Studies, Vol.12, 2017 the JICA study, is not environmentally feasible due to land constraints. Therefore, since 2013 the multi-airport system (MAS) ha

Transcription

1 Studies, Vol.12, 2017 An Analysis of The Greater Jakarta Metropolitan Area Multi-Airport System Pintanugra PERSADANTA a, DEWANTI b a Airport System and Environmental Division, Directorate General of Civil Aviation, Ministry of Transportation, Indonesia; pintarigan@yahoo.com b Department of Civil and Environmental Engineering, Universitas Gadjah Mada, Yogyakarta, Indonesia; dewanti@ugm.ac.id Abstract: Japan International Cooperation Agency in 2012 has conducted a research regarding multi-airport systems in Greater Jakarta which stated that a new airport should be operating in this metropolitan area by 2019 to accommodate future demand. According to the 2012 study, no airports except Soekarno-Hatta Airport were able to handle commercial flights in the city due to regulatory and environmental issues. However, this is not the case. Halim Perdanakusuma Airport, now a secondary airport, has been operating commercially since 2014 and another secondary airport, Pondok Cabe Airport, has the potential to be operational in the future. New air demand forecasts and air traffic distributions with different scenarios have been produced and analysed to reflect current situation in Greater Jakarta. The result of these analyses differs from the Japan International Cooperation Agency study and shows that the operation of the new airport can now be postponed until 2022 at the earliest. Keywords: Market Analysis, Demand Growth Drivers, Traffic Forecast, Sensitivity Analysis, Air Traffic Distributions 1. INTRODUCTION From 2010 to 2012, the Japan International Cooperation Agency (JICA) conducted a master plan study on multiple-airport development for Greater Jakarta Metropolitan Area (GJMA). The background of the JICA study was the overcapacity experienced by the existing terminals in Soekarno-Hatta Airport (CGK). CGK plays a vital role in Indonesia s airport system as a primary hub airport, operating as the main international gateway for the country and the main hub for an extensive amount of domestic air routes. Apart from CGK, there are three other airports in GJMA namely Halim Perdanakusuma (HLP), Pondok Cabe (PCB) and Budiarto (BDR). However, CGK was the only airport which operated commercially at GJMA when the JICA study was conducted. The JICA study (2012) concludes that CGK does not have enough capacity to accommodate future air transport demand and cannot be expanded due to environmental reasons. JICA also suggests that HLP, PCB and BDR cause some negative environmental issues with regard to airport operation and thus cannot be included in the GJMA airport system. Therefore, JICA recommends the Government of Indonesia to establish and operate a new major airport at Karawang in However, the situation in GJMA after the JICA study has moved beyond JICA s assumptions. In 2013, the GOI opened HLP and together with CGK serves the air transport demand for GJMA. The GOI through its Ministry of Transport published Ministerial Decree 369/2013 which withdrew the previous regulation that restricts HLP serving scheduled commercial flights. Moreover, CGK continues to develop the third runway, which based on 632

2 Studies, Vol.12, 2017 the JICA study, is not environmentally feasible due to land constraints. Therefore, since 2013 the multi-airport system (MAS) has already been implemented in GJMA through CGK and HLP to serve air demand in the region. The possibility of the other two non-commercial airports in GJMA, BDR and PCB, serving scheduled commercial flights will also be analysed in this research. By looking at the conditions mentioned above, if the existing airports in GJMA can form a strong system to accommodate the air demand, the urgency for a new airport in 2019 is questionable and will be analysed in this research. 2. LITERATURE REVIEW 2.1 Definition of Multi-Airport Systems According to Horonjeff et al. (2010), an airport system is an integrated system of airports assessed to meet the future air transport demand of an area, region or country. This airport system, where some airports serve the same region, is called a MAS. Theoretically, a MAS is an airport system in which more than one airport serving commercial traffic operates in the same metropolitan area without regard to the ownership or political control of the individual airports (de Neufville et al., 2013). Each airport in a MAS varies in size and characteristics. De Neufville (1995) argues that each airport concentrates on its own market. Therefore, it is critical to perceive the dynamics of the different types of airport within MASs. MASs consist of two airport types: primary and secondary. The former is described as an airport that handles more than 20% of the total passengers in a MAS, while the latter handles less than 20% of the total passengers in the same MAS (Bonnefoy et al., 2010) 2.2 Dynamics Evolution of Multi-Airport Systems MASs have been a model for all metropolitan regions with the originating traffic above a specific traffic threshold. The level of originating traffic has not been constant and is likely to change over the coming years. As of 2010, the minimum level of originating traffic was about 15 million passengers per annum (MPA) for the entire metropolitan area increasing from 10 million originating passengers per year in 1995 (de Neufville et al., 2013). Metropolitan areas with the originating traffic less than the threshold level are still able to develop MASs with technical or political judgments that force these airports to exist. The evolution of MASs is influenced by the following three main factors: the readiness of existing airport infrastructures, the entry of low-cost carriers (LCCs) at underutilised airports, and regulatory and political factors. ICAO (2008) and FAA (2007a) identify two methods regarding the evolution of MASs: (i) optimising the availability of the existing airports; or (ii) building new airports. 2.3 Airport Choice in a Multi-Airport System Airports within a MAS can experience either competition or cooperation among them. Airports in MASs can compete in many ways to attract airlines and passengers if offering a better service compared to the others (Graham, 2014). There are various past studies regarding factors affecting the choice of airports in MASs based on the perspectives of passengers and airlines. 633

3 Studies, Vol.12, The passengers perspective Brooke et al. (1994) state that in a MAS, passengers are able to, and do, make a selection of which airports to fly from due to the relative service offerings at competing airports. Passengers with different travel reasons, either business or leisure, have different factors in selecting airports (Başar and Bhat, 2004). A secondary airport is attractive to passengers if it delivers satisfactory access to the desired air services (de Neufville, 1995) The airlines perspective FAA (2013) states that the business model of an airline affects the airport selection within a MAS. The determining factors are demographic and economic characteristics, the market split between leisure-business travellers and the area in which the airport is located. Moreover, the increase in low-cost airlines is causing a more competitive competition between airports in a MAS where the airports are located relatively close together. 2.4 Traffic Distribution in Multi-airport Systems A MAS generally consists of a primary airport and one or more secondary airports. The primary airport has a high concentration of traffic whilst the secondary airports have considerably less (de Neufville, 2000). The development of traffic at secondary airports will be stronger if it serves an exclusive market that it does not have to share with the primary airport. This specialisation can occur through government regulation, historical precedents, geography and airline strategy (de Neufville, 2000). In many cases, governments or airport operators have tried to force traffic to move from a congested airport to those secondary airports with underutilised capacity (de Neufville et al., 2013). However, they do not have control over forming the air traffic allocation between airports and the market served by each airport has various components with different needs and an unpredictable future. Thus the efforts are impractical except in limited conditions. The dynamics of the market will naturally shift the traffic to secondary airports if the primary airport has exceeded its designed capacity, or due to technical constraints at the primary airport. 2.5 Role of Regulatory and Political Factors Regulatory aspects have taken on an important role in the distribution of traffic among airports in MASs (Bonnefoy, 2008). This solution is often used in order to force the distribution of traffic. In 2013, the Indonesian Ministry of Transport produced a decree regarding the National Airport System which regulates the role, function, hierarchy and classification of Indonesian airports (MoT, 2013). The airport hierarchy in Indonesia is divided into hub and spoke airports. A hub airport is categorised into primary, secondary and tertiary hubs. A primary hub is an airport handling over 5.0 MPA; secondary hubs serve between 1.0 and 5.0 MPA; tertiary hubs handle between 0.5 and 1.0 MPA. The key aim of regulating airport systems is to deliver guidelines for the development of airports in a logical, sustainable and cost-effective manner (IATA, 2004; FAA, 2007b). Therefore, the development of airports in MASs should be financially feasible to accommodate not only future demand but also the development of communities, other airports in the region, other transport modes and the environment (IATA, 2004; Wells and Young, 2004; FAA, 2007b). These regulatory tools can be useful to maintain the operations at 634

4 Studies, Vol.12, 2017 primary airports whilst securing the successful development of secondary airports. 2.6 Planning and Developing Multi-Airport Systems Successful planning of MASs requires knowledge of the dynamic competition between airports in the region; misunderstanding how the MAS evolves will lead to expensive and incorrect planning. According to de Neufville et al. (2013), some issues arise from the development of MASs: (i) an unnecessary and embarrassingly massive expenditure due to insufficient traffic at new airports; (ii) difficulty in closing old airports due to political and economic reasons; (iii) insufficient traffic to support MASs; (iv) difficulties in allocating traffic from a primary airport to the alternative airports in the region; and (v) traffic volatility at secondary airports makes the planning process difficult, i.e. risky to invest and potentially unprofitable. Consequently, forecasting air transport demand is considered to be one of the main elements for avoiding excessive investment in the development of secondary airports (Wells and Young, 2004; de Neufville at al., 2013). There are two common forecasting methods: time series and econometric models. Time series is used if there were no structural changes in the market and if the predicted future has the same characteristics as the past. This method should not be used for forecasting more than two years to avoid drastic fluctuations in the underlying conditions in the future. However, the forecast result may lead the airport to plan for failure due to bias and uncertain conditions (Kwakkel, 2008). It also makes investments in second airports more risky in the short run (de Neufville, 1995). The airport planners should identify the risks, opportunities and key factors affecting air traffic growth based on the market analysis to minimise deviation. Therefore, it is important to adopt dynamic strategic planning and be flexible in developing secondary airports (Bonnefoy et al., 2010). 3. METHODOLOGY 3.1 Research Structure The methodology used in this research employs the following four-step method (Figure 1): (i) identify the existing airport system at GJMA and its characteristics; (ii) analyse the dynamics that govern the existing airport system; (iii) forecast the future traffic demand based on the selected air transport demand drivers; and (iv) analyse the implications of any possible traffic scenarios in order to deliver recommendations for the effective development of the future MAS in GJMA. Figure 1. Research methodology 635

5 Studies, Vol.12, Research Process This research starts with a comprehensive literature review to obtain a deep understanding of MASs worldwide which will be adopted by Greater Jakarta s airport system. Secondly, the data and information about GJMA and its airports are gathered and analysed. Furthermore, the future air traffic demand in GJMA will be forecast by integrating top-down and bottom-up approaches (Figure 2). The top-down approach is based on the identification of the aviation demand drivers, including regression analysis against Indonesia GDP per capita, local currency exchange rate and the analysis of past behaviour changes in the Indonesian airport market environment. The bottom-up analysis aims to check the appropriateness of the traffic forecast in line with supporting indicators such as modal competition, airline strategies and regulatory-policy factors. Figure 2. Forecast methodology Furthermore, a sensitivity analysis is conducted by estimating the future passenger traffic under low, medium and high growth assumptions to obtain an alternative assessment for each case. It is then followed by the analysis of traffic distribution within MASs. The findings from the airport development plans and the prediction of air transport demand in GJMA will then be analysed using a SWOT analysis to provide conclusions and recommendations about the development of a new airport in Greater Jakarta. 3.3 Research Data This research covers a wide range of data such as airport and air traffic data, geographical characteristics of Greater Jakarta, GDP per capita and the exchange rate. All data are taken from sources such as the Indonesian Ministry of Transport, Japan International Cooperation Agency, PT Angkasa Pura I and PT Angkasa Pura II for airport and air traffic data, the Indonesian Ministry of Home Affairs for geographical characteristics, World Bank for GDP per capita and Bank of Indonesia for the exchange rate. 3.4 Research Area This research covers four existing airports and one greenfield airport located in the municipalities and regencies in GJMA including Soekarno-Hatta Airport, Halim Perdanakusuma Airport, Budiarto Airport, Pondok Cabe Airport and Karawang Airport. This research also analyses the market characteristics in the region such as air transport growth, route development, airport segmentation and traffic distribution between airports. 636

6 4. RESULTS AND DISCUSSION 4.1 Market Analysis Historical air traffic The total air passenger volume in Indonesia has increased approximately seven-fold from 25.8 million passengers in 2000 to million passengers in Airports in GJMA have contributed around 40% of the total annual passengers in all Indonesian airports. However, the traffic growth in GJMA shows a deceleration in recent years as the annual growth reached 5.9% p.a. between 2010 and 2014, down from 15.8% p.a. over the period (Figure 3). Figure 3. Cumulated traffic of GJMA airports Route development analysis Figure 4 shows that the number of passengers on domestic and international flights increased steadily to 2013 then remained relatively constant to 2015; however, the proportion of domestic, international and transit passengers remained constant over the period. Figure 4. Air passengers at GJMA airports As shown in Figure 5, the total annual passenger numbers in GJMA for the last ten years consists of approximately 75% domestic passengers, 20% international passengers and 5% transfer/transit passengers. 637

7 Figure 5. Route segmentation Airport segmentation and traffic distribution As can be seen in Table 1, the majority of traffic at HLP comes from destinations in Java-Bali, Sumatera and ASEAN regions as airlines use this airport particularly for short-haul routes. The airlines at CGK serve more distant routes. This typical condition formed a natural airport segmentation based on trip destination. Table 1. Route proportion at GJMA Airports Domestic Routes HLP CGK Sumatera 5.21% 29.13% Java - Bali 6.08% 26.70% Sulawesi 1.15% 12.18% Kalimantan 1.32% 10.49% Nusa Tenggara - Maluku 0.70% 3.30% Papua 0.39% 3.36% Total Domestic 14.84% 85.16% International Routes HLP CGK ASEAN 2.18% 59.82% Asia 0.00% 20.35% Middle East 0.00% 13.76% Europe 0.00% 2.95% Australia 0.00% 0.93% Total International 2.18% 97.82% HLP shows a significant contribution to GJMA as a secondary airport. HLP began serving scheduled commercial flights in 2014 to seven domestic routes with a single airline (Table 2). HLP carried about 1.6 million passengers during its first year. In 2015, the airport handled around 3.1 million passengers with two airlines flying to ten domestic routes. As a primary airport, the air traffic in GJMA is concentrated at CGK. The demand capacity ratio shows that CGK and HLP currently suffer capacity constraints but have a plan to develop their capacity (Table 3). Table 2. Airport operation progress at HLP Year Number of LCCs Number of FSCs Routes served Passengers carried MPA MPA

8 Table 3. Demand capacity ratio Terminal capacity Passengers carried Demand capacity ratio CGK 22 MPA 55.6 MPA 2.53 HLP 1.5 MPA 3.1 MPA Recent evolution of the airport system in Greater Jakarta With regard to the market analysis, it is essential to understand and evaluate how the competitive advantage of airports in GJMA has evolved over recent years. Contrary to CGK and HLP, neither PCB nor BDR opened for scheduled commercial flights. PCB operates particularly for non-commercial and military flights while BDR is an airport specifically for flying school and aircraft maintenance. However, recent traffic conditions at CGK and HLP have forced the government to optimise the role of PCB and BDR rather than build a new major airport. In the case of PCB, Garuda Indonesia is willing to operate and make PCB a base for short-haul routes. But the government has denied permission due to air navigation issues and surface access constraints. 4.2 Demand Growth Drivers Economic factors Table 4 shows that Indonesia s economic growth heavily affects the air transport demand growth in GJMA. The deceleration of GJMA s air transport growth is relatively in line with the decline in economic growth on a national scale, although much more marked. Table 4. GJMA air traffic growth and Indonesia economic growth Year GJMA Air Traffic Growth (%) Indonesia Economic Growth (%) In addition, the Indonesian GDP is used to explain the GJMA s air traffic demand rather than the gross regional domestic product (GRDP) of GJMA for two main reasons: the GRDP of GJMA contributes 26.31% to Indonesia s GDP and the air traffic demand at GJMA amounts to 33.8% of the total air traffic in Indonesia Strong demographic and geographic characteristics The Indonesia s air transport figures show that this sector has an enormous opportunity to develop. Table 5 presents the propensity to fly in Indonesia which is far below other countries in the ASEAN region. 34% of the population using air transport means that aviation in Indonesia has a great opportunity to be developed and thus airports have a chance to improve the utilisation of the terminal capacity in the future. 639

9 Country Table 5. Proportion of air traffic and population Population (million) Air Traffic (million) Proportion (%) Indonesia Malaysia Thailand Singapore Regulatory and political factors The aviation industry, including airports and airlines, operates in a highly regulated political environment. There are two main regulations that regulate the development of airports in Indonesia: the airports master plans and Ministerial Decree 69/2013 regarding the National Airport System. All airports in GJMA, except PCB, due to its main function as a private airport, are mentioned in the Ministerial Decree. The plan for commercial flights at PCB requires a revision to Ministerial Decree 69/2013. This regulation also states the hierarchy of the new airport in GJMA as a primary hub airport. The ASEAN Single Aviation Market (ASAM) has been implemented at selected airports mentioned in the ASEAN open sky agreements since 1 January Five Indonesian airports took part in the agreement including CGK. Recently, the EU commission updated the EU air safety list and removed three Indonesian airlines from their banned list. Moreover, based on Presidential Decree 104/2015, 75 countries have been given a 30 days free entry to Indonesia. These developments will stimulate international flights, particularly to/from European countries and within the ASEAN region. 4.3 Passenger Traffic Forecast The forecast model in this research aims to predict passenger traffic from 2016 until Therefore, econometric analysis is used in the forecast model by applying multiple linear regression techniques. Domestic and international passengers are forecasted in separate econometric models due to distinct traffic characteristics. The domestic forecast model is assumed to follow the linear model. Meanwhile, the international forecast model follows the logarithmic model, which leads to the expected exponential growth from various supporting factors, such as the implementation of open sky policies and free-entry visas (Micco and Serebrisky, 2006). Boeing (2015) argues that some factors have a greater impact on market performance but are difficult to quantify, such as the economic crisis, deregulation and irregular phenomena (e.g. terrorist attacks). These factors are included in forecast models as dummy variables to measure their effect on air transport demand. Dummy variables have the value 0 or 1 to show the absence or presence of the selected factors. Moreover, this research has examined the trend model to predict future passenger numbers by using the observation years as an explanatory variable but this model has an autocorrelation issue and lower value of adjusted-r 2 compared to other models and is thus excluded from this analysis Domestic forecast model The forecast model for domestic traffic involves GDP per capita, exchange rate and dummy variables such as the events of the economic crisis and deregulation. 640

10 Domestic traffic = GDP per capita Exchange rate Economic crisis Deregulation (1) Table 6 shows the domestic model is statistically significant to measure the number of passengers (Significance F-test < 0.05). All independent variables, except the economic crisis, are statistically significant at the 5% significance level with no issue of multicollinearity (VIF < 10). The economic crisis is still included because it mainly affects people s purchasing power. R 2 -value means over 98% variability in passenger numbers can be explained by the model. Moreover, the Durbin-Watson value (1.573) shows no autocorrelation and there is no heteroscedasticity based on Figure 6. Table 6. Statistical test domestic model International forecast model Figure 6. Heteroscedasticity test domestic model The forecast model for international traffic involves GDP per capita, exchange rate and dummy variables such as the events of economic crisis and irregular phenomena such as terrorist attacks. ln(international traffic) = ln(gdp per capita) ln(exchange rate) Economic crisis Irregular phenomena (2) 641

11 Table 7 shows the international model is statistically significant to measure the number of passengers (Significance F-test < 0.05). All independent variables, except the irregular phenomena, are statistically significant at 5% significance level with no issue of multicollinearity (VIF < 10). Irregular phenomena are still included because they mainly affect international travellers going to a country, particularly tourists. R 2 -value means over 98% variability in passenger numbers can be explained by the model. Moreover, the Durbin-Watson value (1.867) shows no autocorrelation and there is no heteroscedasticity based on Figure 7. Table 7. Statistical test international model Figure 7. Heteroscedasticity test international model The Indonesian GDP per capita is used to predict the international traffic, even though the international traffic consists of foreign and Indonesian passengers, due to the fact that the proportion of foreign passengers is only 19% of the total international traffic for the last 15 years or equivalently 3.9% of the total passengers in GJMA airports. Therefore Indonesian GDP per capita relatively suits to reflect the passenger purchasing power to travel abroad Sensitivity analysis of air transport demand The future GDP and exchange rate for the long-term is predicted based on the government s economic assumptions and the historical average growth divided into low, moderate and high growth assumptions (Table 8). These assumptions are then used to conduct the sensitivity analysis. 642

12 Table 8. Forecast framework GDP per capita Exchange rate Case 1 Case 2 Case 3 Case 1 Case 2 Case 3 Annual growth ratio 3% 4% 5% 2% 3% 4% Gradual decrease in GDP/ increase in exchange rate over 15 years 1% 1% 1% 1% 1% 1% Based on the scenarios in Table 8, the number of domestic and international passengers until 2030 can be forecasted. Table 9 and Figure 8 present total future passenger numbers at GJMA by adding domestic and international forecasts. Transit/transfer passengers are not examined in this research but the future growth is assumed to remain at approximately 5% of the total domestic and international passengers. Table 9. Air transport demand forecast in GJMA Air Historical Data Forecast passengers Case 1 72,107,836 87,044, ,757,805 Case 2 41,849,990 56,816,762 77,812, ,650, ,477,196 Case 3 83,795, ,784, ,517,483 Figure 8. Passenger forecast result The annual growth ratio from 2015 to 2030 is 3.9% (Case 1) to 6.8% (Case 3). This growth ratio is in line with the range of 5.9% historical growth (Table 10). Table 10. GJMA Annual Demand Growth Annual Growth (% p.a.) Case 1 Historical Data Forecast Case Case

13 Studies, Vol.12, Comparison with official forecasts Figure 9 presents a comparison of several official forecasts and the research s forecast of the total domestic-international passengers at GJMA, without regard to transit-transfer passengers. However, these official forecasts solely examine CGK as the only commercial airport serving GJMA without considering the possibility of commercialising other airports. Figure 9. Forecasts comparison Sensitivity analysis of air transport supply Since 2008 the plan to develop the terminal capacity of CGK up to 87 million has been approved by the Ministry of Transport and stated on its master plan. However, some issues arose during the implementation regarding socio-economic and environmental issues, especially with regard to land acquisitions. In contrast with CGK, HLP is facing difficulties in expanding its capacity due to its status as a military base. The airport currently has one passenger terminal with a capacity of 1.5 MPA; in 2015, the airport served around 3.1 MPA. With the aim of not competing with its sister company, Citilink, Garuda Indonesia has recently planned to use PCB as their ATR-72 main base, particularly for short-haul routes. As explained previously, the main issue in PCB is the cross airspace with HLP. The solution is to control PCB and HLP under one air navigation controller. Garuda Indonesia, Indonesian Directorate General of Civil Aviation (DGCA) and Angkasa Pura II have been conducting simultaneous meetings to resolve the issue. The opening of PCB will increase the capacity of the GJMA by around 1 MPA. The main functions of BDR, an airport owned and operated by the Indonesian DGCA, are for flight training purposes and maintenance organisations. At the moment, the government has no plans to open BDR for commercial flights. Based on the situations mentioned above, four scenarios regarding total airport capacity in GJMA are examined: Scenario 1: 93 MPA (CGK: 87 MPA, HLP: 5 MPA, PCB: 1 MPA) Scenario 2: 92 MPA (CGK: 87 MPA, HLP: 5 MPA) Scenario 3: 88.5 MPA (CGK: 87 MPA, HLP: 1.5 MPA) Scenario 4: 63.5 MPA (CGK: 62 MPA, HLP: 1.5 MPA) The sensitivity analysis is undertaken for all scenarios using the forecasted results to estimate the year in which airports in GJMA reach their maximum capacity and therefore need to build a new airport (Table 11). 644

14 Studies, Vol.12, 2017 Table 11. Year GJMA airports reach maximum capacity Case 1 (low) Case 2 (moderate) Case 3 (high) Scenario Scenario Scenario Scenario Scenario 4 is a do-nothing scheme where CGK cannot continue the construction of the 4th terminal. Scenario 3 contains the ultimate capacity of CGK with no additional capacity at HLP. Scenarios 1 and 2 assume that the Air Force approves the airport operator of HLP expanding the terminal capacity. To obtain permission from the Air Force will be a difficult task since HLP is also used as a military base. Figure 10. Development phase of terminal capacity Scenario 1 Figure 11. Development phase of terminal capacity Scenario 3 645

15 Considering instructions from the elected President to develop airport infrastructures, Scenario 3 is selected as the most likely capacity at GJMA airports whilst Scenario 4 is the least likely airport development in GJMA as the government strongly supports CGK to achieve its maximum capacity. Therefore, analysis in this research uses Scenario 3 as a base case scenario whilst encouraging the airport capacity optimisation scheme by implementing Scenario 1. The supply-demand analysis using Scenarios 1 and 3 against three different traffic growth assumptions is explained in Figure 10 and Figure Air Traffic Distribution Traffic distribution between airports varies according to the choice of preferred airport. Scenarios 1 and 3 are selected to describe the optimum and the most likely supply capacity for all GJMA airports. Meanwhile, the aim to adopt the high demand growth is to examine the airport potential of handling maximum demand. Moreover, it is assumed that the moderate and low growth rates will deliver a lower impact on the airport capacity than the high demand growth. These also give a longer time for the airport to maintain its capacity before constructing the next development phase Condition-1: scenario 1, high demand, preferred airport: HLP-PCB This condition assumes the air navigation issue at PCB is resolved by applying joint control under HLP s controller and thus can serve commercial flights. If HLP and PCB are the preferred airports, the traffic at HLP and PCB is assumed to be equal to their capacity and CGK receives the remaining traffic. Thus, CGK s traffic equals the total demand at GJMA with HLP and PCB s traffic deducted. As HLP handled over 3 MPA in 2015, it is assumed to have a capacity of 5 MPA in 2018 whilst the capacity of PCB remains constant at 1 MPA. Moreover, it is stated in Ministerial Decree 69/2013 that HLP would handle passengers up to 5 MPA. Table 12 summarises the pattern of traffic distribution between CGK, HLP and PCB if the preferred airports are HLP and PCB. In 2021, CGK, HLP and PCB should be able to serve the traffic at GJMA without any capacity constraints. From 2022 onwards, the demand will surpass the total available capacity and thus the new airport at GJMA is required to be operational. Table 12. Passenger traffic distribution: Condition 1, HLP-PCB preferred Phase Year GJMA Demand Total CGK CGK Share HLP HLP Share PCB PCB Share Extra Required Million Million Million Million Million Million Million Million Million Initial Phase Development Phase Ultimate Phase Condition-2: scenario 1, high demand, preferred airport: CGK In this condition, CGK becomes the major attraction for air traffic in GJMA and CGK s traffic share is assumed to be equal to its capacity. HLP and PCB will receive the spillover traffic from CGK. HLP and PCB s traffic equals the total demand at GJMA with CGK s 646

16 Studies, Vol.12, 2017 traffic deducted. As a primary domestic hub airport and the main gateway to Indonesia, CGK offers more routes and frequencies than HLP and PCB. Considering these factors, HLP and PCB become the least preferred airports and the traffic share at HLP and PCB depends upon GJMA s traffic demand and the remaining traffic left after subtracting it from CGK. Therefore, CGK will always function at its full capacity but HLP will have to experience conditions when the traffic is less than its capacity. Table 13 presents the pattern of traffic distribution between CGK, HLP and PCB if the preferred airport is CGK. Table 13. Passenger traffic distribution: Condition 2, CGK preferred Phase Year GJMA Demand Total CGK CGK Share HLP HLP Share PCB PCB Share Extra Required Million Million Million Million Million Million Million Million Million Initial Phase Development Phase Ultimate Phase Condition-3: scenario 3, high demand, preferred airport: HLP This is the condition that prevails if the government rejects opening PCB for commercial flights due to safety reasons and HLP cannot expand its terminal capacity due to military constraints. If HLP is the preferred airport, the traffic at HLP is assumed equal to its capacity and CGK receives the remaining traffic. Thus, CGK s traffic equals the total demand at GJMA with HLP s traffic deducted. Based on this condition, from 2021 onwards the demand will surpass the optimum capacity and thus a new airport for GJMA is required to be operational. The new airport is suggested to have an initial capacity of 30 MPA to accommodate the forecasted passenger numbers until In 2026, the terminal capacity needs to be expanded to handle the demand until Table 14 shows the pattern of traffic distribution between HLP and CGK if the preferred airport is HLP. Table 14. Passenger traffic distribution: Condition 3, HLP preferred Phase Year GJMA Demand Total CGK CGK Share HLP HLP Share Extra Required Million Million Million Million Million Million Million Initial Phase Development Phase Ultimate Phase Condition-4: scenario 3, high demand, preferred airport: CGK In this condition, CGK becomes the major attraction for air traffic in GJMA and CGK s traffic share is assumed to be equal to its capacity. HLP will receive the spillover traffic from 647

17 Studies, Vol.12, 2017 CGK. The traffic share at HLP depends upon GJMA s traffic demand and the remaining traffic left after subtracting CGK s traffic. Table 15 summarises the pattern of traffic distribution between HLP and CGK if the preferred airport is CGK. Table 15. Passenger traffic distribution: Condition 4, CGK preferred Phase Year GJMA Demand Total CGK CGK Share HLP HLP Share Extra Required Million Million Million Million Million Million Million Initial Phase Development Phase Ultimate Phase SWOT Analysis of GJMA s Multi-Airport Development In order to gain a comprehensive vision of the future development of a MAS in GJMA, the evaluation of Strengths, Weaknesses, Opportunities and Threats to optimise the development of the existing GJMA airports is summarised and presented below (Table 16). Table 16. SWOT analysis of the development of GJMA airports Strengths - The airports are located in the capital region of Indonesia - The airports offer flights to almost all provinces - Passenger unconstrained demand at GJMA is predicted at 124 million annual passengers in 2030 representing a 5.4% annual growth rate ( ) - CGK is a well known international gateway to Indonesia - The government supports CGK to develop its capacity up to the optimum - Public transport is well connected to the city centre (train/bus/taxi) - No air traffic congestion at HLP and PCB - HLP and PCB have sufficient infrastructures - Due to its geographical advantage, HLP can focus on absorbing demand from East GJMA whilst PCB can focus on demand from South GJMA Opportunities - Indonesia s characteristic as an archipelagic country relies heavily on air transport - Future potential demand for international flights due to the implementation of ASEAN Open Sky and free-entry visa policy - Relatively stable economic growth supports the purchasing power to travel using air transport - The government strongly supports CGK to expand its capacity - Significant growth of air traffic demand for the last 15 years Weaknesses - CGK and HLP experience terminal capacity issues - GJMA relies heavily on CGK - Constrained car parking at CGK - Severe road traffic to/from CGK during peak hours - Over dependence on the domestic market - HLP is a joint use airport with the military, therefore several restrictions apply - PCB is surrounded by a densely populated area - The development plan to expand the capacity of CGK and HLP is limited by many factors including land availability, environmental issues and budget constraints - BDR is a dedicated airport for flying schools and aircraft maintenance - The construction cost of a new major airport is estimated at 5.64 billion Threats - Future competition with other transport modes, particularly land transport, will shift the air transport demand - Government plan to develop a new airport in GJMA - Socio-economic and environmental issues arise in developing a green field airport - Airport development is vulnerable to political decision (national scope) - The development of HLP needs approval from the military 648

18 Studies, Vol.12, Good surface access and connectivity between CGK-HLP to the city centre provide passengers with flexibility to choose their preferred airport and widen the airport s catchment area - Garuda Indonesia is willing to invest at PCB - Airline s strategy to order no less than 500 additional aircraft - The air space issue arises against the plan to open PCB for scheduled commercial flights - PCB is not accommodated in the Ministerial Decree regarding the National Airport System - Land transport infrastructure issues to/from PCB and BDR - Airlines are reluctant to move to secondary airports 5. KEY RESEARCH FINDINGS & RECOMMENDATIONS 5.1 Key Research Findings Analysis of the air transport market in GJMA GJMA s air transport demand has a huge opportunity to develop due to the fact that only 34% of Indonesian residents today are using air transport for travelling. Moreover, around 75% of total passenger numbers are domestic passengers, which shows the great potential demand of the domestic market. GJMA operates two out of four airports serving scheduled commercial traffic through CGK and HLP which experience over-capacity with the demand-capacity ratio being 2.53 and 2.07 respectively The air transport demand forecast The forecasted results of air transport demand up to 2030 show that the future air transport demand in GJMA will grow % annually. The forecast model is divided into domestic and international models. The forecast for future air transport demand is conducted under three different growth scenarios: low, moderate and high. By using these assumptions, GJMA s passenger numbers in 2030 are predicted, ranging from million passengers. International passenger growth has a higher increment compared to domestic growth. Open sky agreements and the free-entry visa policy will stimulate the growth of international passengers. Therefore, the growth in domestic passenger numbers is forecasted to be steady at % per annum, slightly lower than the % annual international passenger growth The adequacy analysis of the airport capacity in the current airport system This research recognizes that the existing airport capacity is not able to accommodate the future demand of a maximum 152 million passengers. CGK and HLP have an existing total capacity of 43 MPA. Angkasa Pura II, as the airport operator of CGK and HLP, has planned the expansion of the airport terminal to a certain capacity. There is also a demand for PCB to be able to serve commercial traffic but there is an issue with air navigation. The other airport, BDR, is owned by the government and is aimed at serving training flights and aircraft maintenance, and thus cannot handle scheduled commercial flights Investigation of the potential development of a new airport in GJMA This research has identified and examined four possible capacity developments at GJMA airports. These four development scenarios conclude that the total capacity of GJMA airports can be expanded, starting from 63.5 MPA up to 93 MPA. Concerning the necessity of an 649

19 Studies, Vol.12, 2017 additional airport capacity in GJMA, a discussion should be undertaken between the military, the airport operator and government to decide the future development of HLP and PCB. In addition, the SWOT analysis shows that the construction of a new airport is necessary but the initial plan and time frame proposed by JICA should be reconsidered. JICA s study (2012) showed the need to operate a new airport in 2019 to accommodate the moderate growth of air transport demand. In contrast, the result of this research demonstrates that in order to handle the high growth rate of air transport demand, the operation of a new airport in GJMA can be delayed up till Strategic options for the development of the GJMA's airport system With the optimisation of CGK, HLP and PCB to handle high growth demand, it will take until 2022 before the GJMA will achieve its maximum capacity. If the military does not agree to develop HLP s terminal capacity and PCB s air space issue cannot be resolved then a new airport is required to operate in 2021 in order to accommodate the high growth demand. As there is no strong argument not to open PCB for handling commercial flights and delay the expansion of HLP s terminal capacity, the assumption to optimise CGK, HLP and PCB will be applied to these analyses. An intensive discussion should be undertaken between government, military and the airport operator to decide the future development of HLP and PCB as suggested by SWOT analysis. Therefore, the author recommends postponing the development of a new airport at least until 2022 and maximising the use of existing secondary airports in the metropolitan area such as HLP and PCB. This action requires a policy to revise the national airport system by accommodating PCB as proposed by SWOT analysis. Nevertheless, the increase in air transport demand should be monitored closely as the maximum capacity could be achieved earlier or later than forecasted. Flexible strategic planning is proposed to avoid financially embarrassing failures due to the massive investment required for a completely new airport. 5.2 Conclusions This chapter summarizes conclusions based on the analyses described previously: 1. The development of CGK should consider the growth impact from international flights whilst HLP s development should be more focused on domestic flights. 2. It is important to keep HLP open for commercial flights. Therefore, a discussion should be undertaken between government and the military so that the airport operator is allowed to expand the terminal capacity and the air traffic slots. 3. Ministerial Decree 69/2013 about the National Airport System has to be revised to accommodate PCB. The cross air space issue between PCB and HLP could be resolved without neglecting safety aspects. Thus the utmost airport capacity for the GJMA can be achieved. 4. The optimisation of all existing airports in the GJMA is strongly encouraged in order to avoid unnecessary investment through the construction of a new major airport. This new airport plan should be regarded as the last option. 5. The operation of the new airport can be postponed until 2022 at the earliest, but if the number of passengers is lower than the forecast then this date could be delayed until such time as the forecast is reached. 650

20 Studies, Vol.12, 2017 REFERENCES Başar, G. and Bhat, C. (2004). A parameterized consideration set model for airport choice: an application to the San Francisco Bay Area. Transportation Research Part B: Methodological, 38(10), pp Bonnefoy, P. (2008). Scalability of the Air Transportation System and Development of Multi-Airport Systems: A Worldwide Perspective. PhD, Massachusetts Institute of Technology. Bonnefoy, P., de Neufville, R. and Hansman, R. (2010). Evolution and Development of Multiairport Systems: Worldwide Perspective. Journal of Transportation Engineering, 136(11), pp Brooke, A., Caves, R. and Pitfield, D. (1994). Methodology for predicting European short-haul air transport demand from regional airports. Journal of Air Transport Management, 1(1), pp de Neufville, R. (1995). Management of multi-airport systems. Journal of Air Transport Management, 2(2), pp de Neufville, R. (2000). Planning multi-airport systems in metropolitan regions in the 1990s. [online] Cambridge: Massachusetts Institute of Technology. Available at: PDF [Accessed 6 Jun. 2016]. de Neufville, R., Odoni, A., Belobaba, P. and Reynolds, T. (2013). Airport systems: planning, design, and management (2 nd Ed.). New York: McGraw-Hill. Federal Aviation Administration (2007a). Terminal area forecasts database. [online] Available at: Federal Aviation Administration (2007b). AC 150/5070-6B. Change 1 to Airport Master Plans. [online] Washington, D. C.: U.S. Government Printing Office. Available at: b_chg1.pdf [Accessed 6 Jun. 2016]. Federal Aviation Administration (2013). Understanding Airline and Passenger Choice in Multi-Airport Regions. ACRP Report 98. [online] Washington, DC: Transportation Research Board. Available at: pdf [Accessed 2 Jun. 2016]. Graham, A. (2014). Managing Airports. 4 th ed. London: Routledge. Horonjeff, R., McKelvey, F. X., Sproule, W. J. and Young, S. B. (2010). Planning and design of airports (5 th Ed). New York: McGraw-Hill. IATA (2004). Airport Development Reference Manual. Montreal: IATA. ICAO (2008). Airports core service database. [online] Available at: JICA (2012). Project for the master plan study on multiple-airport development for Greater Jakarta Metropolitan Area in the Republic of Indonesia. Final Report. [online] JICA. Available at: _01.pdf [Accessed 10 Jun. 2016]. Kwakkel, J. H. (2008). The problem of uncertainty in airport master planning. e-zine Edition, editor, Aerlines Magazine, 39. Micco, A. and Serebrisky, T. (2006). Competition regimes and air transport costs: The effects of open skies agreements. Journal of International Economics, 70(1), pp MoT (2013). Ministerial Decree No.69/2013: National Airport System. Jakarta: Ministry of Transport. Wells, A. and Young, S. (2004). Airport planning & management. New York: McGraw-Hill. 651