 22, 25 27, 126 130, 173 174, 277, 283 284, 283f, 285f, 286f average load factor per flight at, 130f average seat capacity per flight at, 129f capacity utilisation development at, 55 61, 57f, 58f development of capacity utilisation by capacity utilisation class at, 97 101 forecast number of aircraft movements for 2030 and 2040 for, 285 286 hourly traffic pattern at, 127f unaccommodated passenger demand at, 284 285 San Francisco (SFO) airport, 97, 182 Santos Dumont (SDU) airport, 279 Sao Paulo (GRU) airport, 97 Sao Paulo Congonhas (CGH) airport, 248 249 Saturation capacity, 10 Seat capacity, raising, 119 126, 121t Secondary slot trading, 302 Seoul (ICN) airport, 95 97 Separation minima, 12 13, 14t Shanghai (SHA) airport, 95 97 Short-term actions, 10 11 SIMMOD, 27 Simulation models, 22, 27 Singapore Changi (SIN) airport, 168, 256 Single runway airport, 11 12 Slot coordinator, 25, 117 South American airports, 277 278 classification of, 278f Southwest Pacific airports, 279 classification of, 280f Spatial distribution of air traffic, 65 Strategic airport planning, 8 Strategies for mitigating airport capacity constraints, 107 Beijing (PEK) airport (case study), 126 130 load factor per flight, 119 126 London Heathrow (LHR) airport (case study), 126 130 new runways, investment in, 110 115 rerouting traffic to under-utilised airports and/or using more off-peak hours, 116 119 San Diego (SAN) airport (case study), 126 130 seat capacity, raising, 119 126, 121t typology of mitigation measures, 107 110

Index Stuttgart (STR) airport, 274 275 Sydney (SYD) airport, 97, 279 282

T Technological innovations, 231, 302 Tel Aviv (TLV) airport, 95 97 Temporal distribution of air traffic, 65 Terminal area airspace, 12 13 Terminal control airspace, 12 13 Theoretical capacity, 10, 13 14 Total Airspace and Airport Modeller, 27 Total passenger volume (TP), 225 growth rates, 149 Tourism receipts, 147 Traffic, rerouting to under-utilised airports, 116 119 Traffic forecast, 225 forecast assumptions for 2030 and 2040, 229 232 mitigation strategies, assessing, 259 282 airport capacity improvements up to 2030 and 2040, 260 266 general mitigation measures in world regions, 266 282 for San Diego (SAN), London Heathrow (LHR), Beijing Capital City (PEK) and Beijing Daxing (PKX), 282 290 aircraft size, 287 288 capacity analyses, 288 290 number of aircraft movements, 285 286 passenger volume, 283 285 for 2016 30, 232 245 aircraft size, 240 flight volume, 236 240 passenger volume, 232 236 top 20 airports in terms of unaccommodated demand in 2030, 240 245 for 2030 40 aircraft size, 252 256 flight volume, 251 252 passenger volume, 246 251 top 20 airports in terms of unaccommodated demand in 2040, 256 259

327

traffic situation in 2016, 225 229 aircraft size, 227 229 flight volume, 226 227 passenger volume, 225 226 Traffic growth, 28, 76t, 77, 100 101, 115, 303 307 Traffic ranking curves, 32, 62, 90, 297 Traffic ranking functions, 62, 296 298 Traffic volumes, 19, 297 298 ‘Typical peak hours’, 33 34 Typology of mitigation measures, 107 110

U Ultimate capacity, 10 Unaccommodated passenger volume, 232, 235 236, 245 forecast share of for the year 2030 by world region, 237t for the year 2040 by world region, 250t top 20 airports in terms of in 2030, 243t in 2040, 257t Unconstrained passenger demand, 235 236 Under-utilised airports, 92 97 rerouting traffic to, 116 119 Urumqi (URC) airport, 97

V Vancouver (YVR) airport, 95 97 Vienna (VIE) airport, 97 Visual flight rules (VFR), 8 9, 14, 20 21, 25 27 Visual meteorological conditions, 14

W Weather conditions, impact on runway capacity, 14 15 Wide-body aircraft, 15 World Development Indicators (WDI), 137 139